| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 4.956190476190476, |
| "eval_steps": 500, |
| "global_step": 325, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.015238095238095238, |
| "grad_norm": 6.41825703019793, |
| "learning_rate": 2.4242424242424244e-06, |
| "loss": 1.0212, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.030476190476190476, |
| "grad_norm": 6.32460329298527, |
| "learning_rate": 4.848484848484849e-06, |
| "loss": 1.0078, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.045714285714285714, |
| "grad_norm": 5.823493836887707, |
| "learning_rate": 7.272727272727273e-06, |
| "loss": 0.9926, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.06095238095238095, |
| "grad_norm": 4.215092249626398, |
| "learning_rate": 9.696969696969698e-06, |
| "loss": 0.9511, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.0761904761904762, |
| "grad_norm": 2.755821783551069, |
| "learning_rate": 1.2121212121212122e-05, |
| "loss": 0.924, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.09142857142857143, |
| "grad_norm": 5.79854862437273, |
| "learning_rate": 1.4545454545454546e-05, |
| "loss": 0.9436, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.10666666666666667, |
| "grad_norm": 6.970153409745764, |
| "learning_rate": 1.6969696969696972e-05, |
| "loss": 0.9199, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.1219047619047619, |
| "grad_norm": 6.398045150850651, |
| "learning_rate": 1.9393939393939395e-05, |
| "loss": 0.9368, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.13714285714285715, |
| "grad_norm": 3.7323118823746717, |
| "learning_rate": 2.1818181818181818e-05, |
| "loss": 0.8624, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.1523809523809524, |
| "grad_norm": 2.4649345087328935, |
| "learning_rate": 2.4242424242424244e-05, |
| "loss": 0.8177, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.1676190476190476, |
| "grad_norm": 2.624350847053939, |
| "learning_rate": 2.6666666666666667e-05, |
| "loss": 0.7966, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.18285714285714286, |
| "grad_norm": 1.495708592085656, |
| "learning_rate": 2.9090909090909093e-05, |
| "loss": 0.7813, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.1980952380952381, |
| "grad_norm": 1.4871227985381805, |
| "learning_rate": 3.151515151515152e-05, |
| "loss": 0.7453, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.21333333333333335, |
| "grad_norm": 2.060287535099846, |
| "learning_rate": 3.3939393939393945e-05, |
| "loss": 0.7371, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.22857142857142856, |
| "grad_norm": 1.4326857447271233, |
| "learning_rate": 3.6363636363636364e-05, |
| "loss": 0.7234, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.2438095238095238, |
| "grad_norm": 2.1165784341208105, |
| "learning_rate": 3.878787878787879e-05, |
| "loss": 0.7143, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.259047619047619, |
| "grad_norm": 1.7376117944761271, |
| "learning_rate": 4.1212121212121216e-05, |
| "loss": 0.7091, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.2742857142857143, |
| "grad_norm": 1.9235494481574669, |
| "learning_rate": 4.3636363636363636e-05, |
| "loss": 0.7033, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.2895238095238095, |
| "grad_norm": 1.2885542410242774, |
| "learning_rate": 4.606060606060607e-05, |
| "loss": 0.693, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.3047619047619048, |
| "grad_norm": 2.0682542365556746, |
| "learning_rate": 4.848484848484849e-05, |
| "loss": 0.6782, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.32, |
| "grad_norm": 1.1294865223043438, |
| "learning_rate": 5.0909090909090914e-05, |
| "loss": 0.6734, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.3352380952380952, |
| "grad_norm": 1.6892295840685134, |
| "learning_rate": 5.333333333333333e-05, |
| "loss": 0.6701, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.3504761904761905, |
| "grad_norm": 1.7015145087043497, |
| "learning_rate": 5.5757575757575766e-05, |
| "loss": 0.6747, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.3657142857142857, |
| "grad_norm": 1.105513245543281, |
| "learning_rate": 5.8181818181818185e-05, |
| "loss": 0.6569, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.38095238095238093, |
| "grad_norm": 1.9301412200235886, |
| "learning_rate": 6.060606060606061e-05, |
| "loss": 0.6672, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.3961904761904762, |
| "grad_norm": 1.5372986311163992, |
| "learning_rate": 6.303030303030304e-05, |
| "loss": 0.6575, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.4114285714285714, |
| "grad_norm": 1.841770844497413, |
| "learning_rate": 6.545454545454546e-05, |
| "loss": 0.6439, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.4266666666666667, |
| "grad_norm": 1.285994127925713, |
| "learning_rate": 6.787878787878789e-05, |
| "loss": 0.6421, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.4419047619047619, |
| "grad_norm": 1.306641316576376, |
| "learning_rate": 7.03030303030303e-05, |
| "loss": 0.6446, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.45714285714285713, |
| "grad_norm": 2.060289337588267, |
| "learning_rate": 7.272727272727273e-05, |
| "loss": 0.6418, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.4723809523809524, |
| "grad_norm": 1.5995121846928928, |
| "learning_rate": 7.515151515151517e-05, |
| "loss": 0.6318, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.4876190476190476, |
| "grad_norm": 1.5361156868763157, |
| "learning_rate": 7.757575757575758e-05, |
| "loss": 0.6356, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.5028571428571429, |
| "grad_norm": 1.9790958633956652, |
| "learning_rate": 8e-05, |
| "loss": 0.6315, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.518095238095238, |
| "grad_norm": 1.0983212679170282, |
| "learning_rate": 7.999768495280586e-05, |
| "loss": 0.6212, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.5333333333333333, |
| "grad_norm": 2.356052428059575, |
| "learning_rate": 7.999074007919565e-05, |
| "loss": 0.6366, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.5485714285714286, |
| "grad_norm": 1.9118106657858278, |
| "learning_rate": 7.997916618305483e-05, |
| "loss": 0.6337, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.5638095238095238, |
| "grad_norm": 2.086875421088417, |
| "learning_rate": 7.996296460408921e-05, |
| "loss": 0.6243, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.579047619047619, |
| "grad_norm": 1.6815412319889804, |
| "learning_rate": 7.994213721766979e-05, |
| "loss": 0.6222, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.5942857142857143, |
| "grad_norm": 1.9619509595749545, |
| "learning_rate": 7.99166864346157e-05, |
| "loss": 0.6185, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.6095238095238096, |
| "grad_norm": 1.1850043703613138, |
| "learning_rate": 7.988661520091513e-05, |
| "loss": 0.6222, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.6247619047619047, |
| "grad_norm": 1.5413225249585591, |
| "learning_rate": 7.985192699738432e-05, |
| "loss": 0.6131, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.64, |
| "grad_norm": 2.2994933443512697, |
| "learning_rate": 7.981262583926472e-05, |
| "loss": 0.6326, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.6552380952380953, |
| "grad_norm": 0.8610367129145644, |
| "learning_rate": 7.976871627575808e-05, |
| "loss": 0.6057, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.6704761904761904, |
| "grad_norm": 2.630926210907478, |
| "learning_rate": 7.972020338950004e-05, |
| "loss": 0.6366, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.6857142857142857, |
| "grad_norm": 1.542557855004854, |
| "learning_rate": 7.96670927959716e-05, |
| "loss": 0.624, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.700952380952381, |
| "grad_norm": 1.9643148826366752, |
| "learning_rate": 7.960939064284934e-05, |
| "loss": 0.619, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.7161904761904762, |
| "grad_norm": 1.248523269079334, |
| "learning_rate": 7.954710360929362e-05, |
| "loss": 0.6108, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.7314285714285714, |
| "grad_norm": 1.225316205009982, |
| "learning_rate": 7.948023890517557e-05, |
| "loss": 0.6126, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.7466666666666667, |
| "grad_norm": 1.4200551325493893, |
| "learning_rate": 7.940880427024243e-05, |
| "loss": 0.6093, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.7619047619047619, |
| "grad_norm": 1.3484644010507392, |
| "learning_rate": 7.933280797322181e-05, |
| "loss": 0.5995, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.7771428571428571, |
| "grad_norm": 1.317639647361387, |
| "learning_rate": 7.925225881086437e-05, |
| "loss": 0.5967, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.7923809523809524, |
| "grad_norm": 1.5340582175990225, |
| "learning_rate": 7.916716610692578e-05, |
| "loss": 0.5981, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.8076190476190476, |
| "grad_norm": 1.0386828999827622, |
| "learning_rate": 7.907753971108728e-05, |
| "loss": 0.5896, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.8228571428571428, |
| "grad_norm": 1.5332648656890953, |
| "learning_rate": 7.898338999781567e-05, |
| "loss": 0.589, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.8380952380952381, |
| "grad_norm": 1.1339766871931323, |
| "learning_rate": 7.888472786516246e-05, |
| "loss": 0.5937, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.8533333333333334, |
| "grad_norm": 1.5164034405411189, |
| "learning_rate": 7.87815647335023e-05, |
| "loss": 0.5883, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.8685714285714285, |
| "grad_norm": 0.9700436403678562, |
| "learning_rate": 7.86739125442111e-05, |
| "loss": 0.5931, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.8838095238095238, |
| "grad_norm": 1.3463613799484717, |
| "learning_rate": 7.856178375828384e-05, |
| "loss": 0.589, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.8990476190476191, |
| "grad_norm": 0.8752856535881799, |
| "learning_rate": 7.844519135489204e-05, |
| "loss": 0.5754, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.9142857142857143, |
| "grad_norm": 1.1854440512880053, |
| "learning_rate": 7.832414882988153e-05, |
| "loss": 0.5894, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.9295238095238095, |
| "grad_norm": 1.7051861411911757, |
| "learning_rate": 7.819867019421023e-05, |
| "loss": 0.5885, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.9447619047619048, |
| "grad_norm": 0.6380913241526577, |
| "learning_rate": 7.806876997232625e-05, |
| "loss": 0.5834, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.96, |
| "grad_norm": 1.3953430391233206, |
| "learning_rate": 7.793446320048687e-05, |
| "loss": 0.5783, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.9752380952380952, |
| "grad_norm": 1.0079170376946163, |
| "learning_rate": 7.779576542501781e-05, |
| "loss": 0.5839, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.9904761904761905, |
| "grad_norm": 1.1878161560623453, |
| "learning_rate": 7.765269270051389e-05, |
| "loss": 0.575, |
| "step": 65 |
| }, |
| { |
| "epoch": 1.0066666666666666, |
| "grad_norm": 1.495861192796852, |
| "learning_rate": 7.750526158798056e-05, |
| "loss": 0.8303, |
| "step": 66 |
| }, |
| { |
| "epoch": 1.0219047619047619, |
| "grad_norm": 1.3683956168224771, |
| "learning_rate": 7.735348915291701e-05, |
| "loss": 0.571, |
| "step": 67 |
| }, |
| { |
| "epoch": 1.0371428571428571, |
| "grad_norm": 0.8831683489698775, |
| "learning_rate": 7.719739296334073e-05, |
| "loss": 0.5548, |
| "step": 68 |
| }, |
| { |
| "epoch": 1.0523809523809524, |
| "grad_norm": 0.9884633873935289, |
| "learning_rate": 7.7036991087754e-05, |
| "loss": 0.5619, |
| "step": 69 |
| }, |
| { |
| "epoch": 1.0676190476190477, |
| "grad_norm": 1.658491152165491, |
| "learning_rate": 7.687230209305242e-05, |
| "loss": 0.5876, |
| "step": 70 |
| }, |
| { |
| "epoch": 1.0828571428571427, |
| "grad_norm": 0.8352075307968011, |
| "learning_rate": 7.670334504237575e-05, |
| "loss": 0.5685, |
| "step": 71 |
| }, |
| { |
| "epoch": 1.098095238095238, |
| "grad_norm": 1.429719144977111, |
| "learning_rate": 7.65301394929013e-05, |
| "loss": 0.5764, |
| "step": 72 |
| }, |
| { |
| "epoch": 1.1133333333333333, |
| "grad_norm": 0.9027318597099093, |
| "learning_rate": 7.635270549358012e-05, |
| "loss": 0.5602, |
| "step": 73 |
| }, |
| { |
| "epoch": 1.1285714285714286, |
| "grad_norm": 1.0536650805805048, |
| "learning_rate": 7.617106358281634e-05, |
| "loss": 0.5765, |
| "step": 74 |
| }, |
| { |
| "epoch": 1.1438095238095238, |
| "grad_norm": 0.8336607908560256, |
| "learning_rate": 7.598523478608974e-05, |
| "loss": 0.5583, |
| "step": 75 |
| }, |
| { |
| "epoch": 1.159047619047619, |
| "grad_norm": 0.7166113488943966, |
| "learning_rate": 7.579524061352206e-05, |
| "loss": 0.551, |
| "step": 76 |
| }, |
| { |
| "epoch": 1.1742857142857144, |
| "grad_norm": 0.8105369080449557, |
| "learning_rate": 7.560110305738708e-05, |
| "loss": 0.5659, |
| "step": 77 |
| }, |
| { |
| "epoch": 1.1895238095238094, |
| "grad_norm": 0.6283210931425559, |
| "learning_rate": 7.540284458956503e-05, |
| "loss": 0.5379, |
| "step": 78 |
| }, |
| { |
| "epoch": 1.2047619047619047, |
| "grad_norm": 0.8126105033449552, |
| "learning_rate": 7.520048815894144e-05, |
| "loss": 0.5476, |
| "step": 79 |
| }, |
| { |
| "epoch": 1.22, |
| "grad_norm": 0.5252506521520861, |
| "learning_rate": 7.49940571887506e-05, |
| "loss": 0.5551, |
| "step": 80 |
| }, |
| { |
| "epoch": 1.2352380952380952, |
| "grad_norm": 0.6609919242663752, |
| "learning_rate": 7.478357557386444e-05, |
| "loss": 0.5407, |
| "step": 81 |
| }, |
| { |
| "epoch": 1.2504761904761905, |
| "grad_norm": 0.576238102776681, |
| "learning_rate": 7.456906767802658e-05, |
| "loss": 0.5465, |
| "step": 82 |
| }, |
| { |
| "epoch": 1.2657142857142858, |
| "grad_norm": 0.5233988001346253, |
| "learning_rate": 7.435055833103213e-05, |
| "loss": 0.55, |
| "step": 83 |
| }, |
| { |
| "epoch": 1.2809523809523808, |
| "grad_norm": 0.44023553757218015, |
| "learning_rate": 7.412807282585362e-05, |
| "loss": 0.5412, |
| "step": 84 |
| }, |
| { |
| "epoch": 1.2961904761904761, |
| "grad_norm": 0.4831178702636027, |
| "learning_rate": 7.390163691571325e-05, |
| "loss": 0.5398, |
| "step": 85 |
| }, |
| { |
| "epoch": 1.3114285714285714, |
| "grad_norm": 0.454634628403172, |
| "learning_rate": 7.367127681110199e-05, |
| "loss": 0.534, |
| "step": 86 |
| }, |
| { |
| "epoch": 1.3266666666666667, |
| "grad_norm": 0.44419681534445155, |
| "learning_rate": 7.343701917674547e-05, |
| "loss": 0.5421, |
| "step": 87 |
| }, |
| { |
| "epoch": 1.341904761904762, |
| "grad_norm": 0.4888479476988804, |
| "learning_rate": 7.319889112851771e-05, |
| "loss": 0.5371, |
| "step": 88 |
| }, |
| { |
| "epoch": 1.3571428571428572, |
| "grad_norm": 0.7143397739870562, |
| "learning_rate": 7.295692023030217e-05, |
| "loss": 0.5377, |
| "step": 89 |
| }, |
| { |
| "epoch": 1.3723809523809525, |
| "grad_norm": 0.9465106459896011, |
| "learning_rate": 7.271113449080131e-05, |
| "loss": 0.5479, |
| "step": 90 |
| }, |
| { |
| "epoch": 1.3876190476190478, |
| "grad_norm": 1.2678262246780991, |
| "learning_rate": 7.246156236029445e-05, |
| "loss": 0.5452, |
| "step": 91 |
| }, |
| { |
| "epoch": 1.4028571428571428, |
| "grad_norm": 0.6281248994163343, |
| "learning_rate": 7.220823272734461e-05, |
| "loss": 0.5401, |
| "step": 92 |
| }, |
| { |
| "epoch": 1.418095238095238, |
| "grad_norm": 0.34247752257094766, |
| "learning_rate": 7.19511749154546e-05, |
| "loss": 0.5394, |
| "step": 93 |
| }, |
| { |
| "epoch": 1.4333333333333333, |
| "grad_norm": 0.7014077226631056, |
| "learning_rate": 7.169041867967272e-05, |
| "loss": 0.5407, |
| "step": 94 |
| }, |
| { |
| "epoch": 1.4485714285714286, |
| "grad_norm": 0.9221215435654039, |
| "learning_rate": 7.142599420314857e-05, |
| "loss": 0.5464, |
| "step": 95 |
| }, |
| { |
| "epoch": 1.4638095238095237, |
| "grad_norm": 0.8724875653280793, |
| "learning_rate": 7.115793209363929e-05, |
| "loss": 0.5379, |
| "step": 96 |
| }, |
| { |
| "epoch": 1.479047619047619, |
| "grad_norm": 0.734302784896464, |
| "learning_rate": 7.088626337996657e-05, |
| "loss": 0.5248, |
| "step": 97 |
| }, |
| { |
| "epoch": 1.4942857142857142, |
| "grad_norm": 0.6804548302488894, |
| "learning_rate": 7.06110195084251e-05, |
| "loss": 0.5403, |
| "step": 98 |
| }, |
| { |
| "epoch": 1.5095238095238095, |
| "grad_norm": 0.8180102079360684, |
| "learning_rate": 7.03322323391425e-05, |
| "loss": 0.5356, |
| "step": 99 |
| }, |
| { |
| "epoch": 1.5247619047619048, |
| "grad_norm": 0.7089574204038593, |
| "learning_rate": 7.004993414239147e-05, |
| "loss": 0.5367, |
| "step": 100 |
| }, |
| { |
| "epoch": 1.54, |
| "grad_norm": 0.4534978500691297, |
| "learning_rate": 6.976415759485443e-05, |
| "loss": 0.5301, |
| "step": 101 |
| }, |
| { |
| "epoch": 1.5552380952380953, |
| "grad_norm": 0.5791842384826449, |
| "learning_rate": 6.947493577584111e-05, |
| "loss": 0.5338, |
| "step": 102 |
| }, |
| { |
| "epoch": 1.5704761904761906, |
| "grad_norm": 0.6373667696236239, |
| "learning_rate": 6.918230216345951e-05, |
| "loss": 0.5255, |
| "step": 103 |
| }, |
| { |
| "epoch": 1.5857142857142859, |
| "grad_norm": 0.5813750117468534, |
| "learning_rate": 6.888629063074082e-05, |
| "loss": 0.5209, |
| "step": 104 |
| }, |
| { |
| "epoch": 1.600952380952381, |
| "grad_norm": 0.5731639034091425, |
| "learning_rate": 6.858693544171845e-05, |
| "loss": 0.5215, |
| "step": 105 |
| }, |
| { |
| "epoch": 1.6161904761904762, |
| "grad_norm": 0.5351450945596543, |
| "learning_rate": 6.828427124746191e-05, |
| "loss": 0.5301, |
| "step": 106 |
| }, |
| { |
| "epoch": 1.6314285714285715, |
| "grad_norm": 0.6016816279659157, |
| "learning_rate": 6.797833308206588e-05, |
| "loss": 0.529, |
| "step": 107 |
| }, |
| { |
| "epoch": 1.6466666666666665, |
| "grad_norm": 0.7659521464895882, |
| "learning_rate": 6.766915635859497e-05, |
| "loss": 0.533, |
| "step": 108 |
| }, |
| { |
| "epoch": 1.6619047619047618, |
| "grad_norm": 0.9630524913355004, |
| "learning_rate": 6.735677686498443e-05, |
| "loss": 0.5297, |
| "step": 109 |
| }, |
| { |
| "epoch": 1.677142857142857, |
| "grad_norm": 1.1492200252998914, |
| "learning_rate": 6.70412307598978e-05, |
| "loss": 0.5264, |
| "step": 110 |
| }, |
| { |
| "epoch": 1.6923809523809523, |
| "grad_norm": 0.5857314920473611, |
| "learning_rate": 6.672255456854135e-05, |
| "loss": 0.5241, |
| "step": 111 |
| }, |
| { |
| "epoch": 1.7076190476190476, |
| "grad_norm": 0.48173626706101424, |
| "learning_rate": 6.640078517843619e-05, |
| "loss": 0.5346, |
| "step": 112 |
| }, |
| { |
| "epoch": 1.7228571428571429, |
| "grad_norm": 0.7730414410041352, |
| "learning_rate": 6.60759598351485e-05, |
| "loss": 0.5229, |
| "step": 113 |
| }, |
| { |
| "epoch": 1.7380952380952381, |
| "grad_norm": 0.6363250099986629, |
| "learning_rate": 6.57481161379783e-05, |
| "loss": 0.5243, |
| "step": 114 |
| }, |
| { |
| "epoch": 1.7533333333333334, |
| "grad_norm": 0.45624023704141803, |
| "learning_rate": 6.54172920356071e-05, |
| "loss": 0.5247, |
| "step": 115 |
| }, |
| { |
| "epoch": 1.7685714285714287, |
| "grad_norm": 0.4715396085670591, |
| "learning_rate": 6.508352582170543e-05, |
| "loss": 0.5144, |
| "step": 116 |
| }, |
| { |
| "epoch": 1.783809523809524, |
| "grad_norm": 0.5945001964397519, |
| "learning_rate": 6.474685613050014e-05, |
| "loss": 0.5279, |
| "step": 117 |
| }, |
| { |
| "epoch": 1.799047619047619, |
| "grad_norm": 0.5588842197173406, |
| "learning_rate": 6.440732193230241e-05, |
| "loss": 0.522, |
| "step": 118 |
| }, |
| { |
| "epoch": 1.8142857142857143, |
| "grad_norm": 0.36397509002448036, |
| "learning_rate": 6.40649625289969e-05, |
| "loss": 0.5206, |
| "step": 119 |
| }, |
| { |
| "epoch": 1.8295238095238096, |
| "grad_norm": 0.5225246500798869, |
| "learning_rate": 6.371981754949241e-05, |
| "loss": 0.5233, |
| "step": 120 |
| }, |
| { |
| "epoch": 1.8447619047619046, |
| "grad_norm": 0.6419719236551851, |
| "learning_rate": 6.337192694513474e-05, |
| "loss": 0.5265, |
| "step": 121 |
| }, |
| { |
| "epoch": 1.8599999999999999, |
| "grad_norm": 0.41672473350375433, |
| "learning_rate": 6.30213309850823e-05, |
| "loss": 0.5165, |
| "step": 122 |
| }, |
| { |
| "epoch": 1.8752380952380951, |
| "grad_norm": 0.29619919377664744, |
| "learning_rate": 6.266807025164472e-05, |
| "loss": 0.5158, |
| "step": 123 |
| }, |
| { |
| "epoch": 1.8904761904761904, |
| "grad_norm": 0.41731867813128104, |
| "learning_rate": 6.231218563558551e-05, |
| "loss": 0.518, |
| "step": 124 |
| }, |
| { |
| "epoch": 1.9057142857142857, |
| "grad_norm": 0.3648760435140768, |
| "learning_rate": 6.195371833138878e-05, |
| "loss": 0.5165, |
| "step": 125 |
| }, |
| { |
| "epoch": 1.920952380952381, |
| "grad_norm": 0.3531467699123105, |
| "learning_rate": 6.159270983249084e-05, |
| "loss": 0.5247, |
| "step": 126 |
| }, |
| { |
| "epoch": 1.9361904761904762, |
| "grad_norm": 0.45588003744063627, |
| "learning_rate": 6.122920192647734e-05, |
| "loss": 0.5197, |
| "step": 127 |
| }, |
| { |
| "epoch": 1.9514285714285715, |
| "grad_norm": 0.5510619573466758, |
| "learning_rate": 6.086323669024616e-05, |
| "loss": 0.5256, |
| "step": 128 |
| }, |
| { |
| "epoch": 1.9666666666666668, |
| "grad_norm": 0.5646853258922581, |
| "learning_rate": 6.049485648513696e-05, |
| "loss": 0.527, |
| "step": 129 |
| }, |
| { |
| "epoch": 1.981904761904762, |
| "grad_norm": 0.4479646401937566, |
| "learning_rate": 6.012410395202774e-05, |
| "loss": 0.5234, |
| "step": 130 |
| }, |
| { |
| "epoch": 1.997142857142857, |
| "grad_norm": 0.575986791359234, |
| "learning_rate": 5.9751022006399095e-05, |
| "loss": 0.7009, |
| "step": 131 |
| }, |
| { |
| "epoch": 2.013333333333333, |
| "grad_norm": 0.814219365805337, |
| "learning_rate": 5.9375653833366585e-05, |
| "loss": 0.5404, |
| "step": 132 |
| }, |
| { |
| "epoch": 2.0285714285714285, |
| "grad_norm": 0.8146199046124171, |
| "learning_rate": 5.8998042882682024e-05, |
| "loss": 0.4877, |
| "step": 133 |
| }, |
| { |
| "epoch": 2.0438095238095237, |
| "grad_norm": 0.7629490780519704, |
| "learning_rate": 5.861823286370396e-05, |
| "loss": 0.4849, |
| "step": 134 |
| }, |
| { |
| "epoch": 2.059047619047619, |
| "grad_norm": 0.6570978207497692, |
| "learning_rate": 5.823626774033836e-05, |
| "loss": 0.4887, |
| "step": 135 |
| }, |
| { |
| "epoch": 2.0742857142857143, |
| "grad_norm": 0.5305731216336181, |
| "learning_rate": 5.7852191725949564e-05, |
| "loss": 0.4881, |
| "step": 136 |
| }, |
| { |
| "epoch": 2.0895238095238096, |
| "grad_norm": 0.512493977688762, |
| "learning_rate": 5.746604927824257e-05, |
| "loss": 0.4844, |
| "step": 137 |
| }, |
| { |
| "epoch": 2.104761904761905, |
| "grad_norm": 0.4828622337949408, |
| "learning_rate": 5.707788509411685e-05, |
| "loss": 0.4887, |
| "step": 138 |
| }, |
| { |
| "epoch": 2.12, |
| "grad_norm": 0.4884027438000281, |
| "learning_rate": 5.668774410449268e-05, |
| "loss": 0.4908, |
| "step": 139 |
| }, |
| { |
| "epoch": 2.1352380952380954, |
| "grad_norm": 0.42776751878992564, |
| "learning_rate": 5.629567146911023e-05, |
| "loss": 0.4872, |
| "step": 140 |
| }, |
| { |
| "epoch": 2.1504761904761907, |
| "grad_norm": 0.36115257273207996, |
| "learning_rate": 5.590171257130225e-05, |
| "loss": 0.4861, |
| "step": 141 |
| }, |
| { |
| "epoch": 2.1657142857142855, |
| "grad_norm": 0.38436733504559367, |
| "learning_rate": 5.550591301274073e-05, |
| "loss": 0.4844, |
| "step": 142 |
| }, |
| { |
| "epoch": 2.1809523809523808, |
| "grad_norm": 0.3435583779900254, |
| "learning_rate": 5.5108318608158595e-05, |
| "loss": 0.4836, |
| "step": 143 |
| }, |
| { |
| "epoch": 2.196190476190476, |
| "grad_norm": 0.2857883944660014, |
| "learning_rate": 5.4708975380046356e-05, |
| "loss": 0.4796, |
| "step": 144 |
| }, |
| { |
| "epoch": 2.2114285714285713, |
| "grad_norm": 0.31512140613431383, |
| "learning_rate": 5.430792955332503e-05, |
| "loss": 0.4863, |
| "step": 145 |
| }, |
| { |
| "epoch": 2.2266666666666666, |
| "grad_norm": 0.31769228977835245, |
| "learning_rate": 5.3905227549995366e-05, |
| "loss": 0.4809, |
| "step": 146 |
| }, |
| { |
| "epoch": 2.241904761904762, |
| "grad_norm": 0.27924100773876076, |
| "learning_rate": 5.350091598376454e-05, |
| "loss": 0.4805, |
| "step": 147 |
| }, |
| { |
| "epoch": 2.257142857142857, |
| "grad_norm": 0.2662485345450648, |
| "learning_rate": 5.3095041654650384e-05, |
| "loss": 0.4831, |
| "step": 148 |
| }, |
| { |
| "epoch": 2.2723809523809524, |
| "grad_norm": 0.23461248952656025, |
| "learning_rate": 5.268765154356425e-05, |
| "loss": 0.4816, |
| "step": 149 |
| }, |
| { |
| "epoch": 2.2876190476190477, |
| "grad_norm": 0.240243576104443, |
| "learning_rate": 5.227879280687281e-05, |
| "loss": 0.4788, |
| "step": 150 |
| }, |
| { |
| "epoch": 2.302857142857143, |
| "grad_norm": 0.24392104367270967, |
| "learning_rate": 5.186851277093961e-05, |
| "loss": 0.4851, |
| "step": 151 |
| }, |
| { |
| "epoch": 2.318095238095238, |
| "grad_norm": 0.19828908319240954, |
| "learning_rate": 5.1456858926646974e-05, |
| "loss": 0.4817, |
| "step": 152 |
| }, |
| { |
| "epoch": 2.3333333333333335, |
| "grad_norm": 0.2354905746190516, |
| "learning_rate": 5.104387892389877e-05, |
| "loss": 0.4816, |
| "step": 153 |
| }, |
| { |
| "epoch": 2.3485714285714288, |
| "grad_norm": 0.24408900516772156, |
| "learning_rate": 5.062962056610478e-05, |
| "loss": 0.4833, |
| "step": 154 |
| }, |
| { |
| "epoch": 2.3638095238095236, |
| "grad_norm": 0.22588988217182931, |
| "learning_rate": 5.021413180464748e-05, |
| "loss": 0.4884, |
| "step": 155 |
| }, |
| { |
| "epoch": 2.379047619047619, |
| "grad_norm": 0.20985654308992252, |
| "learning_rate": 4.979746073333145e-05, |
| "loss": 0.4777, |
| "step": 156 |
| }, |
| { |
| "epoch": 2.394285714285714, |
| "grad_norm": 0.24110384355800776, |
| "learning_rate": 4.9379655582816396e-05, |
| "loss": 0.4867, |
| "step": 157 |
| }, |
| { |
| "epoch": 2.4095238095238094, |
| "grad_norm": 0.20401637758198887, |
| "learning_rate": 4.896076471503439e-05, |
| "loss": 0.4766, |
| "step": 158 |
| }, |
| { |
| "epoch": 2.4247619047619047, |
| "grad_norm": 0.19932449853927872, |
| "learning_rate": 4.854083661759185e-05, |
| "loss": 0.4884, |
| "step": 159 |
| }, |
| { |
| "epoch": 2.44, |
| "grad_norm": 0.24533376047455172, |
| "learning_rate": 4.8119919898156934e-05, |
| "loss": 0.4834, |
| "step": 160 |
| }, |
| { |
| "epoch": 2.455238095238095, |
| "grad_norm": 0.19336383545060395, |
| "learning_rate": 4.76980632788332e-05, |
| "loss": 0.4779, |
| "step": 161 |
| }, |
| { |
| "epoch": 2.4704761904761905, |
| "grad_norm": 0.19824370981981676, |
| "learning_rate": 4.7275315590519765e-05, |
| "loss": 0.4823, |
| "step": 162 |
| }, |
| { |
| "epoch": 2.4857142857142858, |
| "grad_norm": 0.20431330071012052, |
| "learning_rate": 4.6851725767259106e-05, |
| "loss": 0.4781, |
| "step": 163 |
| }, |
| { |
| "epoch": 2.500952380952381, |
| "grad_norm": 0.2082524151757895, |
| "learning_rate": 4.642734284057283e-05, |
| "loss": 0.4755, |
| "step": 164 |
| }, |
| { |
| "epoch": 2.5161904761904763, |
| "grad_norm": 0.2129960852067495, |
| "learning_rate": 4.600221593378611e-05, |
| "loss": 0.4796, |
| "step": 165 |
| }, |
| { |
| "epoch": 2.5314285714285716, |
| "grad_norm": 0.17147652502249622, |
| "learning_rate": 4.5576394256341596e-05, |
| "loss": 0.4784, |
| "step": 166 |
| }, |
| { |
| "epoch": 2.546666666666667, |
| "grad_norm": 0.2328806782705143, |
| "learning_rate": 4.514992709810324e-05, |
| "loss": 0.4839, |
| "step": 167 |
| }, |
| { |
| "epoch": 2.5619047619047617, |
| "grad_norm": 0.2628457128916824, |
| "learning_rate": 4.472286382365096e-05, |
| "loss": 0.4899, |
| "step": 168 |
| }, |
| { |
| "epoch": 2.5771428571428574, |
| "grad_norm": 0.20530497111629914, |
| "learning_rate": 4.429525386656653e-05, |
| "loss": 0.4865, |
| "step": 169 |
| }, |
| { |
| "epoch": 2.5923809523809522, |
| "grad_norm": 0.18615860925149122, |
| "learning_rate": 4.38671467237115e-05, |
| "loss": 0.4841, |
| "step": 170 |
| }, |
| { |
| "epoch": 2.6076190476190475, |
| "grad_norm": 0.181524499480342, |
| "learning_rate": 4.343859194949787e-05, |
| "loss": 0.4792, |
| "step": 171 |
| }, |
| { |
| "epoch": 2.6228571428571428, |
| "grad_norm": 0.2025029996461714, |
| "learning_rate": 4.3009639150152e-05, |
| "loss": 0.4788, |
| "step": 172 |
| }, |
| { |
| "epoch": 2.638095238095238, |
| "grad_norm": 0.19560730560439193, |
| "learning_rate": 4.2580337977972664e-05, |
| "loss": 0.4827, |
| "step": 173 |
| }, |
| { |
| "epoch": 2.6533333333333333, |
| "grad_norm": 0.16741577940585486, |
| "learning_rate": 4.215073812558352e-05, |
| "loss": 0.4885, |
| "step": 174 |
| }, |
| { |
| "epoch": 2.6685714285714286, |
| "grad_norm": 0.1739859347419123, |
| "learning_rate": 4.1720889320181224e-05, |
| "loss": 0.4772, |
| "step": 175 |
| }, |
| { |
| "epoch": 2.683809523809524, |
| "grad_norm": 0.19324214245630159, |
| "learning_rate": 4.1290841317779325e-05, |
| "loss": 0.4884, |
| "step": 176 |
| }, |
| { |
| "epoch": 2.699047619047619, |
| "grad_norm": 0.162133413448313, |
| "learning_rate": 4.0860643897448894e-05, |
| "loss": 0.4789, |
| "step": 177 |
| }, |
| { |
| "epoch": 2.7142857142857144, |
| "grad_norm": 0.1910560430197172, |
| "learning_rate": 4.043034685555647e-05, |
| "loss": 0.4807, |
| "step": 178 |
| }, |
| { |
| "epoch": 2.7295238095238097, |
| "grad_norm": 0.14823981314032625, |
| "learning_rate": 4e-05, |
| "loss": 0.481, |
| "step": 179 |
| }, |
| { |
| "epoch": 2.744761904761905, |
| "grad_norm": 0.1784212235327859, |
| "learning_rate": 3.9569653144443546e-05, |
| "loss": 0.481, |
| "step": 180 |
| }, |
| { |
| "epoch": 2.76, |
| "grad_norm": 0.200733513756011, |
| "learning_rate": 3.913935610255112e-05, |
| "loss": 0.4763, |
| "step": 181 |
| }, |
| { |
| "epoch": 2.7752380952380955, |
| "grad_norm": 0.14938330915090647, |
| "learning_rate": 3.870915868222068e-05, |
| "loss": 0.4819, |
| "step": 182 |
| }, |
| { |
| "epoch": 2.7904761904761903, |
| "grad_norm": 0.1739261226284681, |
| "learning_rate": 3.8279110679818775e-05, |
| "loss": 0.4738, |
| "step": 183 |
| }, |
| { |
| "epoch": 2.8057142857142856, |
| "grad_norm": 0.19838890478019414, |
| "learning_rate": 3.78492618744165e-05, |
| "loss": 0.4783, |
| "step": 184 |
| }, |
| { |
| "epoch": 2.820952380952381, |
| "grad_norm": 0.1528772994448381, |
| "learning_rate": 3.7419662022027356e-05, |
| "loss": 0.4769, |
| "step": 185 |
| }, |
| { |
| "epoch": 2.836190476190476, |
| "grad_norm": 0.1919042097767604, |
| "learning_rate": 3.6990360849848005e-05, |
| "loss": 0.4757, |
| "step": 186 |
| }, |
| { |
| "epoch": 2.8514285714285714, |
| "grad_norm": 0.17566953990872838, |
| "learning_rate": 3.656140805050215e-05, |
| "loss": 0.4786, |
| "step": 187 |
| }, |
| { |
| "epoch": 2.8666666666666667, |
| "grad_norm": 0.1637970789000741, |
| "learning_rate": 3.6132853276288506e-05, |
| "loss": 0.4739, |
| "step": 188 |
| }, |
| { |
| "epoch": 2.881904761904762, |
| "grad_norm": 0.18549958217752188, |
| "learning_rate": 3.570474613343348e-05, |
| "loss": 0.4768, |
| "step": 189 |
| }, |
| { |
| "epoch": 2.8971428571428572, |
| "grad_norm": 0.18002674770058677, |
| "learning_rate": 3.527713617634904e-05, |
| "loss": 0.4733, |
| "step": 190 |
| }, |
| { |
| "epoch": 2.9123809523809525, |
| "grad_norm": 0.15898767756027013, |
| "learning_rate": 3.4850072901896766e-05, |
| "loss": 0.4802, |
| "step": 191 |
| }, |
| { |
| "epoch": 2.9276190476190473, |
| "grad_norm": 0.17974934358276728, |
| "learning_rate": 3.442360574365843e-05, |
| "loss": 0.4834, |
| "step": 192 |
| }, |
| { |
| "epoch": 2.942857142857143, |
| "grad_norm": 0.14203142815133346, |
| "learning_rate": 3.39977840662139e-05, |
| "loss": 0.4747, |
| "step": 193 |
| }, |
| { |
| "epoch": 2.958095238095238, |
| "grad_norm": 0.17952115377014885, |
| "learning_rate": 3.3572657159427186e-05, |
| "loss": 0.4808, |
| "step": 194 |
| }, |
| { |
| "epoch": 2.9733333333333336, |
| "grad_norm": 0.17353334331590287, |
| "learning_rate": 3.31482742327409e-05, |
| "loss": 0.4751, |
| "step": 195 |
| }, |
| { |
| "epoch": 2.9885714285714284, |
| "grad_norm": 0.1528037773950146, |
| "learning_rate": 3.272468440948025e-05, |
| "loss": 0.4734, |
| "step": 196 |
| }, |
| { |
| "epoch": 3.0047619047619047, |
| "grad_norm": 0.20480536463226645, |
| "learning_rate": 3.230193672116681e-05, |
| "loss": 0.6928, |
| "step": 197 |
| }, |
| { |
| "epoch": 3.02, |
| "grad_norm": 0.19151123516717333, |
| "learning_rate": 3.188008010184306e-05, |
| "loss": 0.4452, |
| "step": 198 |
| }, |
| { |
| "epoch": 3.0352380952380953, |
| "grad_norm": 0.2104984649163227, |
| "learning_rate": 3.145916338240816e-05, |
| "loss": 0.4546, |
| "step": 199 |
| }, |
| { |
| "epoch": 3.0504761904761906, |
| "grad_norm": 0.19741714449829997, |
| "learning_rate": 3.103923528496562e-05, |
| "loss": 0.4452, |
| "step": 200 |
| }, |
| { |
| "epoch": 3.065714285714286, |
| "grad_norm": 0.19623857285807467, |
| "learning_rate": 3.062034441718362e-05, |
| "loss": 0.4483, |
| "step": 201 |
| }, |
| { |
| "epoch": 3.080952380952381, |
| "grad_norm": 0.25035024771569697, |
| "learning_rate": 3.0202539266668568e-05, |
| "loss": 0.4506, |
| "step": 202 |
| }, |
| { |
| "epoch": 3.0961904761904764, |
| "grad_norm": 0.16968910324281286, |
| "learning_rate": 2.9785868195352525e-05, |
| "loss": 0.4449, |
| "step": 203 |
| }, |
| { |
| "epoch": 3.111428571428571, |
| "grad_norm": 0.23581488026041375, |
| "learning_rate": 2.9370379433895228e-05, |
| "loss": 0.4512, |
| "step": 204 |
| }, |
| { |
| "epoch": 3.1266666666666665, |
| "grad_norm": 0.15672745746884265, |
| "learning_rate": 2.8956121076101244e-05, |
| "loss": 0.449, |
| "step": 205 |
| }, |
| { |
| "epoch": 3.1419047619047618, |
| "grad_norm": 0.20702186752340365, |
| "learning_rate": 2.8543141073353026e-05, |
| "loss": 0.4439, |
| "step": 206 |
| }, |
| { |
| "epoch": 3.157142857142857, |
| "grad_norm": 0.16133838526688454, |
| "learning_rate": 2.8131487229060387e-05, |
| "loss": 0.4427, |
| "step": 207 |
| }, |
| { |
| "epoch": 3.1723809523809523, |
| "grad_norm": 0.1854062242780086, |
| "learning_rate": 2.7721207193127212e-05, |
| "loss": 0.4472, |
| "step": 208 |
| }, |
| { |
| "epoch": 3.1876190476190476, |
| "grad_norm": 0.17182738319155882, |
| "learning_rate": 2.7312348456435766e-05, |
| "loss": 0.4453, |
| "step": 209 |
| }, |
| { |
| "epoch": 3.202857142857143, |
| "grad_norm": 0.16697225397908244, |
| "learning_rate": 2.6904958345349633e-05, |
| "loss": 0.4561, |
| "step": 210 |
| }, |
| { |
| "epoch": 3.218095238095238, |
| "grad_norm": 0.16821522604659328, |
| "learning_rate": 2.6499084016235475e-05, |
| "loss": 0.444, |
| "step": 211 |
| }, |
| { |
| "epoch": 3.2333333333333334, |
| "grad_norm": 0.14853827941880265, |
| "learning_rate": 2.609477245000464e-05, |
| "loss": 0.44, |
| "step": 212 |
| }, |
| { |
| "epoch": 3.2485714285714287, |
| "grad_norm": 0.1502037122279702, |
| "learning_rate": 2.569207044667498e-05, |
| "loss": 0.4494, |
| "step": 213 |
| }, |
| { |
| "epoch": 3.263809523809524, |
| "grad_norm": 0.16216155417715333, |
| "learning_rate": 2.529102461995364e-05, |
| "loss": 0.4471, |
| "step": 214 |
| }, |
| { |
| "epoch": 3.279047619047619, |
| "grad_norm": 0.1586744564208733, |
| "learning_rate": 2.48916813918414e-05, |
| "loss": 0.4486, |
| "step": 215 |
| }, |
| { |
| "epoch": 3.2942857142857145, |
| "grad_norm": 0.14092622353201845, |
| "learning_rate": 2.449408698725928e-05, |
| "loss": 0.445, |
| "step": 216 |
| }, |
| { |
| "epoch": 3.3095238095238093, |
| "grad_norm": 0.15562276697944483, |
| "learning_rate": 2.409828742869777e-05, |
| "loss": 0.4415, |
| "step": 217 |
| }, |
| { |
| "epoch": 3.3247619047619046, |
| "grad_norm": 0.14631955274946826, |
| "learning_rate": 2.370432853088978e-05, |
| "loss": 0.454, |
| "step": 218 |
| }, |
| { |
| "epoch": 3.34, |
| "grad_norm": 0.14585582091880941, |
| "learning_rate": 2.3312255895507336e-05, |
| "loss": 0.4473, |
| "step": 219 |
| }, |
| { |
| "epoch": 3.355238095238095, |
| "grad_norm": 0.1458201114996886, |
| "learning_rate": 2.2922114905883167e-05, |
| "loss": 0.4444, |
| "step": 220 |
| }, |
| { |
| "epoch": 3.3704761904761904, |
| "grad_norm": 0.1402608135797642, |
| "learning_rate": 2.2533950721757447e-05, |
| "loss": 0.4465, |
| "step": 221 |
| }, |
| { |
| "epoch": 3.3857142857142857, |
| "grad_norm": 0.14672569191797688, |
| "learning_rate": 2.2147808274050436e-05, |
| "loss": 0.4484, |
| "step": 222 |
| }, |
| { |
| "epoch": 3.400952380952381, |
| "grad_norm": 0.15377591576947733, |
| "learning_rate": 2.1763732259661657e-05, |
| "loss": 0.4479, |
| "step": 223 |
| }, |
| { |
| "epoch": 3.416190476190476, |
| "grad_norm": 0.14739415052119614, |
| "learning_rate": 2.138176713629605e-05, |
| "loss": 0.4475, |
| "step": 224 |
| }, |
| { |
| "epoch": 3.4314285714285715, |
| "grad_norm": 0.14607883746358746, |
| "learning_rate": 2.1001957117318002e-05, |
| "loss": 0.4421, |
| "step": 225 |
| }, |
| { |
| "epoch": 3.4466666666666668, |
| "grad_norm": 0.1465432249850391, |
| "learning_rate": 2.0624346166633425e-05, |
| "loss": 0.4396, |
| "step": 226 |
| }, |
| { |
| "epoch": 3.461904761904762, |
| "grad_norm": 0.12349201931738121, |
| "learning_rate": 2.0248977993600912e-05, |
| "loss": 0.4491, |
| "step": 227 |
| }, |
| { |
| "epoch": 3.4771428571428573, |
| "grad_norm": 0.14981067132432477, |
| "learning_rate": 1.987589604797227e-05, |
| "loss": 0.4448, |
| "step": 228 |
| }, |
| { |
| "epoch": 3.4923809523809526, |
| "grad_norm": 0.1137656845152472, |
| "learning_rate": 1.9505143514863056e-05, |
| "loss": 0.4489, |
| "step": 229 |
| }, |
| { |
| "epoch": 3.5076190476190474, |
| "grad_norm": 0.1398849247762614, |
| "learning_rate": 1.913676330975385e-05, |
| "loss": 0.4546, |
| "step": 230 |
| }, |
| { |
| "epoch": 3.522857142857143, |
| "grad_norm": 0.12820525613752842, |
| "learning_rate": 1.8770798073522657e-05, |
| "loss": 0.447, |
| "step": 231 |
| }, |
| { |
| "epoch": 3.538095238095238, |
| "grad_norm": 0.13773852598915348, |
| "learning_rate": 1.8407290167509163e-05, |
| "loss": 0.4543, |
| "step": 232 |
| }, |
| { |
| "epoch": 3.5533333333333332, |
| "grad_norm": 0.12262425169332791, |
| "learning_rate": 1.8046281668611242e-05, |
| "loss": 0.444, |
| "step": 233 |
| }, |
| { |
| "epoch": 3.5685714285714285, |
| "grad_norm": 0.11207677036541486, |
| "learning_rate": 1.7687814364414508e-05, |
| "loss": 0.4451, |
| "step": 234 |
| }, |
| { |
| "epoch": 3.583809523809524, |
| "grad_norm": 0.11456531045490899, |
| "learning_rate": 1.73319297483553e-05, |
| "loss": 0.4465, |
| "step": 235 |
| }, |
| { |
| "epoch": 3.599047619047619, |
| "grad_norm": 0.12011483524657235, |
| "learning_rate": 1.6978669014917723e-05, |
| "loss": 0.4406, |
| "step": 236 |
| }, |
| { |
| "epoch": 3.6142857142857143, |
| "grad_norm": 0.11200693227331737, |
| "learning_rate": 1.6628073054865262e-05, |
| "loss": 0.4539, |
| "step": 237 |
| }, |
| { |
| "epoch": 3.6295238095238096, |
| "grad_norm": 0.11945653235133909, |
| "learning_rate": 1.6280182450507593e-05, |
| "loss": 0.4505, |
| "step": 238 |
| }, |
| { |
| "epoch": 3.644761904761905, |
| "grad_norm": 0.11328522908106976, |
| "learning_rate": 1.593503747100309e-05, |
| "loss": 0.44, |
| "step": 239 |
| }, |
| { |
| "epoch": 3.66, |
| "grad_norm": 0.11584353620347927, |
| "learning_rate": 1.5592678067697594e-05, |
| "loss": 0.4401, |
| "step": 240 |
| }, |
| { |
| "epoch": 3.675238095238095, |
| "grad_norm": 0.11338268232248719, |
| "learning_rate": 1.5253143869499875e-05, |
| "loss": 0.4526, |
| "step": 241 |
| }, |
| { |
| "epoch": 3.6904761904761907, |
| "grad_norm": 0.11825521687064344, |
| "learning_rate": 1.4916474178294578e-05, |
| "loss": 0.4472, |
| "step": 242 |
| }, |
| { |
| "epoch": 3.7057142857142855, |
| "grad_norm": 0.10393717835800384, |
| "learning_rate": 1.4582707964392909e-05, |
| "loss": 0.4503, |
| "step": 243 |
| }, |
| { |
| "epoch": 3.720952380952381, |
| "grad_norm": 0.11344323679399201, |
| "learning_rate": 1.4251883862021715e-05, |
| "loss": 0.4507, |
| "step": 244 |
| }, |
| { |
| "epoch": 3.736190476190476, |
| "grad_norm": 0.11795276330820621, |
| "learning_rate": 1.3924040164851497e-05, |
| "loss": 0.4509, |
| "step": 245 |
| }, |
| { |
| "epoch": 3.7514285714285713, |
| "grad_norm": 0.10446068928724489, |
| "learning_rate": 1.359921482156382e-05, |
| "loss": 0.453, |
| "step": 246 |
| }, |
| { |
| "epoch": 3.7666666666666666, |
| "grad_norm": 0.11963166659223025, |
| "learning_rate": 1.3277445431458653e-05, |
| "loss": 0.4423, |
| "step": 247 |
| }, |
| { |
| "epoch": 3.781904761904762, |
| "grad_norm": 0.11133235152901447, |
| "learning_rate": 1.2958769240102206e-05, |
| "loss": 0.4465, |
| "step": 248 |
| }, |
| { |
| "epoch": 3.797142857142857, |
| "grad_norm": 0.10488189725095455, |
| "learning_rate": 1.264322313501559e-05, |
| "loss": 0.4495, |
| "step": 249 |
| }, |
| { |
| "epoch": 3.8123809523809524, |
| "grad_norm": 0.09788048647298295, |
| "learning_rate": 1.2330843641405057e-05, |
| "loss": 0.4451, |
| "step": 250 |
| }, |
| { |
| "epoch": 3.8276190476190477, |
| "grad_norm": 0.1068932171231341, |
| "learning_rate": 1.2021666917934125e-05, |
| "loss": 0.4473, |
| "step": 251 |
| }, |
| { |
| "epoch": 3.842857142857143, |
| "grad_norm": 0.09965606097681808, |
| "learning_rate": 1.1715728752538103e-05, |
| "loss": 0.4443, |
| "step": 252 |
| }, |
| { |
| "epoch": 3.8580952380952382, |
| "grad_norm": 0.10053631876045145, |
| "learning_rate": 1.1413064558281564e-05, |
| "loss": 0.4468, |
| "step": 253 |
| }, |
| { |
| "epoch": 3.873333333333333, |
| "grad_norm": 0.09801079877642455, |
| "learning_rate": 1.1113709369259187e-05, |
| "loss": 0.4426, |
| "step": 254 |
| }, |
| { |
| "epoch": 3.888571428571429, |
| "grad_norm": 0.09483370655189176, |
| "learning_rate": 1.0817697836540493e-05, |
| "loss": 0.4493, |
| "step": 255 |
| }, |
| { |
| "epoch": 3.9038095238095236, |
| "grad_norm": 0.09358289511323944, |
| "learning_rate": 1.052506422415891e-05, |
| "loss": 0.4415, |
| "step": 256 |
| }, |
| { |
| "epoch": 3.919047619047619, |
| "grad_norm": 0.09937735008586661, |
| "learning_rate": 1.0235842405145587e-05, |
| "loss": 0.447, |
| "step": 257 |
| }, |
| { |
| "epoch": 3.934285714285714, |
| "grad_norm": 0.09713725112426527, |
| "learning_rate": 9.950065857608537e-06, |
| "loss": 0.4449, |
| "step": 258 |
| }, |
| { |
| "epoch": 3.9495238095238094, |
| "grad_norm": 0.10061824714900855, |
| "learning_rate": 9.667767660857508e-06, |
| "loss": 0.4378, |
| "step": 259 |
| }, |
| { |
| "epoch": 3.9647619047619047, |
| "grad_norm": 0.09975526240303212, |
| "learning_rate": 9.388980491574901e-06, |
| "loss": 0.4485, |
| "step": 260 |
| }, |
| { |
| "epoch": 3.98, |
| "grad_norm": 0.09583315993538538, |
| "learning_rate": 9.11373662003343e-06, |
| "loss": 0.4518, |
| "step": 261 |
| }, |
| { |
| "epoch": 3.9952380952380953, |
| "grad_norm": 0.12153372041343533, |
| "learning_rate": 8.842067906360716e-06, |
| "loss": 0.559, |
| "step": 262 |
| }, |
| { |
| "epoch": 4.011428571428572, |
| "grad_norm": 0.14086095004056465, |
| "learning_rate": 8.574005796851419e-06, |
| "loss": 0.5283, |
| "step": 263 |
| }, |
| { |
| "epoch": 4.026666666666666, |
| "grad_norm": 0.11788004209905123, |
| "learning_rate": 8.309581320327282e-06, |
| "loss": 0.4248, |
| "step": 264 |
| }, |
| { |
| "epoch": 4.041904761904762, |
| "grad_norm": 0.11207857563036874, |
| "learning_rate": 8.048825084545413e-06, |
| "loss": 0.4208, |
| "step": 265 |
| }, |
| { |
| "epoch": 4.057142857142857, |
| "grad_norm": 0.11130142780853505, |
| "learning_rate": 7.791767272655403e-06, |
| "loss": 0.4282, |
| "step": 266 |
| }, |
| { |
| "epoch": 4.072380952380953, |
| "grad_norm": 0.11636579346899405, |
| "learning_rate": 7.5384376397055644e-06, |
| "loss": 0.4275, |
| "step": 267 |
| }, |
| { |
| "epoch": 4.0876190476190475, |
| "grad_norm": 0.12911191155727517, |
| "learning_rate": 7.2888655091986995e-06, |
| "loss": 0.4296, |
| "step": 268 |
| }, |
| { |
| "epoch": 4.102857142857143, |
| "grad_norm": 0.1207757027022162, |
| "learning_rate": 7.043079769697833e-06, |
| "loss": 0.4301, |
| "step": 269 |
| }, |
| { |
| "epoch": 4.118095238095238, |
| "grad_norm": 0.10942499693335449, |
| "learning_rate": 6.801108871482287e-06, |
| "loss": 0.432, |
| "step": 270 |
| }, |
| { |
| "epoch": 4.133333333333334, |
| "grad_norm": 0.11446276362313004, |
| "learning_rate": 6.562980823254519e-06, |
| "loss": 0.4313, |
| "step": 271 |
| }, |
| { |
| "epoch": 4.148571428571429, |
| "grad_norm": 0.11355557887705738, |
| "learning_rate": 6.328723188898029e-06, |
| "loss": 0.4279, |
| "step": 272 |
| }, |
| { |
| "epoch": 4.163809523809523, |
| "grad_norm": 0.1062761145201427, |
| "learning_rate": 6.098363084286765e-06, |
| "loss": 0.4259, |
| "step": 273 |
| }, |
| { |
| "epoch": 4.179047619047619, |
| "grad_norm": 0.10273617911128365, |
| "learning_rate": 5.8719271741464056e-06, |
| "loss": 0.4225, |
| "step": 274 |
| }, |
| { |
| "epoch": 4.194285714285714, |
| "grad_norm": 0.10086062111767927, |
| "learning_rate": 5.6494416689678855e-06, |
| "loss": 0.4189, |
| "step": 275 |
| }, |
| { |
| "epoch": 4.20952380952381, |
| "grad_norm": 0.09559781846200713, |
| "learning_rate": 5.430932321973425e-06, |
| "loss": 0.4158, |
| "step": 276 |
| }, |
| { |
| "epoch": 4.2247619047619045, |
| "grad_norm": 0.09465874774217785, |
| "learning_rate": 5.21642442613556e-06, |
| "loss": 0.4299, |
| "step": 277 |
| }, |
| { |
| "epoch": 4.24, |
| "grad_norm": 0.09527880228230132, |
| "learning_rate": 5.005942811249412e-06, |
| "loss": 0.4243, |
| "step": 278 |
| }, |
| { |
| "epoch": 4.255238095238095, |
| "grad_norm": 0.09428002792877421, |
| "learning_rate": 4.799511841058575e-06, |
| "loss": 0.4262, |
| "step": 279 |
| }, |
| { |
| "epoch": 4.270476190476191, |
| "grad_norm": 0.09763553814742547, |
| "learning_rate": 4.597155410434972e-06, |
| "loss": 0.4247, |
| "step": 280 |
| }, |
| { |
| "epoch": 4.285714285714286, |
| "grad_norm": 0.09287137908820702, |
| "learning_rate": 4.398896942612934e-06, |
| "loss": 0.4238, |
| "step": 281 |
| }, |
| { |
| "epoch": 4.300952380952381, |
| "grad_norm": 0.09888369046987704, |
| "learning_rate": 4.204759386477961e-06, |
| "loss": 0.4391, |
| "step": 282 |
| }, |
| { |
| "epoch": 4.316190476190476, |
| "grad_norm": 0.09609824332962343, |
| "learning_rate": 4.01476521391027e-06, |
| "loss": 0.4352, |
| "step": 283 |
| }, |
| { |
| "epoch": 4.331428571428571, |
| "grad_norm": 0.09233308942677047, |
| "learning_rate": 3.828936417183671e-06, |
| "loss": 0.43, |
| "step": 284 |
| }, |
| { |
| "epoch": 4.346666666666667, |
| "grad_norm": 0.0890288964909307, |
| "learning_rate": 3.6472945064198827e-06, |
| "loss": 0.427, |
| "step": 285 |
| }, |
| { |
| "epoch": 4.3619047619047615, |
| "grad_norm": 0.08989519773450554, |
| "learning_rate": 3.4698605070987077e-06, |
| "loss": 0.4257, |
| "step": 286 |
| }, |
| { |
| "epoch": 4.377142857142857, |
| "grad_norm": 0.09526649707374434, |
| "learning_rate": 3.2966549576242477e-06, |
| "loss": 0.4357, |
| "step": 287 |
| }, |
| { |
| "epoch": 4.392380952380952, |
| "grad_norm": 0.09131870660723178, |
| "learning_rate": 3.127697906947584e-06, |
| "loss": 0.4272, |
| "step": 288 |
| }, |
| { |
| "epoch": 4.407619047619048, |
| "grad_norm": 0.08945756240639792, |
| "learning_rate": 2.963008912246008e-06, |
| "loss": 0.4347, |
| "step": 289 |
| }, |
| { |
| "epoch": 4.422857142857143, |
| "grad_norm": 0.08483664610068463, |
| "learning_rate": 2.802607036659284e-06, |
| "loss": 0.4294, |
| "step": 290 |
| }, |
| { |
| "epoch": 4.438095238095238, |
| "grad_norm": 0.08486549550090346, |
| "learning_rate": 2.646510847082997e-06, |
| "loss": 0.4303, |
| "step": 291 |
| }, |
| { |
| "epoch": 4.453333333333333, |
| "grad_norm": 0.084355078151828, |
| "learning_rate": 2.494738412019442e-06, |
| "loss": 0.423, |
| "step": 292 |
| }, |
| { |
| "epoch": 4.468571428571429, |
| "grad_norm": 0.08518364314817348, |
| "learning_rate": 2.3473072994861167e-06, |
| "loss": 0.427, |
| "step": 293 |
| }, |
| { |
| "epoch": 4.483809523809524, |
| "grad_norm": 0.08287931866258626, |
| "learning_rate": 2.2042345749821913e-06, |
| "loss": 0.4216, |
| "step": 294 |
| }, |
| { |
| "epoch": 4.499047619047619, |
| "grad_norm": 0.08289838691379334, |
| "learning_rate": 2.065536799513135e-06, |
| "loss": 0.4288, |
| "step": 295 |
| }, |
| { |
| "epoch": 4.514285714285714, |
| "grad_norm": 0.0850326818939402, |
| "learning_rate": 1.9312300276737516e-06, |
| "loss": 0.4235, |
| "step": 296 |
| }, |
| { |
| "epoch": 4.529523809523809, |
| "grad_norm": 0.08392478938826013, |
| "learning_rate": 1.8013298057897932e-06, |
| "loss": 0.4293, |
| "step": 297 |
| }, |
| { |
| "epoch": 4.544761904761905, |
| "grad_norm": 0.0892781985318678, |
| "learning_rate": 1.6758511701184809e-06, |
| "loss": 0.4238, |
| "step": 298 |
| }, |
| { |
| "epoch": 4.5600000000000005, |
| "grad_norm": 0.08180406525124756, |
| "learning_rate": 1.5548086451079747e-06, |
| "loss": 0.4384, |
| "step": 299 |
| }, |
| { |
| "epoch": 4.575238095238095, |
| "grad_norm": 0.08011550631601072, |
| "learning_rate": 1.4382162417161748e-06, |
| "loss": 0.4206, |
| "step": 300 |
| }, |
| { |
| "epoch": 4.59047619047619, |
| "grad_norm": 0.08066581660175723, |
| "learning_rate": 1.3260874557889002e-06, |
| "loss": 0.4287, |
| "step": 301 |
| }, |
| { |
| "epoch": 4.605714285714286, |
| "grad_norm": 0.0806215224586864, |
| "learning_rate": 1.2184352664977105e-06, |
| "loss": 0.4175, |
| "step": 302 |
| }, |
| { |
| "epoch": 4.620952380952381, |
| "grad_norm": 0.07969848747023248, |
| "learning_rate": 1.115272134837544e-06, |
| "loss": 0.4312, |
| "step": 303 |
| }, |
| { |
| "epoch": 4.636190476190476, |
| "grad_norm": 0.0797095157185979, |
| "learning_rate": 1.0166100021843283e-06, |
| "loss": 0.4346, |
| "step": 304 |
| }, |
| { |
| "epoch": 4.651428571428571, |
| "grad_norm": 0.0792961159548818, |
| "learning_rate": 9.224602889127321e-07, |
| "loss": 0.4286, |
| "step": 305 |
| }, |
| { |
| "epoch": 4.666666666666667, |
| "grad_norm": 0.08285323866652071, |
| "learning_rate": 8.328338930742296e-07, |
| "loss": 0.4267, |
| "step": 306 |
| }, |
| { |
| "epoch": 4.681904761904762, |
| "grad_norm": 0.07823875843722773, |
| "learning_rate": 7.477411891356268e-07, |
| "loss": 0.4276, |
| "step": 307 |
| }, |
| { |
| "epoch": 4.6971428571428575, |
| "grad_norm": 0.07763022247982207, |
| "learning_rate": 6.671920267782029e-07, |
| "loss": 0.4282, |
| "step": 308 |
| }, |
| { |
| "epoch": 4.712380952380952, |
| "grad_norm": 0.07871343041787499, |
| "learning_rate": 5.911957297575743e-07, |
| "loss": 0.4323, |
| "step": 309 |
| }, |
| { |
| "epoch": 4.727619047619047, |
| "grad_norm": 0.07681136995775628, |
| "learning_rate": 5.197610948244469e-07, |
| "loss": 0.4256, |
| "step": 310 |
| }, |
| { |
| "epoch": 4.742857142857143, |
| "grad_norm": 0.07665250706053243, |
| "learning_rate": 4.5289639070638103e-07, |
| "loss": 0.4171, |
| "step": 311 |
| }, |
| { |
| "epoch": 4.758095238095238, |
| "grad_norm": 0.07835687105047397, |
| "learning_rate": 3.906093571506597e-07, |
| "loss": 0.428, |
| "step": 312 |
| }, |
| { |
| "epoch": 4.773333333333333, |
| "grad_norm": 0.07949715755664716, |
| "learning_rate": 3.329072040284009e-07, |
| "loss": 0.4208, |
| "step": 313 |
| }, |
| { |
| "epoch": 4.788571428571428, |
| "grad_norm": 0.0773084553588578, |
| "learning_rate": 2.797966104999805e-07, |
| "loss": 0.4332, |
| "step": 314 |
| }, |
| { |
| "epoch": 4.803809523809524, |
| "grad_norm": 0.07795013678746698, |
| "learning_rate": 2.3128372424192635e-07, |
| "loss": 0.4377, |
| "step": 315 |
| }, |
| { |
| "epoch": 4.819047619047619, |
| "grad_norm": 0.07666311315306534, |
| "learning_rate": 1.8737416073529635e-07, |
| "loss": 0.4272, |
| "step": 316 |
| }, |
| { |
| "epoch": 4.8342857142857145, |
| "grad_norm": 0.07693103942282913, |
| "learning_rate": 1.4807300261568735e-07, |
| "loss": 0.4287, |
| "step": 317 |
| }, |
| { |
| "epoch": 4.849523809523809, |
| "grad_norm": 0.08092920067828374, |
| "learning_rate": 1.1338479908488798e-07, |
| "loss": 0.4337, |
| "step": 318 |
| }, |
| { |
| "epoch": 4.864761904761905, |
| "grad_norm": 0.07921080037622698, |
| "learning_rate": 8.331356538431313e-08, |
| "loss": 0.4262, |
| "step": 319 |
| }, |
| { |
| "epoch": 4.88, |
| "grad_norm": 0.08216408996820164, |
| "learning_rate": 5.7862782330215804e-08, |
| "loss": 0.4317, |
| "step": 320 |
| }, |
| { |
| "epoch": 4.895238095238096, |
| "grad_norm": 0.07591222440722502, |
| "learning_rate": 3.703539591080052e-08, |
| "loss": 0.4232, |
| "step": 321 |
| }, |
| { |
| "epoch": 4.91047619047619, |
| "grad_norm": 0.08045135907636206, |
| "learning_rate": 2.0833816945184936e-08, |
| "loss": 0.4297, |
| "step": 322 |
| }, |
| { |
| "epoch": 4.925714285714285, |
| "grad_norm": 0.07684321088822434, |
| "learning_rate": 9.25992080436533e-09, |
| "loss": 0.4276, |
| "step": 323 |
| }, |
| { |
| "epoch": 4.940952380952381, |
| "grad_norm": 0.07680665420588856, |
| "learning_rate": 2.315047194136888e-09, |
| "loss": 0.4271, |
| "step": 324 |
| }, |
| { |
| "epoch": 4.956190476190476, |
| "grad_norm": 0.07663586105720761, |
| "learning_rate": 0.0, |
| "loss": 0.4248, |
| "step": 325 |
| }, |
| { |
| "epoch": 4.956190476190476, |
| "step": 325, |
| "total_flos": 8.641809758367515e+18, |
| "train_loss": 0.5190348960803105, |
| "train_runtime": 75842.3225, |
| "train_samples_per_second": 2.214, |
| "train_steps_per_second": 0.004 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 325, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 5, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 8.641809758367515e+18, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
