| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 112500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.044444444444444446, | |
| "grad_norm": 2.7033166885375977, | |
| "learning_rate": 1.9911644444444447e-05, | |
| "loss": 2.2268, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.08888888888888889, | |
| "grad_norm": 2.958282232284546, | |
| "learning_rate": 1.9822755555555557e-05, | |
| "loss": 2.0732, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.13333333333333333, | |
| "grad_norm": 3.01822829246521, | |
| "learning_rate": 1.9733866666666668e-05, | |
| "loss": 2.0296, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.17777777777777778, | |
| "grad_norm": 2.8041861057281494, | |
| "learning_rate": 1.9644977777777778e-05, | |
| "loss": 2.0176, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 2.194178342819214, | |
| "learning_rate": 1.9556088888888892e-05, | |
| "loss": 2.0181, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "grad_norm": 2.7587132453918457, | |
| "learning_rate": 1.9467200000000002e-05, | |
| "loss": 1.9864, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.3111111111111111, | |
| "grad_norm": 2.8509016036987305, | |
| "learning_rate": 1.9378311111111113e-05, | |
| "loss": 1.9701, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 2.180856704711914, | |
| "learning_rate": 1.9289422222222223e-05, | |
| "loss": 1.936, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 2.2034873962402344, | |
| "learning_rate": 1.9200533333333337e-05, | |
| "loss": 1.9455, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 3.1408488750457764, | |
| "learning_rate": 1.9111644444444447e-05, | |
| "loss": 1.9438, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.4888888888888889, | |
| "grad_norm": 3.070190191268921, | |
| "learning_rate": 1.9022755555555558e-05, | |
| "loss": 1.9491, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 2.0312423706054688, | |
| "learning_rate": 1.893386666666667e-05, | |
| "loss": 1.9236, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.5777777777777777, | |
| "grad_norm": 2.183950424194336, | |
| "learning_rate": 1.884497777777778e-05, | |
| "loss": 1.9111, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.6222222222222222, | |
| "grad_norm": 2.13474702835083, | |
| "learning_rate": 1.875608888888889e-05, | |
| "loss": 1.9158, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 2.055859327316284, | |
| "learning_rate": 1.866737777777778e-05, | |
| "loss": 1.8952, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 2.280942916870117, | |
| "learning_rate": 1.857848888888889e-05, | |
| "loss": 1.9023, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.7555555555555555, | |
| "grad_norm": 3.20082426071167, | |
| "learning_rate": 1.84896e-05, | |
| "loss": 1.8711, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.5550222396850586, | |
| "learning_rate": 1.8400711111111114e-05, | |
| "loss": 1.8865, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.8444444444444444, | |
| "grad_norm": 3.101032018661499, | |
| "learning_rate": 1.8311822222222224e-05, | |
| "loss": 1.8942, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 1.9913034439086914, | |
| "learning_rate": 1.8222933333333335e-05, | |
| "loss": 1.8823, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.9333333333333333, | |
| "grad_norm": 1.8786649703979492, | |
| "learning_rate": 1.8134044444444445e-05, | |
| "loss": 1.8903, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.9777777777777777, | |
| "grad_norm": 2.1240804195404053, | |
| "learning_rate": 1.804515555555556e-05, | |
| "loss": 1.8759, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_Sacrebleu": 10.927683654987915, | |
| "eval_loss": 1.7730144262313843, | |
| "eval_runtime": 5155.9444, | |
| "eval_samples_per_second": 1.94, | |
| "eval_steps_per_second": 0.242, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 1.0222222222222221, | |
| "grad_norm": 3.4717421531677246, | |
| "learning_rate": 1.795626666666667e-05, | |
| "loss": 1.809, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 2.8455264568328857, | |
| "learning_rate": 1.786755555555556e-05, | |
| "loss": 1.7839, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 2.3160150051116943, | |
| "learning_rate": 1.7778844444444446e-05, | |
| "loss": 1.7898, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.1555555555555554, | |
| "grad_norm": 2.7126030921936035, | |
| "learning_rate": 1.768995555555556e-05, | |
| "loss": 1.7653, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 2.833963394165039, | |
| "learning_rate": 1.760106666666667e-05, | |
| "loss": 1.7937, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.2444444444444445, | |
| "grad_norm": 2.582491159439087, | |
| "learning_rate": 1.751217777777778e-05, | |
| "loss": 1.7672, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.2888888888888888, | |
| "grad_norm": 2.6020431518554688, | |
| "learning_rate": 1.742346666666667e-05, | |
| "loss": 1.7788, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 2.6280031204223633, | |
| "learning_rate": 1.733457777777778e-05, | |
| "loss": 1.7617, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.3777777777777778, | |
| "grad_norm": 2.258315324783325, | |
| "learning_rate": 1.724568888888889e-05, | |
| "loss": 1.7603, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.4222222222222223, | |
| "grad_norm": 2.546867847442627, | |
| "learning_rate": 1.71568e-05, | |
| "loss": 1.7639, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.4666666666666668, | |
| "grad_norm": 2.596524715423584, | |
| "learning_rate": 1.7067911111111112e-05, | |
| "loss": 1.7787, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.511111111111111, | |
| "grad_norm": 2.728980541229248, | |
| "learning_rate": 1.6979022222222222e-05, | |
| "loss": 1.7728, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.5555555555555556, | |
| "grad_norm": 1.9743603467941284, | |
| "learning_rate": 1.6890133333333333e-05, | |
| "loss": 1.7436, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 2.306102991104126, | |
| "learning_rate": 1.6801422222222223e-05, | |
| "loss": 1.7593, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.6444444444444444, | |
| "grad_norm": 2.595468282699585, | |
| "learning_rate": 1.6712533333333333e-05, | |
| "loss": 1.7587, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 1.6888888888888889, | |
| "grad_norm": 1.8225383758544922, | |
| "learning_rate": 1.6623644444444447e-05, | |
| "loss": 1.7852, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.7333333333333334, | |
| "grad_norm": 2.581843852996826, | |
| "learning_rate": 1.6534755555555557e-05, | |
| "loss": 1.7582, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 1.9384685754776, | |
| "learning_rate": 1.6445866666666668e-05, | |
| "loss": 1.7544, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.8222222222222222, | |
| "grad_norm": 2.542980670928955, | |
| "learning_rate": 1.6356977777777778e-05, | |
| "loss": 1.7592, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 1.8666666666666667, | |
| "grad_norm": 2.4207804203033447, | |
| "learning_rate": 1.6268088888888892e-05, | |
| "loss": 1.7685, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.911111111111111, | |
| "grad_norm": 2.7720110416412354, | |
| "learning_rate": 1.617937777777778e-05, | |
| "loss": 1.7654, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 1.9555555555555557, | |
| "grad_norm": 2.8182454109191895, | |
| "learning_rate": 1.6090488888888892e-05, | |
| "loss": 1.7577, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.3921427726745605, | |
| "learning_rate": 1.6001600000000003e-05, | |
| "loss": 1.7666, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_Sacrebleu": 11.504795316238502, | |
| "eval_loss": 1.7359962463378906, | |
| "eval_runtime": 4723.0089, | |
| "eval_samples_per_second": 2.117, | |
| "eval_steps_per_second": 0.265, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.0444444444444443, | |
| "grad_norm": 2.5132415294647217, | |
| "learning_rate": 1.5912711111111113e-05, | |
| "loss": 1.6699, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 2.088888888888889, | |
| "grad_norm": 2.244852066040039, | |
| "learning_rate": 1.5823822222222224e-05, | |
| "loss": 1.6636, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 2.456984281539917, | |
| "learning_rate": 1.5734933333333334e-05, | |
| "loss": 1.6812, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.1777777777777776, | |
| "grad_norm": 2.611400842666626, | |
| "learning_rate": 1.5646044444444445e-05, | |
| "loss": 1.6813, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 2.504777431488037, | |
| "learning_rate": 1.5557155555555555e-05, | |
| "loss": 1.6698, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.2666666666666666, | |
| "grad_norm": 1.9886395931243896, | |
| "learning_rate": 1.5468444444444445e-05, | |
| "loss": 1.6649, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 2.311111111111111, | |
| "grad_norm": 2.4218804836273193, | |
| "learning_rate": 1.5379555555555555e-05, | |
| "loss": 1.6827, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.3555555555555556, | |
| "grad_norm": 2.380363941192627, | |
| "learning_rate": 1.5290666666666666e-05, | |
| "loss": 1.6821, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 2.9390811920166016, | |
| "learning_rate": 1.520177777777778e-05, | |
| "loss": 1.6517, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.4444444444444446, | |
| "grad_norm": 2.278156042098999, | |
| "learning_rate": 1.511288888888889e-05, | |
| "loss": 1.6551, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 2.488888888888889, | |
| "grad_norm": 2.5583651065826416, | |
| "learning_rate": 1.5024e-05, | |
| "loss": 1.6386, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.533333333333333, | |
| "grad_norm": 2.4227840900421143, | |
| "learning_rate": 1.493528888888889e-05, | |
| "loss": 1.6808, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 2.5777777777777775, | |
| "grad_norm": 2.4636645317077637, | |
| "learning_rate": 1.48464e-05, | |
| "loss": 1.6675, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 2.6222222222222222, | |
| "grad_norm": 1.932691216468811, | |
| "learning_rate": 1.4757511111111111e-05, | |
| "loss": 1.6649, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 2.324248790740967, | |
| "learning_rate": 1.4668622222222223e-05, | |
| "loss": 1.641, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.7111111111111112, | |
| "grad_norm": 2.5001227855682373, | |
| "learning_rate": 1.4579733333333335e-05, | |
| "loss": 1.6593, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 2.7555555555555555, | |
| "grad_norm": 2.053122043609619, | |
| "learning_rate": 1.4490844444444446e-05, | |
| "loss": 1.6679, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 2.3404734134674072, | |
| "learning_rate": 1.4401955555555556e-05, | |
| "loss": 1.6705, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 2.8444444444444446, | |
| "grad_norm": 1.811047911643982, | |
| "learning_rate": 1.4313066666666669e-05, | |
| "loss": 1.6573, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 2.888888888888889, | |
| "grad_norm": 2.5269010066986084, | |
| "learning_rate": 1.4224355555555555e-05, | |
| "loss": 1.6659, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 2.9333333333333336, | |
| "grad_norm": 3.2324156761169434, | |
| "learning_rate": 1.4135466666666669e-05, | |
| "loss": 1.6509, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 2.977777777777778, | |
| "grad_norm": 2.910116195678711, | |
| "learning_rate": 1.404657777777778e-05, | |
| "loss": 1.6799, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_Sacrebleu": 11.461615098924428, | |
| "eval_loss": 1.7215642929077148, | |
| "eval_runtime": 4737.4626, | |
| "eval_samples_per_second": 2.111, | |
| "eval_steps_per_second": 0.264, | |
| "step": 33750 | |
| }, | |
| { | |
| "epoch": 3.022222222222222, | |
| "grad_norm": 2.388887643814087, | |
| "learning_rate": 1.395768888888889e-05, | |
| "loss": 1.6205, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 3.066666666666667, | |
| "grad_norm": 2.422229290008545, | |
| "learning_rate": 1.38688e-05, | |
| "loss": 1.5935, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 3.111111111111111, | |
| "grad_norm": 2.689824104309082, | |
| "learning_rate": 1.378008888888889e-05, | |
| "loss": 1.5677, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 3.1555555555555554, | |
| "grad_norm": 2.5040597915649414, | |
| "learning_rate": 1.369137777777778e-05, | |
| "loss": 1.5982, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 2.0994420051574707, | |
| "learning_rate": 1.360248888888889e-05, | |
| "loss": 1.6006, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 3.2444444444444445, | |
| "grad_norm": 2.4700512886047363, | |
| "learning_rate": 1.35136e-05, | |
| "loss": 1.594, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 3.2888888888888888, | |
| "grad_norm": 2.679499864578247, | |
| "learning_rate": 1.3424711111111113e-05, | |
| "loss": 1.5901, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 2.6343369483947754, | |
| "learning_rate": 1.3335822222222223e-05, | |
| "loss": 1.5875, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 3.3777777777777778, | |
| "grad_norm": 2.5335209369659424, | |
| "learning_rate": 1.3247111111111113e-05, | |
| "loss": 1.6139, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 3.422222222222222, | |
| "grad_norm": 2.9107964038848877, | |
| "learning_rate": 1.3158222222222223e-05, | |
| "loss": 1.6033, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 3.466666666666667, | |
| "grad_norm": 2.534843921661377, | |
| "learning_rate": 1.3069333333333334e-05, | |
| "loss": 1.584, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 3.511111111111111, | |
| "grad_norm": 2.409266710281372, | |
| "learning_rate": 1.2980444444444444e-05, | |
| "loss": 1.5854, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 2.509253740310669, | |
| "learning_rate": 1.2891555555555556e-05, | |
| "loss": 1.5809, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 1.9488080739974976, | |
| "learning_rate": 1.2802666666666667e-05, | |
| "loss": 1.5889, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 3.6444444444444444, | |
| "grad_norm": 2.8996152877807617, | |
| "learning_rate": 1.2713777777777779e-05, | |
| "loss": 1.5807, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 3.688888888888889, | |
| "grad_norm": 2.444199800491333, | |
| "learning_rate": 1.2624888888888891e-05, | |
| "loss": 1.6047, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 3.7333333333333334, | |
| "grad_norm": 2.627521276473999, | |
| "learning_rate": 1.2536000000000002e-05, | |
| "loss": 1.582, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 3.7777777777777777, | |
| "grad_norm": 2.4339983463287354, | |
| "learning_rate": 1.2447288888888891e-05, | |
| "loss": 1.5887, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 3.822222222222222, | |
| "grad_norm": 2.3496522903442383, | |
| "learning_rate": 1.2358577777777778e-05, | |
| "loss": 1.5878, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 3.8666666666666667, | |
| "grad_norm": 2.325495719909668, | |
| "learning_rate": 1.2269688888888892e-05, | |
| "loss": 1.5826, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 3.911111111111111, | |
| "grad_norm": 2.1297905445098877, | |
| "learning_rate": 1.2180800000000002e-05, | |
| "loss": 1.6244, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 3.9555555555555557, | |
| "grad_norm": 2.5512731075286865, | |
| "learning_rate": 1.2091911111111112e-05, | |
| "loss": 1.6075, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.9816246032714844, | |
| "learning_rate": 1.2003200000000002e-05, | |
| "loss": 1.5647, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_Sacrebleu": 11.693143227743814, | |
| "eval_loss": 1.7151726484298706, | |
| "eval_runtime": 4558.3443, | |
| "eval_samples_per_second": 2.194, | |
| "eval_steps_per_second": 0.274, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 4.044444444444444, | |
| "grad_norm": 2.9890973567962646, | |
| "learning_rate": 1.1914311111111113e-05, | |
| "loss": 1.5201, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 4.088888888888889, | |
| "grad_norm": 2.128161668777466, | |
| "learning_rate": 1.1825422222222223e-05, | |
| "loss": 1.5305, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 4.133333333333334, | |
| "grad_norm": 2.3265063762664795, | |
| "learning_rate": 1.1736533333333335e-05, | |
| "loss": 1.5326, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 4.177777777777778, | |
| "grad_norm": 2.2641632556915283, | |
| "learning_rate": 1.1647644444444446e-05, | |
| "loss": 1.5413, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 4.222222222222222, | |
| "grad_norm": 2.4255621433258057, | |
| "learning_rate": 1.1558755555555556e-05, | |
| "loss": 1.5391, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 4.266666666666667, | |
| "grad_norm": 2.454061269760132, | |
| "learning_rate": 1.1469866666666667e-05, | |
| "loss": 1.5202, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 4.311111111111111, | |
| "grad_norm": 2.373842477798462, | |
| "learning_rate": 1.138097777777778e-05, | |
| "loss": 1.5369, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 4.355555555555555, | |
| "grad_norm": 2.1090219020843506, | |
| "learning_rate": 1.1292088888888891e-05, | |
| "loss": 1.541, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 2.7172622680664062, | |
| "learning_rate": 1.1203200000000001e-05, | |
| "loss": 1.5172, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 4.444444444444445, | |
| "grad_norm": 2.557300329208374, | |
| "learning_rate": 1.1114311111111112e-05, | |
| "loss": 1.5325, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 4.488888888888889, | |
| "grad_norm": 2.321272850036621, | |
| "learning_rate": 1.1025422222222224e-05, | |
| "loss": 1.5424, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 4.533333333333333, | |
| "grad_norm": 2.1814475059509277, | |
| "learning_rate": 1.0936533333333334e-05, | |
| "loss": 1.5219, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 4.5777777777777775, | |
| "grad_norm": 2.5891165733337402, | |
| "learning_rate": 1.0847822222222224e-05, | |
| "loss": 1.5294, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 4.622222222222222, | |
| "grad_norm": 2.655404567718506, | |
| "learning_rate": 1.0758933333333335e-05, | |
| "loss": 1.5272, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 4.666666666666667, | |
| "grad_norm": 2.0418145656585693, | |
| "learning_rate": 1.0670044444444445e-05, | |
| "loss": 1.5323, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 4.711111111111111, | |
| "grad_norm": 2.315140962600708, | |
| "learning_rate": 1.0581155555555556e-05, | |
| "loss": 1.5201, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 4.7555555555555555, | |
| "grad_norm": 2.4892210960388184, | |
| "learning_rate": 1.0492266666666668e-05, | |
| "loss": 1.5065, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 3.2457361221313477, | |
| "learning_rate": 1.0403555555555556e-05, | |
| "loss": 1.531, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 4.844444444444444, | |
| "grad_norm": 2.579188346862793, | |
| "learning_rate": 1.0314666666666668e-05, | |
| "loss": 1.5093, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 4.888888888888889, | |
| "grad_norm": 2.9643566608428955, | |
| "learning_rate": 1.0225777777777778e-05, | |
| "loss": 1.5394, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 4.933333333333334, | |
| "grad_norm": 2.824070930480957, | |
| "learning_rate": 1.0136888888888889e-05, | |
| "loss": 1.5352, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 4.977777777777778, | |
| "grad_norm": 2.714317560195923, | |
| "learning_rate": 1.0048e-05, | |
| "loss": 1.5055, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_Sacrebleu": 11.699486858073387, | |
| "eval_loss": 1.7135735750198364, | |
| "eval_runtime": 6965.0506, | |
| "eval_samples_per_second": 1.436, | |
| "eval_steps_per_second": 0.179, | |
| "step": 56250 | |
| }, | |
| { | |
| "epoch": 5.022222222222222, | |
| "grad_norm": 2.3869211673736572, | |
| "learning_rate": 9.959288888888889e-06, | |
| "loss": 1.5236, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 5.066666666666666, | |
| "grad_norm": 2.824490785598755, | |
| "learning_rate": 9.870400000000001e-06, | |
| "loss": 1.4628, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 5.111111111111111, | |
| "grad_norm": 2.378420829772949, | |
| "learning_rate": 9.781511111111112e-06, | |
| "loss": 1.4761, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 5.155555555555556, | |
| "grad_norm": 2.934887647628784, | |
| "learning_rate": 9.692622222222224e-06, | |
| "loss": 1.4744, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 2.5306599140167236, | |
| "learning_rate": 9.603733333333334e-06, | |
| "loss": 1.4719, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 5.2444444444444445, | |
| "grad_norm": 2.27069354057312, | |
| "learning_rate": 9.515022222222224e-06, | |
| "loss": 1.4717, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 5.288888888888889, | |
| "grad_norm": 2.79362154006958, | |
| "learning_rate": 9.426133333333335e-06, | |
| "loss": 1.4671, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 5.333333333333333, | |
| "grad_norm": 2.4642298221588135, | |
| "learning_rate": 9.337244444444445e-06, | |
| "loss": 1.4962, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 5.377777777777778, | |
| "grad_norm": 2.8472516536712646, | |
| "learning_rate": 9.248355555555555e-06, | |
| "loss": 1.4997, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 5.4222222222222225, | |
| "grad_norm": 2.2296738624572754, | |
| "learning_rate": 9.159644444444445e-06, | |
| "loss": 1.4713, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 5.466666666666667, | |
| "grad_norm": 2.8742563724517822, | |
| "learning_rate": 9.070755555555556e-06, | |
| "loss": 1.4629, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 5.511111111111111, | |
| "grad_norm": 2.6657145023345947, | |
| "learning_rate": 8.982044444444445e-06, | |
| "loss": 1.4653, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 5.555555555555555, | |
| "grad_norm": 2.7092838287353516, | |
| "learning_rate": 8.893155555555556e-06, | |
| "loss": 1.4888, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 2.277348041534424, | |
| "learning_rate": 8.804266666666668e-06, | |
| "loss": 1.4786, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 5.644444444444445, | |
| "grad_norm": 1.995211124420166, | |
| "learning_rate": 8.715377777777778e-06, | |
| "loss": 1.4685, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 5.688888888888889, | |
| "grad_norm": 2.569850444793701, | |
| "learning_rate": 8.62648888888889e-06, | |
| "loss": 1.4687, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 5.733333333333333, | |
| "grad_norm": 2.3745999336242676, | |
| "learning_rate": 8.537600000000001e-06, | |
| "loss": 1.4949, | |
| "step": 64500 | |
| }, | |
| { | |
| "epoch": 5.777777777777778, | |
| "grad_norm": 2.3172736167907715, | |
| "learning_rate": 8.448711111111112e-06, | |
| "loss": 1.4807, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 5.822222222222222, | |
| "grad_norm": 2.316258192062378, | |
| "learning_rate": 8.359822222222222e-06, | |
| "loss": 1.4994, | |
| "step": 65500 | |
| }, | |
| { | |
| "epoch": 5.866666666666667, | |
| "grad_norm": 2.303201913833618, | |
| "learning_rate": 8.270933333333334e-06, | |
| "loss": 1.4887, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 5.911111111111111, | |
| "grad_norm": 2.1658709049224854, | |
| "learning_rate": 8.182222222222222e-06, | |
| "loss": 1.4751, | |
| "step": 66500 | |
| }, | |
| { | |
| "epoch": 5.955555555555556, | |
| "grad_norm": 2.177354574203491, | |
| "learning_rate": 8.093333333333334e-06, | |
| "loss": 1.4751, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 2.559067726135254, | |
| "learning_rate": 8.004444444444445e-06, | |
| "loss": 1.4748, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_Sacrebleu": 11.844348456691092, | |
| "eval_loss": 1.7136952877044678, | |
| "eval_runtime": 4383.3868, | |
| "eval_samples_per_second": 2.281, | |
| "eval_steps_per_second": 0.285, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 6.044444444444444, | |
| "grad_norm": 2.618431329727173, | |
| "learning_rate": 7.915555555555557e-06, | |
| "loss": 1.4243, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 6.088888888888889, | |
| "grad_norm": 2.2536802291870117, | |
| "learning_rate": 7.826666666666667e-06, | |
| "loss": 1.4415, | |
| "step": 68500 | |
| }, | |
| { | |
| "epoch": 6.133333333333334, | |
| "grad_norm": 2.866265058517456, | |
| "learning_rate": 7.737777777777778e-06, | |
| "loss": 1.429, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 6.177777777777778, | |
| "grad_norm": 2.799807071685791, | |
| "learning_rate": 7.648888888888888e-06, | |
| "loss": 1.4456, | |
| "step": 69500 | |
| }, | |
| { | |
| "epoch": 6.222222222222222, | |
| "grad_norm": 2.6799798011779785, | |
| "learning_rate": 7.5600000000000005e-06, | |
| "loss": 1.4302, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 6.266666666666667, | |
| "grad_norm": 2.8705008029937744, | |
| "learning_rate": 7.471288888888889e-06, | |
| "loss": 1.4334, | |
| "step": 70500 | |
| }, | |
| { | |
| "epoch": 6.311111111111111, | |
| "grad_norm": 2.6502320766448975, | |
| "learning_rate": 7.382400000000001e-06, | |
| "loss": 1.4338, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 6.355555555555555, | |
| "grad_norm": 2.9277572631835938, | |
| "learning_rate": 7.293511111111111e-06, | |
| "loss": 1.4364, | |
| "step": 71500 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 2.3213584423065186, | |
| "learning_rate": 7.2046222222222224e-06, | |
| "loss": 1.4258, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 6.444444444444445, | |
| "grad_norm": 3.05753231048584, | |
| "learning_rate": 7.115911111111111e-06, | |
| "loss": 1.4159, | |
| "step": 72500 | |
| }, | |
| { | |
| "epoch": 6.488888888888889, | |
| "grad_norm": 2.879302978515625, | |
| "learning_rate": 7.027022222222223e-06, | |
| "loss": 1.4575, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 6.533333333333333, | |
| "grad_norm": 2.283719539642334, | |
| "learning_rate": 6.938133333333333e-06, | |
| "loss": 1.4355, | |
| "step": 73500 | |
| }, | |
| { | |
| "epoch": 6.5777777777777775, | |
| "grad_norm": 2.5252952575683594, | |
| "learning_rate": 6.849422222222223e-06, | |
| "loss": 1.4346, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 6.622222222222222, | |
| "grad_norm": 2.6116600036621094, | |
| "learning_rate": 6.760533333333333e-06, | |
| "loss": 1.4646, | |
| "step": 74500 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 2.494943380355835, | |
| "learning_rate": 6.671644444444445e-06, | |
| "loss": 1.4385, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 6.711111111111111, | |
| "grad_norm": 2.3194327354431152, | |
| "learning_rate": 6.582755555555556e-06, | |
| "loss": 1.4508, | |
| "step": 75500 | |
| }, | |
| { | |
| "epoch": 6.7555555555555555, | |
| "grad_norm": 2.8426125049591064, | |
| "learning_rate": 6.494044444444445e-06, | |
| "loss": 1.4412, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "grad_norm": 3.0324361324310303, | |
| "learning_rate": 6.405155555555555e-06, | |
| "loss": 1.423, | |
| "step": 76500 | |
| }, | |
| { | |
| "epoch": 6.844444444444444, | |
| "grad_norm": 2.176151990890503, | |
| "learning_rate": 6.3162666666666674e-06, | |
| "loss": 1.4474, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 6.888888888888889, | |
| "grad_norm": 2.6893413066864014, | |
| "learning_rate": 6.227377777777778e-06, | |
| "loss": 1.4415, | |
| "step": 77500 | |
| }, | |
| { | |
| "epoch": 6.933333333333334, | |
| "grad_norm": 2.657773017883301, | |
| "learning_rate": 6.138488888888889e-06, | |
| "loss": 1.4257, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 6.977777777777778, | |
| "grad_norm": 2.4668519496917725, | |
| "learning_rate": 6.049600000000001e-06, | |
| "loss": 1.424, | |
| "step": 78500 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_Sacrebleu": 11.861485151142872, | |
| "eval_loss": 1.7204111814498901, | |
| "eval_runtime": 4008.7534, | |
| "eval_samples_per_second": 2.495, | |
| "eval_steps_per_second": 0.312, | |
| "step": 78750 | |
| }, | |
| { | |
| "epoch": 7.022222222222222, | |
| "grad_norm": 3.0832223892211914, | |
| "learning_rate": 5.960711111111112e-06, | |
| "loss": 1.4192, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 7.066666666666666, | |
| "grad_norm": 2.553757667541504, | |
| "learning_rate": 5.871822222222223e-06, | |
| "loss": 1.4091, | |
| "step": 79500 | |
| }, | |
| { | |
| "epoch": 7.111111111111111, | |
| "grad_norm": 2.9473392963409424, | |
| "learning_rate": 5.7829333333333336e-06, | |
| "loss": 1.4289, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 7.155555555555556, | |
| "grad_norm": 3.1997127532958984, | |
| "learning_rate": 5.694044444444446e-06, | |
| "loss": 1.3862, | |
| "step": 80500 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 3.077010154724121, | |
| "learning_rate": 5.605155555555556e-06, | |
| "loss": 1.4042, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 7.2444444444444445, | |
| "grad_norm": 3.0450375080108643, | |
| "learning_rate": 5.5162666666666675e-06, | |
| "loss": 1.4109, | |
| "step": 81500 | |
| }, | |
| { | |
| "epoch": 7.288888888888889, | |
| "grad_norm": 2.3486075401306152, | |
| "learning_rate": 5.4275555555555555e-06, | |
| "loss": 1.3827, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 7.333333333333333, | |
| "grad_norm": 1.9607141017913818, | |
| "learning_rate": 5.338666666666668e-06, | |
| "loss": 1.4181, | |
| "step": 82500 | |
| }, | |
| { | |
| "epoch": 7.377777777777778, | |
| "grad_norm": 2.3723082542419434, | |
| "learning_rate": 5.249777777777778e-06, | |
| "loss": 1.386, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 7.4222222222222225, | |
| "grad_norm": 3.01910138130188, | |
| "learning_rate": 5.1608888888888894e-06, | |
| "loss": 1.4034, | |
| "step": 83500 | |
| }, | |
| { | |
| "epoch": 7.466666666666667, | |
| "grad_norm": 3.99428129196167, | |
| "learning_rate": 5.072e-06, | |
| "loss": 1.4088, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 7.511111111111111, | |
| "grad_norm": 2.6704437732696533, | |
| "learning_rate": 4.983111111111111e-06, | |
| "loss": 1.4215, | |
| "step": 84500 | |
| }, | |
| { | |
| "epoch": 7.555555555555555, | |
| "grad_norm": 2.2187724113464355, | |
| "learning_rate": 4.8944e-06, | |
| "loss": 1.4019, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 2.6798274517059326, | |
| "learning_rate": 4.805511111111111e-06, | |
| "loss": 1.4033, | |
| "step": 85500 | |
| }, | |
| { | |
| "epoch": 7.644444444444445, | |
| "grad_norm": 2.2254135608673096, | |
| "learning_rate": 4.716622222222223e-06, | |
| "loss": 1.3873, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 7.688888888888889, | |
| "grad_norm": 3.0042669773101807, | |
| "learning_rate": 4.627733333333333e-06, | |
| "loss": 1.3814, | |
| "step": 86500 | |
| }, | |
| { | |
| "epoch": 7.733333333333333, | |
| "grad_norm": 2.9878363609313965, | |
| "learning_rate": 4.539022222222222e-06, | |
| "loss": 1.4171, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 7.777777777777778, | |
| "grad_norm": 2.8947293758392334, | |
| "learning_rate": 4.450133333333333e-06, | |
| "loss": 1.3992, | |
| "step": 87500 | |
| }, | |
| { | |
| "epoch": 7.822222222222222, | |
| "grad_norm": 2.1072092056274414, | |
| "learning_rate": 4.361422222222222e-06, | |
| "loss": 1.4128, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 7.866666666666667, | |
| "grad_norm": 2.7761893272399902, | |
| "learning_rate": 4.272533333333334e-06, | |
| "loss": 1.4229, | |
| "step": 88500 | |
| }, | |
| { | |
| "epoch": 7.911111111111111, | |
| "grad_norm": 2.724802255630493, | |
| "learning_rate": 4.183644444444445e-06, | |
| "loss": 1.4147, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 7.955555555555556, | |
| "grad_norm": 2.1179940700531006, | |
| "learning_rate": 4.094755555555555e-06, | |
| "loss": 1.4154, | |
| "step": 89500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.6914615631103516, | |
| "learning_rate": 4.005866666666667e-06, | |
| "loss": 1.4058, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_Sacrebleu": 11.669051917717454, | |
| "eval_loss": 1.7239112854003906, | |
| "eval_runtime": 3810.4584, | |
| "eval_samples_per_second": 2.624, | |
| "eval_steps_per_second": 0.328, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 8.044444444444444, | |
| "grad_norm": 2.054410457611084, | |
| "learning_rate": 3.916977777777778e-06, | |
| "loss": 1.3687, | |
| "step": 90500 | |
| }, | |
| { | |
| "epoch": 8.088888888888889, | |
| "grad_norm": 3.7907662391662598, | |
| "learning_rate": 3.828088888888889e-06, | |
| "loss": 1.3839, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 8.133333333333333, | |
| "grad_norm": 2.704850435256958, | |
| "learning_rate": 3.7392e-06, | |
| "loss": 1.3677, | |
| "step": 91500 | |
| }, | |
| { | |
| "epoch": 8.177777777777777, | |
| "grad_norm": 2.888054132461548, | |
| "learning_rate": 3.650311111111111e-06, | |
| "loss": 1.382, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 8.222222222222221, | |
| "grad_norm": 3.2732253074645996, | |
| "learning_rate": 3.5616e-06, | |
| "loss": 1.4076, | |
| "step": 92500 | |
| }, | |
| { | |
| "epoch": 8.266666666666667, | |
| "grad_norm": 2.5939693450927734, | |
| "learning_rate": 3.4727111111111112e-06, | |
| "loss": 1.376, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 8.311111111111112, | |
| "grad_norm": 3.650665760040283, | |
| "learning_rate": 3.383822222222222e-06, | |
| "loss": 1.3869, | |
| "step": 93500 | |
| }, | |
| { | |
| "epoch": 8.355555555555556, | |
| "grad_norm": 2.4687626361846924, | |
| "learning_rate": 3.294933333333334e-06, | |
| "loss": 1.3643, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "grad_norm": 2.497544765472412, | |
| "learning_rate": 3.206044444444445e-06, | |
| "loss": 1.3783, | |
| "step": 94500 | |
| }, | |
| { | |
| "epoch": 8.444444444444445, | |
| "grad_norm": 2.620575189590454, | |
| "learning_rate": 3.117333333333333e-06, | |
| "loss": 1.3929, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 8.488888888888889, | |
| "grad_norm": 1.9510403871536255, | |
| "learning_rate": 3.028444444444445e-06, | |
| "loss": 1.3757, | |
| "step": 95500 | |
| }, | |
| { | |
| "epoch": 8.533333333333333, | |
| "grad_norm": 2.253645896911621, | |
| "learning_rate": 2.9395555555555562e-06, | |
| "loss": 1.3843, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 8.577777777777778, | |
| "grad_norm": 2.5692851543426514, | |
| "learning_rate": 2.850666666666667e-06, | |
| "loss": 1.3617, | |
| "step": 96500 | |
| }, | |
| { | |
| "epoch": 8.622222222222222, | |
| "grad_norm": 2.7318949699401855, | |
| "learning_rate": 2.761955555555556e-06, | |
| "loss": 1.3965, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 8.666666666666666, | |
| "grad_norm": 2.397948741912842, | |
| "learning_rate": 2.6730666666666673e-06, | |
| "loss": 1.3901, | |
| "step": 97500 | |
| }, | |
| { | |
| "epoch": 8.71111111111111, | |
| "grad_norm": 2.227858781814575, | |
| "learning_rate": 2.584177777777778e-06, | |
| "loss": 1.3631, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 8.755555555555556, | |
| "grad_norm": 2.7916066646575928, | |
| "learning_rate": 2.495288888888889e-06, | |
| "loss": 1.3957, | |
| "step": 98500 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "grad_norm": 3.0856597423553467, | |
| "learning_rate": 2.4064e-06, | |
| "loss": 1.3941, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 8.844444444444445, | |
| "grad_norm": 2.253394603729248, | |
| "learning_rate": 2.317688888888889e-06, | |
| "loss": 1.3731, | |
| "step": 99500 | |
| }, | |
| { | |
| "epoch": 8.88888888888889, | |
| "grad_norm": 2.375483274459839, | |
| "learning_rate": 2.2288e-06, | |
| "loss": 1.4075, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 8.933333333333334, | |
| "grad_norm": 2.5208873748779297, | |
| "learning_rate": 2.139911111111111e-06, | |
| "loss": 1.36, | |
| "step": 100500 | |
| }, | |
| { | |
| "epoch": 8.977777777777778, | |
| "grad_norm": 2.9750773906707764, | |
| "learning_rate": 2.0510222222222223e-06, | |
| "loss": 1.3803, | |
| "step": 101000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_Sacrebleu": 11.812893236671785, | |
| "eval_loss": 1.7298730611801147, | |
| "eval_runtime": 3842.5674, | |
| "eval_samples_per_second": 2.602, | |
| "eval_steps_per_second": 0.325, | |
| "step": 101250 | |
| }, | |
| { | |
| "epoch": 9.022222222222222, | |
| "grad_norm": 2.893308162689209, | |
| "learning_rate": 1.9621333333333332e-06, | |
| "loss": 1.3477, | |
| "step": 101500 | |
| }, | |
| { | |
| "epoch": 9.066666666666666, | |
| "grad_norm": 2.4525198936462402, | |
| "learning_rate": 1.8732444444444445e-06, | |
| "loss": 1.3894, | |
| "step": 102000 | |
| }, | |
| { | |
| "epoch": 9.11111111111111, | |
| "grad_norm": 2.138702392578125, | |
| "learning_rate": 1.7843555555555556e-06, | |
| "loss": 1.3871, | |
| "step": 102500 | |
| }, | |
| { | |
| "epoch": 9.155555555555555, | |
| "grad_norm": 2.6823806762695312, | |
| "learning_rate": 1.6956444444444445e-06, | |
| "loss": 1.3678, | |
| "step": 103000 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "grad_norm": 3.6477270126342773, | |
| "learning_rate": 1.6067555555555556e-06, | |
| "loss": 1.3709, | |
| "step": 103500 | |
| }, | |
| { | |
| "epoch": 9.244444444444444, | |
| "grad_norm": 2.2082371711730957, | |
| "learning_rate": 1.5182222222222223e-06, | |
| "loss": 1.376, | |
| "step": 104000 | |
| }, | |
| { | |
| "epoch": 9.28888888888889, | |
| "grad_norm": 2.7231674194335938, | |
| "learning_rate": 1.4293333333333334e-06, | |
| "loss": 1.3372, | |
| "step": 104500 | |
| }, | |
| { | |
| "epoch": 9.333333333333334, | |
| "grad_norm": 2.383312940597534, | |
| "learning_rate": 1.3404444444444445e-06, | |
| "loss": 1.3507, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 9.377777777777778, | |
| "grad_norm": 2.908069610595703, | |
| "learning_rate": 1.2515555555555556e-06, | |
| "loss": 1.371, | |
| "step": 105500 | |
| }, | |
| { | |
| "epoch": 9.422222222222222, | |
| "grad_norm": 2.7813098430633545, | |
| "learning_rate": 1.1626666666666667e-06, | |
| "loss": 1.355, | |
| "step": 106000 | |
| }, | |
| { | |
| "epoch": 9.466666666666667, | |
| "grad_norm": 2.4911601543426514, | |
| "learning_rate": 1.0737777777777778e-06, | |
| "loss": 1.3662, | |
| "step": 106500 | |
| }, | |
| { | |
| "epoch": 9.511111111111111, | |
| "grad_norm": 2.429072856903076, | |
| "learning_rate": 9.85066666666667e-07, | |
| "loss": 1.3643, | |
| "step": 107000 | |
| }, | |
| { | |
| "epoch": 9.555555555555555, | |
| "grad_norm": 2.832702875137329, | |
| "learning_rate": 8.961777777777779e-07, | |
| "loss": 1.3664, | |
| "step": 107500 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 2.4679746627807617, | |
| "learning_rate": 8.07288888888889e-07, | |
| "loss": 1.3599, | |
| "step": 108000 | |
| }, | |
| { | |
| "epoch": 9.644444444444444, | |
| "grad_norm": 2.522490978240967, | |
| "learning_rate": 7.184000000000001e-07, | |
| "loss": 1.3691, | |
| "step": 108500 | |
| }, | |
| { | |
| "epoch": 9.688888888888888, | |
| "grad_norm": 2.7089273929595947, | |
| "learning_rate": 6.295111111111112e-07, | |
| "loss": 1.373, | |
| "step": 109000 | |
| }, | |
| { | |
| "epoch": 9.733333333333333, | |
| "grad_norm": 2.0823116302490234, | |
| "learning_rate": 5.406222222222223e-07, | |
| "loss": 1.369, | |
| "step": 109500 | |
| }, | |
| { | |
| "epoch": 9.777777777777779, | |
| "grad_norm": 1.9741628170013428, | |
| "learning_rate": 4.517333333333334e-07, | |
| "loss": 1.3559, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 9.822222222222223, | |
| "grad_norm": 2.4756247997283936, | |
| "learning_rate": 3.628444444444445e-07, | |
| "loss": 1.379, | |
| "step": 110500 | |
| }, | |
| { | |
| "epoch": 9.866666666666667, | |
| "grad_norm": 2.6208088397979736, | |
| "learning_rate": 2.739555555555556e-07, | |
| "loss": 1.3585, | |
| "step": 111000 | |
| }, | |
| { | |
| "epoch": 9.911111111111111, | |
| "grad_norm": 3.2721610069274902, | |
| "learning_rate": 1.8506666666666668e-07, | |
| "loss": 1.3772, | |
| "step": 111500 | |
| }, | |
| { | |
| "epoch": 9.955555555555556, | |
| "grad_norm": 2.5334367752075195, | |
| "learning_rate": 9.617777777777777e-08, | |
| "loss": 1.3576, | |
| "step": 112000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 2.59663987159729, | |
| "learning_rate": 7.288888888888889e-09, | |
| "loss": 1.3678, | |
| "step": 112500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 112500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "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": 2.437992677376e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |