| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 2.9952, |
| "eval_steps": 500, |
| "global_step": 312, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0096, |
| "grad_norm": 6.273941516876221, |
| "learning_rate": 3.125e-07, |
| "loss": 1.0433, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.0192, |
| "grad_norm": 5.883962154388428, |
| "learning_rate": 6.25e-07, |
| "loss": 1.0078, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.0288, |
| "grad_norm": 6.3777031898498535, |
| "learning_rate": 9.375000000000001e-07, |
| "loss": 1.0117, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.0384, |
| "grad_norm": 6.135955810546875, |
| "learning_rate": 1.25e-06, |
| "loss": 0.9995, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.048, |
| "grad_norm": 6.0287981033325195, |
| "learning_rate": 1.5625e-06, |
| "loss": 1.016, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.0576, |
| "grad_norm": 5.598588466644287, |
| "learning_rate": 1.8750000000000003e-06, |
| "loss": 1.0019, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.0672, |
| "grad_norm": 4.601566314697266, |
| "learning_rate": 2.1875000000000002e-06, |
| "loss": 0.9724, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.0768, |
| "grad_norm": 4.467075824737549, |
| "learning_rate": 2.5e-06, |
| "loss": 0.9682, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.0864, |
| "grad_norm": 2.5521304607391357, |
| "learning_rate": 2.8125e-06, |
| "loss": 0.9056, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.096, |
| "grad_norm": 2.4211952686309814, |
| "learning_rate": 3.125e-06, |
| "loss": 0.8879, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.1056, |
| "grad_norm": 2.1752817630767822, |
| "learning_rate": 3.4375e-06, |
| "loss": 0.8818, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.1152, |
| "grad_norm": 3.6486101150512695, |
| "learning_rate": 3.7500000000000005e-06, |
| "loss": 0.902, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.1248, |
| "grad_norm": 4.0764851570129395, |
| "learning_rate": 4.0625000000000005e-06, |
| "loss": 0.9259, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.1344, |
| "grad_norm": 3.5974338054656982, |
| "learning_rate": 4.3750000000000005e-06, |
| "loss": 0.8718, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.144, |
| "grad_norm": 3.368156909942627, |
| "learning_rate": 4.6875000000000004e-06, |
| "loss": 0.855, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.1536, |
| "grad_norm": 2.510629653930664, |
| "learning_rate": 5e-06, |
| "loss": 0.8303, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.1632, |
| "grad_norm": 2.1214306354522705, |
| "learning_rate": 5.3125e-06, |
| "loss": 0.7902, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.1728, |
| "grad_norm": 1.9427293539047241, |
| "learning_rate": 5.625e-06, |
| "loss": 0.8608, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.1824, |
| "grad_norm": 1.5806338787078857, |
| "learning_rate": 5.9375e-06, |
| "loss": 0.7988, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.192, |
| "grad_norm": 1.4685497283935547, |
| "learning_rate": 6.25e-06, |
| "loss": 0.7877, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.2016, |
| "grad_norm": 1.1793313026428223, |
| "learning_rate": 6.5625e-06, |
| "loss": 0.7172, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.2112, |
| "grad_norm": 1.2347708940505981, |
| "learning_rate": 6.875e-06, |
| "loss": 0.7375, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.2208, |
| "grad_norm": 1.3614697456359863, |
| "learning_rate": 7.1875e-06, |
| "loss": 0.7647, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.2304, |
| "grad_norm": 1.2697902917861938, |
| "learning_rate": 7.500000000000001e-06, |
| "loss": 0.7701, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.24, |
| "grad_norm": 1.0271424055099487, |
| "learning_rate": 7.8125e-06, |
| "loss": 0.7413, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.2496, |
| "grad_norm": 1.1519695520401, |
| "learning_rate": 8.125000000000001e-06, |
| "loss": 0.7155, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.2592, |
| "grad_norm": 1.3047678470611572, |
| "learning_rate": 8.4375e-06, |
| "loss": 0.7273, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.2688, |
| "grad_norm": 1.0531271696090698, |
| "learning_rate": 8.750000000000001e-06, |
| "loss": 0.7263, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.2784, |
| "grad_norm": 0.7886083722114563, |
| "learning_rate": 9.0625e-06, |
| "loss": 0.7221, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.288, |
| "grad_norm": 1.0374627113342285, |
| "learning_rate": 9.375000000000001e-06, |
| "loss": 0.7395, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.2976, |
| "grad_norm": 0.9843682050704956, |
| "learning_rate": 9.6875e-06, |
| "loss": 0.7197, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.3072, |
| "grad_norm": 0.8767869472503662, |
| "learning_rate": 1e-05, |
| "loss": 0.714, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.3168, |
| "grad_norm": 0.9439571499824524, |
| "learning_rate": 9.999685283773504e-06, |
| "loss": 0.7318, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.3264, |
| "grad_norm": 0.9337266087532043, |
| "learning_rate": 9.998741174712534e-06, |
| "loss": 0.6993, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.336, |
| "grad_norm": 0.8179473280906677, |
| "learning_rate": 9.997167791667668e-06, |
| "loss": 0.6659, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.3456, |
| "grad_norm": 0.7900959849357605, |
| "learning_rate": 9.994965332706574e-06, |
| "loss": 0.6875, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.3552, |
| "grad_norm": 0.8492732048034668, |
| "learning_rate": 9.992134075089085e-06, |
| "loss": 0.698, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.3648, |
| "grad_norm": 0.7290406823158264, |
| "learning_rate": 9.98867437523228e-06, |
| "loss": 0.6955, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.3744, |
| "grad_norm": 0.7271062135696411, |
| "learning_rate": 9.984586668665641e-06, |
| "loss": 0.6812, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.384, |
| "grad_norm": 0.8993069529533386, |
| "learning_rate": 9.979871469976197e-06, |
| "loss": 0.7021, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.3936, |
| "grad_norm": 0.8171764016151428, |
| "learning_rate": 9.974529372743762e-06, |
| "loss": 0.7173, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.4032, |
| "grad_norm": 0.7342078685760498, |
| "learning_rate": 9.968561049466214e-06, |
| "loss": 0.6812, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.4128, |
| "grad_norm": 0.7545952796936035, |
| "learning_rate": 9.961967251474823e-06, |
| "loss": 0.6869, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.4224, |
| "grad_norm": 0.8997238278388977, |
| "learning_rate": 9.954748808839675e-06, |
| "loss": 0.6997, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.432, |
| "grad_norm": 0.8454792499542236, |
| "learning_rate": 9.946906630265184e-06, |
| "loss": 0.6714, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.4416, |
| "grad_norm": 1.486637830734253, |
| "learning_rate": 9.938441702975689e-06, |
| "loss": 0.6751, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.4512, |
| "grad_norm": 0.8794205784797668, |
| "learning_rate": 9.92935509259118e-06, |
| "loss": 0.6635, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.4608, |
| "grad_norm": 0.7148414850234985, |
| "learning_rate": 9.91964794299315e-06, |
| "loss": 0.6805, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.4704, |
| "grad_norm": 0.7541925311088562, |
| "learning_rate": 9.909321476180594e-06, |
| "loss": 0.6903, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.48, |
| "grad_norm": 0.721092700958252, |
| "learning_rate": 9.898376992116179e-06, |
| "loss": 0.6909, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.4896, |
| "grad_norm": 0.6446923613548279, |
| "learning_rate": 9.886815868562596e-06, |
| "loss": 0.6737, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.4992, |
| "grad_norm": 0.7050280570983887, |
| "learning_rate": 9.874639560909118e-06, |
| "loss": 0.697, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.5088, |
| "grad_norm": 0.723859429359436, |
| "learning_rate": 9.861849601988384e-06, |
| "loss": 0.6446, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.5184, |
| "grad_norm": 0.6872769594192505, |
| "learning_rate": 9.848447601883436e-06, |
| "loss": 0.6863, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.528, |
| "grad_norm": 0.6015923023223877, |
| "learning_rate": 9.834435247725032e-06, |
| "loss": 0.6665, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.5376, |
| "grad_norm": 0.7028643488883972, |
| "learning_rate": 9.819814303479268e-06, |
| "loss": 0.6796, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.5472, |
| "grad_norm": 0.732957124710083, |
| "learning_rate": 9.804586609725499e-06, |
| "loss": 0.6661, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.5568, |
| "grad_norm": 0.5806607007980347, |
| "learning_rate": 9.788754083424654e-06, |
| "loss": 0.6395, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.5664, |
| "grad_norm": 0.6523646116256714, |
| "learning_rate": 9.772318717677905e-06, |
| "loss": 0.6519, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.576, |
| "grad_norm": 0.7675644159317017, |
| "learning_rate": 9.755282581475769e-06, |
| "loss": 0.6528, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.5856, |
| "grad_norm": 0.5650181770324707, |
| "learning_rate": 9.737647819437645e-06, |
| "loss": 0.6515, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.5952, |
| "grad_norm": 0.5440039038658142, |
| "learning_rate": 9.719416651541839e-06, |
| "loss": 0.6438, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.6048, |
| "grad_norm": 0.6638810634613037, |
| "learning_rate": 9.700591372846096e-06, |
| "loss": 0.6287, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.6144, |
| "grad_norm": 0.6809440851211548, |
| "learning_rate": 9.681174353198687e-06, |
| "loss": 0.6463, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.624, |
| "grad_norm": 0.5411990284919739, |
| "learning_rate": 9.661168036940071e-06, |
| "loss": 0.6379, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.6336, |
| "grad_norm": 0.6200382709503174, |
| "learning_rate": 9.640574942595195e-06, |
| "loss": 0.6714, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.6432, |
| "grad_norm": 0.6965516805648804, |
| "learning_rate": 9.619397662556434e-06, |
| "loss": 0.6324, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.6528, |
| "grad_norm": 0.6105510592460632, |
| "learning_rate": 9.597638862757255e-06, |
| "loss": 0.6635, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.6624, |
| "grad_norm": 0.6869755387306213, |
| "learning_rate": 9.5753012823366e-06, |
| "loss": 0.6325, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.672, |
| "grad_norm": 0.6416379809379578, |
| "learning_rate": 9.552387733294081e-06, |
| "loss": 0.6309, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.6816, |
| "grad_norm": 0.6986199617385864, |
| "learning_rate": 9.528901100135971e-06, |
| "loss": 0.652, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.6912, |
| "grad_norm": 0.6804450750350952, |
| "learning_rate": 9.504844339512096e-06, |
| "loss": 0.6527, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.7008, |
| "grad_norm": 0.6506721377372742, |
| "learning_rate": 9.480220479843627e-06, |
| "loss": 0.638, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.7104, |
| "grad_norm": 0.6433568596839905, |
| "learning_rate": 9.45503262094184e-06, |
| "loss": 0.6268, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.72, |
| "grad_norm": 0.6668800115585327, |
| "learning_rate": 9.4292839336179e-06, |
| "loss": 0.6343, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.7296, |
| "grad_norm": 0.6119109392166138, |
| "learning_rate": 9.40297765928369e-06, |
| "loss": 0.6333, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.7392, |
| "grad_norm": 0.5585500001907349, |
| "learning_rate": 9.376117109543769e-06, |
| "loss": 0.6533, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.7488, |
| "grad_norm": 0.5532995462417603, |
| "learning_rate": 9.348705665778479e-06, |
| "loss": 0.6317, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.7584, |
| "grad_norm": 0.6779194474220276, |
| "learning_rate": 9.320746778718274e-06, |
| "loss": 0.6414, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.768, |
| "grad_norm": 0.5922401547431946, |
| "learning_rate": 9.292243968009332e-06, |
| "loss": 0.6368, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.7776, |
| "grad_norm": 0.5234724879264832, |
| "learning_rate": 9.263200821770462e-06, |
| "loss": 0.6325, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.7872, |
| "grad_norm": 0.609754741191864, |
| "learning_rate": 9.233620996141421e-06, |
| "loss": 0.6311, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.7968, |
| "grad_norm": 0.5881137847900391, |
| "learning_rate": 9.203508214822652e-06, |
| "loss": 0.6509, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.8064, |
| "grad_norm": 0.535968542098999, |
| "learning_rate": 9.172866268606514e-06, |
| "loss": 0.6374, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.816, |
| "grad_norm": 0.6534250378608704, |
| "learning_rate": 9.141699014900084e-06, |
| "loss": 0.6461, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.8256, |
| "grad_norm": 0.5637496113777161, |
| "learning_rate": 9.110010377239552e-06, |
| "loss": 0.6293, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.8352, |
| "grad_norm": 0.6735429167747498, |
| "learning_rate": 9.077804344796302e-06, |
| "loss": 0.6732, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.8448, |
| "grad_norm": 0.6145018339157104, |
| "learning_rate": 9.045084971874738e-06, |
| "loss": 0.6459, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.8544, |
| "grad_norm": 0.5470115542411804, |
| "learning_rate": 9.011856377401891e-06, |
| "loss": 0.6385, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.864, |
| "grad_norm": 0.5681822299957275, |
| "learning_rate": 8.978122744408905e-06, |
| "loss": 0.6416, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.8736, |
| "grad_norm": 0.610372006893158, |
| "learning_rate": 8.943888319504456e-06, |
| "loss": 0.6575, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.8832, |
| "grad_norm": 0.5592759251594543, |
| "learning_rate": 8.90915741234015e-06, |
| "loss": 0.6521, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.8928, |
| "grad_norm": 0.5998126864433289, |
| "learning_rate": 8.873934395068006e-06, |
| "loss": 0.6426, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.9024, |
| "grad_norm": 0.5752755999565125, |
| "learning_rate": 8.838223701790057e-06, |
| "loss": 0.6347, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.912, |
| "grad_norm": 0.6242445707321167, |
| "learning_rate": 8.802029828000157e-06, |
| "loss": 0.6441, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.9216, |
| "grad_norm": 0.5645915865898132, |
| "learning_rate": 8.765357330018056e-06, |
| "loss": 0.6182, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.9312, |
| "grad_norm": 0.6026539206504822, |
| "learning_rate": 8.728210824415829e-06, |
| "loss": 0.669, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.9408, |
| "grad_norm": 0.6019846200942993, |
| "learning_rate": 8.690594987436705e-06, |
| "loss": 0.6275, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.9504, |
| "grad_norm": 0.5070683360099792, |
| "learning_rate": 8.652514554406388e-06, |
| "loss": 0.6107, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.96, |
| "grad_norm": 0.5749801993370056, |
| "learning_rate": 8.613974319136959e-06, |
| "loss": 0.6525, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.9696, |
| "grad_norm": 0.5775330066680908, |
| "learning_rate": 8.574979133323378e-06, |
| "loss": 0.629, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.9792, |
| "grad_norm": 0.6243678331375122, |
| "learning_rate": 8.535533905932739e-06, |
| "loss": 0.6408, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.9888, |
| "grad_norm": 0.6030294895172119, |
| "learning_rate": 8.495643602586287e-06, |
| "loss": 0.6078, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.9984, |
| "grad_norm": 0.5548235177993774, |
| "learning_rate": 8.455313244934324e-06, |
| "loss": 0.657, |
| "step": 104 |
| }, |
| { |
| "epoch": 1.008, |
| "grad_norm": 1.418408751487732, |
| "learning_rate": 8.414547910024035e-06, |
| "loss": 1.1845, |
| "step": 105 |
| }, |
| { |
| "epoch": 1.0176, |
| "grad_norm": 0.5272033214569092, |
| "learning_rate": 8.373352729660373e-06, |
| "loss": 0.5266, |
| "step": 106 |
| }, |
| { |
| "epoch": 1.0272, |
| "grad_norm": 0.6736261248588562, |
| "learning_rate": 8.331732889760021e-06, |
| "loss": 0.6198, |
| "step": 107 |
| }, |
| { |
| "epoch": 1.0368, |
| "grad_norm": 0.650242269039154, |
| "learning_rate": 8.289693629698564e-06, |
| "loss": 0.5922, |
| "step": 108 |
| }, |
| { |
| "epoch": 1.0464, |
| "grad_norm": 0.5418078303337097, |
| "learning_rate": 8.247240241650918e-06, |
| "loss": 0.6109, |
| "step": 109 |
| }, |
| { |
| "epoch": 1.056, |
| "grad_norm": 0.6298655867576599, |
| "learning_rate": 8.204378069925121e-06, |
| "loss": 0.5743, |
| "step": 110 |
| }, |
| { |
| "epoch": 1.0656, |
| "grad_norm": 0.6450831294059753, |
| "learning_rate": 8.16111251028955e-06, |
| "loss": 0.5738, |
| "step": 111 |
| }, |
| { |
| "epoch": 1.0752, |
| "grad_norm": 0.5031427145004272, |
| "learning_rate": 8.117449009293668e-06, |
| "loss": 0.5397, |
| "step": 112 |
| }, |
| { |
| "epoch": 1.0848, |
| "grad_norm": 0.7727047801017761, |
| "learning_rate": 8.073393063582386e-06, |
| "loss": 0.6133, |
| "step": 113 |
| }, |
| { |
| "epoch": 1.0944, |
| "grad_norm": 0.5162041783332825, |
| "learning_rate": 8.0289502192041e-06, |
| "loss": 0.5389, |
| "step": 114 |
| }, |
| { |
| "epoch": 1.104, |
| "grad_norm": 0.6630980372428894, |
| "learning_rate": 7.984126070912519e-06, |
| "loss": 0.5997, |
| "step": 115 |
| }, |
| { |
| "epoch": 1.1136, |
| "grad_norm": 0.5223315358161926, |
| "learning_rate": 7.938926261462366e-06, |
| "loss": 0.5643, |
| "step": 116 |
| }, |
| { |
| "epoch": 1.1232, |
| "grad_norm": 0.6221403479576111, |
| "learning_rate": 7.89335648089903e-06, |
| "loss": 0.5627, |
| "step": 117 |
| }, |
| { |
| "epoch": 1.1328, |
| "grad_norm": 0.6523832678794861, |
| "learning_rate": 7.84742246584226e-06, |
| "loss": 0.6218, |
| "step": 118 |
| }, |
| { |
| "epoch": 1.1424, |
| "grad_norm": 0.5404381155967712, |
| "learning_rate": 7.801129998764014e-06, |
| "loss": 0.5941, |
| "step": 119 |
| }, |
| { |
| "epoch": 1.152, |
| "grad_norm": 0.6009036898612976, |
| "learning_rate": 7.754484907260513e-06, |
| "loss": 0.5785, |
| "step": 120 |
| }, |
| { |
| "epoch": 1.1616, |
| "grad_norm": 0.5859388113021851, |
| "learning_rate": 7.70749306331863e-06, |
| "loss": 0.5607, |
| "step": 121 |
| }, |
| { |
| "epoch": 1.1712, |
| "grad_norm": 0.5393742322921753, |
| "learning_rate": 7.660160382576683e-06, |
| "loss": 0.6196, |
| "step": 122 |
| }, |
| { |
| "epoch": 1.1808, |
| "grad_norm": 0.5670076012611389, |
| "learning_rate": 7.612492823579744e-06, |
| "loss": 0.5725, |
| "step": 123 |
| }, |
| { |
| "epoch": 1.1904, |
| "grad_norm": 0.5758752822875977, |
| "learning_rate": 7.564496387029532e-06, |
| "loss": 0.6359, |
| "step": 124 |
| }, |
| { |
| "epoch": 1.2, |
| "grad_norm": 0.47169139981269836, |
| "learning_rate": 7.516177115029002e-06, |
| "loss": 0.5587, |
| "step": 125 |
| }, |
| { |
| "epoch": 1.2096, |
| "grad_norm": 0.6857102513313293, |
| "learning_rate": 7.467541090321735e-06, |
| "loss": 0.6034, |
| "step": 126 |
| }, |
| { |
| "epoch": 1.2192, |
| "grad_norm": 0.5409797430038452, |
| "learning_rate": 7.4185944355261996e-06, |
| "loss": 0.5597, |
| "step": 127 |
| }, |
| { |
| "epoch": 1.2288000000000001, |
| "grad_norm": 0.5294946432113647, |
| "learning_rate": 7.369343312364994e-06, |
| "loss": 0.5969, |
| "step": 128 |
| }, |
| { |
| "epoch": 1.2384, |
| "grad_norm": 0.614827036857605, |
| "learning_rate": 7.319793920889171e-06, |
| "loss": 0.579, |
| "step": 129 |
| }, |
| { |
| "epoch": 1.248, |
| "grad_norm": 0.6321758031845093, |
| "learning_rate": 7.269952498697734e-06, |
| "loss": 0.6055, |
| "step": 130 |
| }, |
| { |
| "epoch": 1.2576, |
| "grad_norm": 0.6027973890304565, |
| "learning_rate": 7.219825320152411e-06, |
| "loss": 0.6426, |
| "step": 131 |
| }, |
| { |
| "epoch": 1.2671999999999999, |
| "grad_norm": 0.5362632870674133, |
| "learning_rate": 7.169418695587791e-06, |
| "loss": 0.4882, |
| "step": 132 |
| }, |
| { |
| "epoch": 1.2768, |
| "grad_norm": 0.6094551086425781, |
| "learning_rate": 7.118738970516944e-06, |
| "loss": 0.6389, |
| "step": 133 |
| }, |
| { |
| "epoch": 1.2864, |
| "grad_norm": 0.5218535661697388, |
| "learning_rate": 7.067792524832604e-06, |
| "loss": 0.5509, |
| "step": 134 |
| }, |
| { |
| "epoch": 1.296, |
| "grad_norm": 0.5705574154853821, |
| "learning_rate": 7.016585772004026e-06, |
| "loss": 0.5877, |
| "step": 135 |
| }, |
| { |
| "epoch": 1.3056, |
| "grad_norm": 0.6876828074455261, |
| "learning_rate": 6.965125158269619e-06, |
| "loss": 0.5648, |
| "step": 136 |
| }, |
| { |
| "epoch": 1.3152, |
| "grad_norm": 0.5774961709976196, |
| "learning_rate": 6.913417161825449e-06, |
| "loss": 0.5945, |
| "step": 137 |
| }, |
| { |
| "epoch": 1.3248, |
| "grad_norm": 0.5037078857421875, |
| "learning_rate": 6.8614682920097265e-06, |
| "loss": 0.5164, |
| "step": 138 |
| }, |
| { |
| "epoch": 1.3344, |
| "grad_norm": 0.7171159982681274, |
| "learning_rate": 6.809285088483361e-06, |
| "loss": 0.6381, |
| "step": 139 |
| }, |
| { |
| "epoch": 1.3439999999999999, |
| "grad_norm": 0.5094574689865112, |
| "learning_rate": 6.7568741204067145e-06, |
| "loss": 0.5812, |
| "step": 140 |
| }, |
| { |
| "epoch": 1.3536000000000001, |
| "grad_norm": 0.5302834510803223, |
| "learning_rate": 6.704241985612625e-06, |
| "loss": 0.5652, |
| "step": 141 |
| }, |
| { |
| "epoch": 1.3632, |
| "grad_norm": 0.5682352781295776, |
| "learning_rate": 6.651395309775837e-06, |
| "loss": 0.5677, |
| "step": 142 |
| }, |
| { |
| "epoch": 1.3728, |
| "grad_norm": 0.5153688192367554, |
| "learning_rate": 6.598340745578908e-06, |
| "loss": 0.5978, |
| "step": 143 |
| }, |
| { |
| "epoch": 1.3824, |
| "grad_norm": 0.48127686977386475, |
| "learning_rate": 6.545084971874738e-06, |
| "loss": 0.5694, |
| "step": 144 |
| }, |
| { |
| "epoch": 1.392, |
| "grad_norm": 0.5677974224090576, |
| "learning_rate": 6.491634692845781e-06, |
| "loss": 0.5847, |
| "step": 145 |
| }, |
| { |
| "epoch": 1.4016, |
| "grad_norm": 0.6412763595581055, |
| "learning_rate": 6.437996637160086e-06, |
| "loss": 0.5749, |
| "step": 146 |
| }, |
| { |
| "epoch": 1.4112, |
| "grad_norm": 0.5909486413002014, |
| "learning_rate": 6.384177557124247e-06, |
| "loss": 0.6475, |
| "step": 147 |
| }, |
| { |
| "epoch": 1.4208, |
| "grad_norm": 0.5015473961830139, |
| "learning_rate": 6.330184227833376e-06, |
| "loss": 0.5557, |
| "step": 148 |
| }, |
| { |
| "epoch": 1.4304000000000001, |
| "grad_norm": 0.5241117477416992, |
| "learning_rate": 6.276023446318214e-06, |
| "loss": 0.5615, |
| "step": 149 |
| }, |
| { |
| "epoch": 1.44, |
| "grad_norm": 0.5613789558410645, |
| "learning_rate": 6.2217020306894705e-06, |
| "loss": 0.6395, |
| "step": 150 |
| }, |
| { |
| "epoch": 1.4496, |
| "grad_norm": 0.46702006459236145, |
| "learning_rate": 6.1672268192795285e-06, |
| "loss": 0.5018, |
| "step": 151 |
| }, |
| { |
| "epoch": 1.4592, |
| "grad_norm": 0.5378797650337219, |
| "learning_rate": 6.112604669781572e-06, |
| "loss": 0.5834, |
| "step": 152 |
| }, |
| { |
| "epoch": 1.4687999999999999, |
| "grad_norm": 0.5464416742324829, |
| "learning_rate": 6.057842458386315e-06, |
| "loss": 0.6352, |
| "step": 153 |
| }, |
| { |
| "epoch": 1.4784, |
| "grad_norm": 0.5009851455688477, |
| "learning_rate": 6.002947078916365e-06, |
| "loss": 0.5688, |
| "step": 154 |
| }, |
| { |
| "epoch": 1.488, |
| "grad_norm": 0.5060755014419556, |
| "learning_rate": 5.947925441958393e-06, |
| "loss": 0.5453, |
| "step": 155 |
| }, |
| { |
| "epoch": 1.4976, |
| "grad_norm": 0.45909592509269714, |
| "learning_rate": 5.892784473993184e-06, |
| "loss": 0.5977, |
| "step": 156 |
| }, |
| { |
| "epoch": 1.5072, |
| "grad_norm": 0.5472169518470764, |
| "learning_rate": 5.837531116523683e-06, |
| "loss": 0.6206, |
| "step": 157 |
| }, |
| { |
| "epoch": 1.5168, |
| "grad_norm": 0.48066723346710205, |
| "learning_rate": 5.782172325201155e-06, |
| "loss": 0.5773, |
| "step": 158 |
| }, |
| { |
| "epoch": 1.5264, |
| "grad_norm": 0.4873718321323395, |
| "learning_rate": 5.726715068949564e-06, |
| "loss": 0.5291, |
| "step": 159 |
| }, |
| { |
| "epoch": 1.536, |
| "grad_norm": 0.4835259020328522, |
| "learning_rate": 5.671166329088278e-06, |
| "loss": 0.5829, |
| "step": 160 |
| }, |
| { |
| "epoch": 1.5455999999999999, |
| "grad_norm": 0.47710105776786804, |
| "learning_rate": 5.615533098453215e-06, |
| "loss": 0.5708, |
| "step": 161 |
| }, |
| { |
| "epoch": 1.5552000000000001, |
| "grad_norm": 0.46555811166763306, |
| "learning_rate": 5.559822380516539e-06, |
| "loss": 0.5642, |
| "step": 162 |
| }, |
| { |
| "epoch": 1.5648, |
| "grad_norm": 0.5088909864425659, |
| "learning_rate": 5.504041188505022e-06, |
| "loss": 0.622, |
| "step": 163 |
| }, |
| { |
| "epoch": 1.5744, |
| "grad_norm": 0.48639026284217834, |
| "learning_rate": 5.448196544517168e-06, |
| "loss": 0.554, |
| "step": 164 |
| }, |
| { |
| "epoch": 1.584, |
| "grad_norm": 0.46070578694343567, |
| "learning_rate": 5.392295478639226e-06, |
| "loss": 0.5633, |
| "step": 165 |
| }, |
| { |
| "epoch": 1.5936, |
| "grad_norm": 0.519040048122406, |
| "learning_rate": 5.336345028060199e-06, |
| "loss": 0.6354, |
| "step": 166 |
| }, |
| { |
| "epoch": 1.6032, |
| "grad_norm": 0.46838951110839844, |
| "learning_rate": 5.2803522361859596e-06, |
| "loss": 0.5565, |
| "step": 167 |
| }, |
| { |
| "epoch": 1.6128, |
| "grad_norm": 0.47723791003227234, |
| "learning_rate": 5.224324151752575e-06, |
| "loss": 0.5827, |
| "step": 168 |
| }, |
| { |
| "epoch": 1.6223999999999998, |
| "grad_norm": 0.44898900389671326, |
| "learning_rate": 5.168267827938971e-06, |
| "loss": 0.5256, |
| "step": 169 |
| }, |
| { |
| "epoch": 1.6320000000000001, |
| "grad_norm": 0.48522815108299255, |
| "learning_rate": 5.112190321479026e-06, |
| "loss": 0.5899, |
| "step": 170 |
| }, |
| { |
| "epoch": 1.6416, |
| "grad_norm": 0.5246551632881165, |
| "learning_rate": 5.05609869177323e-06, |
| "loss": 0.5808, |
| "step": 171 |
| }, |
| { |
| "epoch": 1.6512, |
| "grad_norm": 0.4806668162345886, |
| "learning_rate": 5e-06, |
| "loss": 0.5526, |
| "step": 172 |
| }, |
| { |
| "epoch": 1.6608, |
| "grad_norm": 0.4761640429496765, |
| "learning_rate": 4.943901308226771e-06, |
| "loss": 0.5866, |
| "step": 173 |
| }, |
| { |
| "epoch": 1.6703999999999999, |
| "grad_norm": 0.41892242431640625, |
| "learning_rate": 4.887809678520976e-06, |
| "loss": 0.5488, |
| "step": 174 |
| }, |
| { |
| "epoch": 1.6800000000000002, |
| "grad_norm": 0.518663227558136, |
| "learning_rate": 4.831732172061032e-06, |
| "loss": 0.6082, |
| "step": 175 |
| }, |
| { |
| "epoch": 1.6896, |
| "grad_norm": 0.4798682630062103, |
| "learning_rate": 4.775675848247427e-06, |
| "loss": 0.5804, |
| "step": 176 |
| }, |
| { |
| "epoch": 1.6992, |
| "grad_norm": 0.46521127223968506, |
| "learning_rate": 4.719647763814041e-06, |
| "loss": 0.6152, |
| "step": 177 |
| }, |
| { |
| "epoch": 1.7088, |
| "grad_norm": 0.478670597076416, |
| "learning_rate": 4.663654971939802e-06, |
| "loss": 0.575, |
| "step": 178 |
| }, |
| { |
| "epoch": 1.7184, |
| "grad_norm": 0.44683733582496643, |
| "learning_rate": 4.6077045213607765e-06, |
| "loss": 0.5496, |
| "step": 179 |
| }, |
| { |
| "epoch": 1.728, |
| "grad_norm": 0.5028547644615173, |
| "learning_rate": 4.551803455482833e-06, |
| "loss": 0.6111, |
| "step": 180 |
| }, |
| { |
| "epoch": 1.7376, |
| "grad_norm": 0.47960808873176575, |
| "learning_rate": 4.4959588114949785e-06, |
| "loss": 0.6407, |
| "step": 181 |
| }, |
| { |
| "epoch": 1.7471999999999999, |
| "grad_norm": 0.45923641324043274, |
| "learning_rate": 4.4401776194834615e-06, |
| "loss": 0.5702, |
| "step": 182 |
| }, |
| { |
| "epoch": 1.7568000000000001, |
| "grad_norm": 0.5147970914840698, |
| "learning_rate": 4.384466901546786e-06, |
| "loss": 0.5645, |
| "step": 183 |
| }, |
| { |
| "epoch": 1.7664, |
| "grad_norm": 0.48651739954948425, |
| "learning_rate": 4.3288336709117246e-06, |
| "loss": 0.5922, |
| "step": 184 |
| }, |
| { |
| "epoch": 1.776, |
| "grad_norm": 0.46390658617019653, |
| "learning_rate": 4.273284931050438e-06, |
| "loss": 0.5706, |
| "step": 185 |
| }, |
| { |
| "epoch": 1.7856, |
| "grad_norm": 0.4808691442012787, |
| "learning_rate": 4.217827674798845e-06, |
| "loss": 0.5703, |
| "step": 186 |
| }, |
| { |
| "epoch": 1.7952, |
| "grad_norm": 0.49175235629081726, |
| "learning_rate": 4.162468883476319e-06, |
| "loss": 0.6144, |
| "step": 187 |
| }, |
| { |
| "epoch": 1.8048, |
| "grad_norm": 0.4273785948753357, |
| "learning_rate": 4.107215526006818e-06, |
| "loss": 0.5436, |
| "step": 188 |
| }, |
| { |
| "epoch": 1.8144, |
| "grad_norm": 0.44895637035369873, |
| "learning_rate": 4.052074558041608e-06, |
| "loss": 0.5621, |
| "step": 189 |
| }, |
| { |
| "epoch": 1.8239999999999998, |
| "grad_norm": 0.4793059825897217, |
| "learning_rate": 3.997052921083637e-06, |
| "loss": 0.5379, |
| "step": 190 |
| }, |
| { |
| "epoch": 1.8336000000000001, |
| "grad_norm": 0.5026523470878601, |
| "learning_rate": 3.9421575416136866e-06, |
| "loss": 0.6023, |
| "step": 191 |
| }, |
| { |
| "epoch": 1.8432, |
| "grad_norm": 0.4519679844379425, |
| "learning_rate": 3.887395330218429e-06, |
| "loss": 0.5087, |
| "step": 192 |
| }, |
| { |
| "epoch": 1.8528, |
| "grad_norm": 0.49146002531051636, |
| "learning_rate": 3.832773180720475e-06, |
| "loss": 0.624, |
| "step": 193 |
| }, |
| { |
| "epoch": 1.8624, |
| "grad_norm": 0.5234208703041077, |
| "learning_rate": 3.778297969310529e-06, |
| "loss": 0.5437, |
| "step": 194 |
| }, |
| { |
| "epoch": 1.8719999999999999, |
| "grad_norm": 0.47663241624832153, |
| "learning_rate": 3.723976553681787e-06, |
| "loss": 0.5639, |
| "step": 195 |
| }, |
| { |
| "epoch": 1.8816000000000002, |
| "grad_norm": 0.48648905754089355, |
| "learning_rate": 3.669815772166625e-06, |
| "loss": 0.6106, |
| "step": 196 |
| }, |
| { |
| "epoch": 1.8912, |
| "grad_norm": 0.4679903984069824, |
| "learning_rate": 3.6158224428757538e-06, |
| "loss": 0.5526, |
| "step": 197 |
| }, |
| { |
| "epoch": 1.9008, |
| "grad_norm": 0.49357014894485474, |
| "learning_rate": 3.562003362839914e-06, |
| "loss": 0.566, |
| "step": 198 |
| }, |
| { |
| "epoch": 1.9104, |
| "grad_norm": 0.45932838320732117, |
| "learning_rate": 3.50836530715422e-06, |
| "loss": 0.6054, |
| "step": 199 |
| }, |
| { |
| "epoch": 1.92, |
| "grad_norm": 0.43980926275253296, |
| "learning_rate": 3.4549150281252635e-06, |
| "loss": 0.5719, |
| "step": 200 |
| }, |
| { |
| "epoch": 1.9296, |
| "grad_norm": 0.44906845688819885, |
| "learning_rate": 3.4016592544210937e-06, |
| "loss": 0.592, |
| "step": 201 |
| }, |
| { |
| "epoch": 1.9392, |
| "grad_norm": 0.4851302206516266, |
| "learning_rate": 3.3486046902241663e-06, |
| "loss": 0.5758, |
| "step": 202 |
| }, |
| { |
| "epoch": 1.9487999999999999, |
| "grad_norm": 0.45291900634765625, |
| "learning_rate": 3.295758014387375e-06, |
| "loss": 0.5487, |
| "step": 203 |
| }, |
| { |
| "epoch": 1.9584000000000001, |
| "grad_norm": 0.4484531879425049, |
| "learning_rate": 3.2431258795932863e-06, |
| "loss": 0.5636, |
| "step": 204 |
| }, |
| { |
| "epoch": 1.968, |
| "grad_norm": 0.48709994554519653, |
| "learning_rate": 3.1907149115166403e-06, |
| "loss": 0.632, |
| "step": 205 |
| }, |
| { |
| "epoch": 1.9776, |
| "grad_norm": 0.4566728174686432, |
| "learning_rate": 3.1385317079902743e-06, |
| "loss": 0.5902, |
| "step": 206 |
| }, |
| { |
| "epoch": 1.9872, |
| "grad_norm": 0.48063844442367554, |
| "learning_rate": 3.0865828381745515e-06, |
| "loss": 0.5484, |
| "step": 207 |
| }, |
| { |
| "epoch": 1.9968, |
| "grad_norm": 0.42609474062919617, |
| "learning_rate": 3.0348748417303826e-06, |
| "loss": 0.5969, |
| "step": 208 |
| }, |
| { |
| "epoch": 2.0064, |
| "grad_norm": 1.4519919157028198, |
| "learning_rate": 2.9834142279959754e-06, |
| "loss": 1.0623, |
| "step": 209 |
| }, |
| { |
| "epoch": 2.016, |
| "grad_norm": 0.4431564509868622, |
| "learning_rate": 2.932207475167398e-06, |
| "loss": 0.4855, |
| "step": 210 |
| }, |
| { |
| "epoch": 2.0256, |
| "grad_norm": 0.5621756315231323, |
| "learning_rate": 2.8812610294830568e-06, |
| "loss": 0.58, |
| "step": 211 |
| }, |
| { |
| "epoch": 2.0352, |
| "grad_norm": 0.45312392711639404, |
| "learning_rate": 2.83058130441221e-06, |
| "loss": 0.457, |
| "step": 212 |
| }, |
| { |
| "epoch": 2.0448, |
| "grad_norm": 0.48333898186683655, |
| "learning_rate": 2.7801746798475905e-06, |
| "loss": 0.5953, |
| "step": 213 |
| }, |
| { |
| "epoch": 2.0544, |
| "grad_norm": 0.44254645705223083, |
| "learning_rate": 2.7300475013022666e-06, |
| "loss": 0.5382, |
| "step": 214 |
| }, |
| { |
| "epoch": 2.064, |
| "grad_norm": 0.4953754246234894, |
| "learning_rate": 2.6802060791108304e-06, |
| "loss": 0.5842, |
| "step": 215 |
| }, |
| { |
| "epoch": 2.0736, |
| "grad_norm": 0.4534076154232025, |
| "learning_rate": 2.6306566876350072e-06, |
| "loss": 0.4798, |
| "step": 216 |
| }, |
| { |
| "epoch": 2.0832, |
| "grad_norm": 0.47140488028526306, |
| "learning_rate": 2.5814055644738013e-06, |
| "loss": 0.5516, |
| "step": 217 |
| }, |
| { |
| "epoch": 2.0928, |
| "grad_norm": 0.4574890732765198, |
| "learning_rate": 2.532458909678266e-06, |
| "loss": 0.5355, |
| "step": 218 |
| }, |
| { |
| "epoch": 2.1024, |
| "grad_norm": 0.4405193328857422, |
| "learning_rate": 2.483822884971e-06, |
| "loss": 0.5414, |
| "step": 219 |
| }, |
| { |
| "epoch": 2.112, |
| "grad_norm": 0.4610828757286072, |
| "learning_rate": 2.43550361297047e-06, |
| "loss": 0.54, |
| "step": 220 |
| }, |
| { |
| "epoch": 2.1216, |
| "grad_norm": 0.5018312335014343, |
| "learning_rate": 2.387507176420256e-06, |
| "loss": 0.5172, |
| "step": 221 |
| }, |
| { |
| "epoch": 2.1312, |
| "grad_norm": 0.4683016240596771, |
| "learning_rate": 2.339839617423318e-06, |
| "loss": 0.571, |
| "step": 222 |
| }, |
| { |
| "epoch": 2.1408, |
| "grad_norm": 0.4281395375728607, |
| "learning_rate": 2.2925069366813718e-06, |
| "loss": 0.5108, |
| "step": 223 |
| }, |
| { |
| "epoch": 2.1504, |
| "grad_norm": 0.4399942457675934, |
| "learning_rate": 2.245515092739488e-06, |
| "loss": 0.5313, |
| "step": 224 |
| }, |
| { |
| "epoch": 2.16, |
| "grad_norm": 0.4210817515850067, |
| "learning_rate": 2.1988700012359865e-06, |
| "loss": 0.509, |
| "step": 225 |
| }, |
| { |
| "epoch": 2.1696, |
| "grad_norm": 0.4607466757297516, |
| "learning_rate": 2.1525775341577404e-06, |
| "loss": 0.5383, |
| "step": 226 |
| }, |
| { |
| "epoch": 2.1792, |
| "grad_norm": 0.44297847151756287, |
| "learning_rate": 2.1066435191009717e-06, |
| "loss": 0.5753, |
| "step": 227 |
| }, |
| { |
| "epoch": 2.1888, |
| "grad_norm": 0.4266228973865509, |
| "learning_rate": 2.061073738537635e-06, |
| "loss": 0.5555, |
| "step": 228 |
| }, |
| { |
| "epoch": 2.1984, |
| "grad_norm": 0.4219799339771271, |
| "learning_rate": 2.0158739290874822e-06, |
| "loss": 0.5427, |
| "step": 229 |
| }, |
| { |
| "epoch": 2.208, |
| "grad_norm": 0.4710475504398346, |
| "learning_rate": 1.971049780795901e-06, |
| "loss": 0.5859, |
| "step": 230 |
| }, |
| { |
| "epoch": 2.2176, |
| "grad_norm": 0.388872355222702, |
| "learning_rate": 1.9266069364176144e-06, |
| "loss": 0.4457, |
| "step": 231 |
| }, |
| { |
| "epoch": 2.2272, |
| "grad_norm": 0.40161120891571045, |
| "learning_rate": 1.8825509907063328e-06, |
| "loss": 0.5098, |
| "step": 232 |
| }, |
| { |
| "epoch": 2.2368, |
| "grad_norm": 0.4724157452583313, |
| "learning_rate": 1.838887489710452e-06, |
| "loss": 0.5875, |
| "step": 233 |
| }, |
| { |
| "epoch": 2.2464, |
| "grad_norm": 0.42020970582962036, |
| "learning_rate": 1.7956219300748796e-06, |
| "loss": 0.4659, |
| "step": 234 |
| }, |
| { |
| "epoch": 2.2560000000000002, |
| "grad_norm": 0.4303283393383026, |
| "learning_rate": 1.7527597583490825e-06, |
| "loss": 0.5505, |
| "step": 235 |
| }, |
| { |
| "epoch": 2.2656, |
| "grad_norm": 0.42594870924949646, |
| "learning_rate": 1.7103063703014372e-06, |
| "loss": 0.5662, |
| "step": 236 |
| }, |
| { |
| "epoch": 2.2752, |
| "grad_norm": 0.3780567944049835, |
| "learning_rate": 1.6682671102399806e-06, |
| "loss": 0.4679, |
| "step": 237 |
| }, |
| { |
| "epoch": 2.2848, |
| "grad_norm": 0.4052414894104004, |
| "learning_rate": 1.6266472703396286e-06, |
| "loss": 0.5236, |
| "step": 238 |
| }, |
| { |
| "epoch": 2.2944, |
| "grad_norm": 0.4502789080142975, |
| "learning_rate": 1.5854520899759656e-06, |
| "loss": 0.5755, |
| "step": 239 |
| }, |
| { |
| "epoch": 2.304, |
| "grad_norm": 0.4016181528568268, |
| "learning_rate": 1.544686755065677e-06, |
| "loss": 0.4941, |
| "step": 240 |
| }, |
| { |
| "epoch": 2.3136, |
| "grad_norm": 0.3988753855228424, |
| "learning_rate": 1.5043563974137132e-06, |
| "loss": 0.5166, |
| "step": 241 |
| }, |
| { |
| "epoch": 2.3232, |
| "grad_norm": 0.4084855616092682, |
| "learning_rate": 1.4644660940672628e-06, |
| "loss": 0.5531, |
| "step": 242 |
| }, |
| { |
| "epoch": 2.3327999999999998, |
| "grad_norm": 0.3739737868309021, |
| "learning_rate": 1.4250208666766235e-06, |
| "loss": 0.5247, |
| "step": 243 |
| }, |
| { |
| "epoch": 2.3424, |
| "grad_norm": 0.39243167638778687, |
| "learning_rate": 1.3860256808630429e-06, |
| "loss": 0.5311, |
| "step": 244 |
| }, |
| { |
| "epoch": 2.352, |
| "grad_norm": 0.4281586706638336, |
| "learning_rate": 1.3474854455936126e-06, |
| "loss": 0.5644, |
| "step": 245 |
| }, |
| { |
| "epoch": 2.3616, |
| "grad_norm": 0.40751615166664124, |
| "learning_rate": 1.3094050125632973e-06, |
| "loss": 0.5418, |
| "step": 246 |
| }, |
| { |
| "epoch": 2.3712, |
| "grad_norm": 0.415406733751297, |
| "learning_rate": 1.2717891755841722e-06, |
| "loss": 0.5533, |
| "step": 247 |
| }, |
| { |
| "epoch": 2.3808, |
| "grad_norm": 0.39046332240104675, |
| "learning_rate": 1.234642669981946e-06, |
| "loss": 0.5763, |
| "step": 248 |
| }, |
| { |
| "epoch": 2.3904, |
| "grad_norm": 0.33598846197128296, |
| "learning_rate": 1.1979701719998454e-06, |
| "loss": 0.439, |
| "step": 249 |
| }, |
| { |
| "epoch": 2.4, |
| "grad_norm": 0.40691110491752625, |
| "learning_rate": 1.1617762982099446e-06, |
| "loss": 0.5848, |
| "step": 250 |
| }, |
| { |
| "epoch": 2.4096, |
| "grad_norm": 0.4017358124256134, |
| "learning_rate": 1.1260656049319957e-06, |
| "loss": 0.5434, |
| "step": 251 |
| }, |
| { |
| "epoch": 2.4192, |
| "grad_norm": 0.38580864667892456, |
| "learning_rate": 1.0908425876598512e-06, |
| "loss": 0.5361, |
| "step": 252 |
| }, |
| { |
| "epoch": 2.4288, |
| "grad_norm": 0.39464429020881653, |
| "learning_rate": 1.0561116804955451e-06, |
| "loss": 0.5276, |
| "step": 253 |
| }, |
| { |
| "epoch": 2.4384, |
| "grad_norm": 0.37000352144241333, |
| "learning_rate": 1.0218772555910955e-06, |
| "loss": 0.4963, |
| "step": 254 |
| }, |
| { |
| "epoch": 2.448, |
| "grad_norm": 0.4111914336681366, |
| "learning_rate": 9.881436225981107e-07, |
| "loss": 0.5652, |
| "step": 255 |
| }, |
| { |
| "epoch": 2.4576000000000002, |
| "grad_norm": 0.36240270733833313, |
| "learning_rate": 9.549150281252633e-07, |
| "loss": 0.4727, |
| "step": 256 |
| }, |
| { |
| "epoch": 2.4672, |
| "grad_norm": 0.39575356245040894, |
| "learning_rate": 9.221956552036992e-07, |
| "loss": 0.5634, |
| "step": 257 |
| }, |
| { |
| "epoch": 2.4768, |
| "grad_norm": 0.4152328670024872, |
| "learning_rate": 8.899896227604509e-07, |
| "loss": 0.5376, |
| "step": 258 |
| }, |
| { |
| "epoch": 2.4864, |
| "grad_norm": 0.39109793305397034, |
| "learning_rate": 8.58300985099918e-07, |
| "loss": 0.5368, |
| "step": 259 |
| }, |
| { |
| "epoch": 2.496, |
| "grad_norm": 0.4161483347415924, |
| "learning_rate": 8.271337313934869e-07, |
| "loss": 0.5086, |
| "step": 260 |
| }, |
| { |
| "epoch": 2.5056000000000003, |
| "grad_norm": 0.3931094706058502, |
| "learning_rate": 7.964917851773496e-07, |
| "loss": 0.5171, |
| "step": 261 |
| }, |
| { |
| "epoch": 2.5152, |
| "grad_norm": 0.3966805338859558, |
| "learning_rate": 7.663790038585794e-07, |
| "loss": 0.5109, |
| "step": 262 |
| }, |
| { |
| "epoch": 2.5248, |
| "grad_norm": 0.437146931886673, |
| "learning_rate": 7.367991782295392e-07, |
| "loss": 0.54, |
| "step": 263 |
| }, |
| { |
| "epoch": 2.5343999999999998, |
| "grad_norm": 0.39987269043922424, |
| "learning_rate": 7.077560319906696e-07, |
| "loss": 0.5305, |
| "step": 264 |
| }, |
| { |
| "epoch": 2.544, |
| "grad_norm": 0.43939921259880066, |
| "learning_rate": 6.792532212817271e-07, |
| "loss": 0.5249, |
| "step": 265 |
| }, |
| { |
| "epoch": 2.5536, |
| "grad_norm": 0.41188380122184753, |
| "learning_rate": 6.512943342215234e-07, |
| "loss": 0.5354, |
| "step": 266 |
| }, |
| { |
| "epoch": 2.5632, |
| "grad_norm": 0.38573968410491943, |
| "learning_rate": 6.238828904562316e-07, |
| "loss": 0.4925, |
| "step": 267 |
| }, |
| { |
| "epoch": 2.5728, |
| "grad_norm": 0.3975042700767517, |
| "learning_rate": 5.9702234071631e-07, |
| "loss": 0.5772, |
| "step": 268 |
| }, |
| { |
| "epoch": 2.5824, |
| "grad_norm": 0.3729751706123352, |
| "learning_rate": 5.707160663821009e-07, |
| "loss": 0.5495, |
| "step": 269 |
| }, |
| { |
| "epoch": 2.592, |
| "grad_norm": 0.3656686544418335, |
| "learning_rate": 5.449673790581611e-07, |
| "loss": 0.5153, |
| "step": 270 |
| }, |
| { |
| "epoch": 2.6016, |
| "grad_norm": 0.36057794094085693, |
| "learning_rate": 5.197795201563744e-07, |
| "loss": 0.4986, |
| "step": 271 |
| }, |
| { |
| "epoch": 2.6112, |
| "grad_norm": 0.3950442671775818, |
| "learning_rate": 4.951556604879049e-07, |
| "loss": 0.574, |
| "step": 272 |
| }, |
| { |
| "epoch": 2.6208, |
| "grad_norm": 0.39079609513282776, |
| "learning_rate": 4.710988998640298e-07, |
| "loss": 0.5261, |
| "step": 273 |
| }, |
| { |
| "epoch": 2.6304, |
| "grad_norm": 0.36852559447288513, |
| "learning_rate": 4.4761226670592074e-07, |
| "loss": 0.5647, |
| "step": 274 |
| }, |
| { |
| "epoch": 2.64, |
| "grad_norm": 0.36895784735679626, |
| "learning_rate": 4.2469871766340096e-07, |
| "loss": 0.5256, |
| "step": 275 |
| }, |
| { |
| "epoch": 2.6496, |
| "grad_norm": 0.39882954955101013, |
| "learning_rate": 4.0236113724274716e-07, |
| "loss": 0.5158, |
| "step": 276 |
| }, |
| { |
| "epoch": 2.6592000000000002, |
| "grad_norm": 0.3873459994792938, |
| "learning_rate": 3.8060233744356634e-07, |
| "loss": 0.5607, |
| "step": 277 |
| }, |
| { |
| "epoch": 2.6688, |
| "grad_norm": 0.36382660269737244, |
| "learning_rate": 3.5942505740480583e-07, |
| "loss": 0.4725, |
| "step": 278 |
| }, |
| { |
| "epoch": 2.6784, |
| "grad_norm": 0.4002612233161926, |
| "learning_rate": 3.3883196305992906e-07, |
| "loss": 0.596, |
| "step": 279 |
| }, |
| { |
| "epoch": 2.6879999999999997, |
| "grad_norm": 0.3573722541332245, |
| "learning_rate": 3.18825646801314e-07, |
| "loss": 0.4755, |
| "step": 280 |
| }, |
| { |
| "epoch": 2.6976, |
| "grad_norm": 0.36770984530448914, |
| "learning_rate": 2.9940862715390483e-07, |
| "loss": 0.5095, |
| "step": 281 |
| }, |
| { |
| "epoch": 2.7072000000000003, |
| "grad_norm": 0.42423680424690247, |
| "learning_rate": 2.8058334845816214e-07, |
| "loss": 0.5671, |
| "step": 282 |
| }, |
| { |
| "epoch": 2.7168, |
| "grad_norm": 0.38146907091140747, |
| "learning_rate": 2.6235218056235633e-07, |
| "loss": 0.549, |
| "step": 283 |
| }, |
| { |
| "epoch": 2.7264, |
| "grad_norm": 0.41113874316215515, |
| "learning_rate": 2.447174185242324e-07, |
| "loss": 0.5609, |
| "step": 284 |
| }, |
| { |
| "epoch": 2.7359999999999998, |
| "grad_norm": 0.3916343152523041, |
| "learning_rate": 2.276812823220964e-07, |
| "loss": 0.5369, |
| "step": 285 |
| }, |
| { |
| "epoch": 2.7456, |
| "grad_norm": 0.3916230797767639, |
| "learning_rate": 2.1124591657534776e-07, |
| "loss": 0.5398, |
| "step": 286 |
| }, |
| { |
| "epoch": 2.7552, |
| "grad_norm": 0.3757457137107849, |
| "learning_rate": 1.9541339027450256e-07, |
| "loss": 0.5394, |
| "step": 287 |
| }, |
| { |
| "epoch": 2.7648, |
| "grad_norm": 0.38745611906051636, |
| "learning_rate": 1.801856965207338e-07, |
| "loss": 0.5854, |
| "step": 288 |
| }, |
| { |
| "epoch": 2.7744, |
| "grad_norm": 0.35113686323165894, |
| "learning_rate": 1.6556475227496816e-07, |
| "loss": 0.496, |
| "step": 289 |
| }, |
| { |
| "epoch": 2.784, |
| "grad_norm": 0.34780657291412354, |
| "learning_rate": 1.5155239811656562e-07, |
| "loss": 0.5218, |
| "step": 290 |
| }, |
| { |
| "epoch": 2.7936, |
| "grad_norm": 0.4079935848712921, |
| "learning_rate": 1.3815039801161723e-07, |
| "loss": 0.5628, |
| "step": 291 |
| }, |
| { |
| "epoch": 2.8032, |
| "grad_norm": 0.4048634171485901, |
| "learning_rate": 1.253604390908819e-07, |
| "loss": 0.5707, |
| "step": 292 |
| }, |
| { |
| "epoch": 2.8128, |
| "grad_norm": 0.38722601532936096, |
| "learning_rate": 1.1318413143740436e-07, |
| "loss": 0.5215, |
| "step": 293 |
| }, |
| { |
| "epoch": 2.8224, |
| "grad_norm": 0.35742488503456116, |
| "learning_rate": 1.0162300788382263e-07, |
| "loss": 0.4985, |
| "step": 294 |
| }, |
| { |
| "epoch": 2.832, |
| "grad_norm": 0.3644346296787262, |
| "learning_rate": 9.0678523819408e-08, |
| "loss": 0.5123, |
| "step": 295 |
| }, |
| { |
| "epoch": 2.8416, |
| "grad_norm": 0.40099066495895386, |
| "learning_rate": 8.035205700685167e-08, |
| "loss": 0.4777, |
| "step": 296 |
| }, |
| { |
| "epoch": 2.8512, |
| "grad_norm": 0.38466301560401917, |
| "learning_rate": 7.064490740882057e-08, |
| "loss": 0.5441, |
| "step": 297 |
| }, |
| { |
| "epoch": 2.8608000000000002, |
| "grad_norm": 0.3442007601261139, |
| "learning_rate": 6.15582970243117e-08, |
| "loss": 0.5, |
| "step": 298 |
| }, |
| { |
| "epoch": 2.8704, |
| "grad_norm": 0.3603901267051697, |
| "learning_rate": 5.3093369734816824e-08, |
| "loss": 0.5128, |
| "step": 299 |
| }, |
| { |
| "epoch": 2.88, |
| "grad_norm": 0.36819523572921753, |
| "learning_rate": 4.52511911603265e-08, |
| "loss": 0.5467, |
| "step": 300 |
| }, |
| { |
| "epoch": 2.8895999999999997, |
| "grad_norm": 0.4061967432498932, |
| "learning_rate": 3.8032748525179684e-08, |
| "loss": 0.5714, |
| "step": 301 |
| }, |
| { |
| "epoch": 2.8992, |
| "grad_norm": 0.3809182047843933, |
| "learning_rate": 3.143895053378698e-08, |
| "loss": 0.5694, |
| "step": 302 |
| }, |
| { |
| "epoch": 2.9088000000000003, |
| "grad_norm": 0.35255205631256104, |
| "learning_rate": 2.547062725623828e-08, |
| "loss": 0.5109, |
| "step": 303 |
| }, |
| { |
| "epoch": 2.9184, |
| "grad_norm": 0.3475726246833801, |
| "learning_rate": 2.012853002380466e-08, |
| "loss": 0.4722, |
| "step": 304 |
| }, |
| { |
| "epoch": 2.928, |
| "grad_norm": 0.3844062089920044, |
| "learning_rate": 1.541333133436018e-08, |
| "loss": 0.5612, |
| "step": 305 |
| }, |
| { |
| "epoch": 2.9375999999999998, |
| "grad_norm": 0.39250507950782776, |
| "learning_rate": 1.132562476771959e-08, |
| "loss": 0.5593, |
| "step": 306 |
| }, |
| { |
| "epoch": 2.9472, |
| "grad_norm": 0.4356182813644409, |
| "learning_rate": 7.865924910916977e-09, |
| "loss": 0.5713, |
| "step": 307 |
| }, |
| { |
| "epoch": 2.9568, |
| "grad_norm": 0.3923850655555725, |
| "learning_rate": 5.034667293427053e-09, |
| "loss": 0.4919, |
| "step": 308 |
| }, |
| { |
| "epoch": 2.9664, |
| "grad_norm": 0.38620316982269287, |
| "learning_rate": 2.8322083323334417e-09, |
| "loss": 0.5161, |
| "step": 309 |
| }, |
| { |
| "epoch": 2.976, |
| "grad_norm": 0.4041982889175415, |
| "learning_rate": 1.2588252874673469e-09, |
| "loss": 0.5558, |
| "step": 310 |
| }, |
| { |
| "epoch": 2.9856, |
| "grad_norm": 0.38204097747802734, |
| "learning_rate": 3.147162264971471e-10, |
| "loss": 0.5324, |
| "step": 311 |
| }, |
| { |
| "epoch": 2.9952, |
| "grad_norm": 0.3853088915348053, |
| "learning_rate": 0.0, |
| "loss": 0.5285, |
| "step": 312 |
| }, |
| { |
| "epoch": 2.9952, |
| "step": 312, |
| "total_flos": 4.700708732613427e+17, |
| "train_loss": 0.6130588305875276, |
| "train_runtime": 28658.3303, |
| "train_samples_per_second": 1.047, |
| "train_steps_per_second": 0.011 |
| } |
| ], |
| "logging_steps": 1.0, |
| "max_steps": 312, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "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": 4.700708732613427e+17, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
