| { | |
| "best_global_step": 1260, | |
| "best_metric": 0.2602008879184723, | |
| "best_model_checkpoint": "saves_multiple/lntuning/llama-3-8b-instruct/train_copa_789_1760637876/checkpoint-1260", | |
| "epoch": 20.0, | |
| "eval_steps": 90, | |
| "global_step": 1800, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05555555555555555, | |
| "grad_norm": 12.625, | |
| "learning_rate": 1.1111111111111112e-06, | |
| "loss": 0.5563, | |
| "num_input_tokens_seen": 1632, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 11.75, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.6401, | |
| "num_input_tokens_seen": 3232, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.16666666666666666, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 3.888888888888889e-06, | |
| "loss": 0.669, | |
| "num_input_tokens_seen": 4832, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 11.625, | |
| "learning_rate": 5.277777777777778e-06, | |
| "loss": 0.7562, | |
| "num_input_tokens_seen": 6432, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2777777777777778, | |
| "grad_norm": 11.875, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.7057, | |
| "num_input_tokens_seen": 7968, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 11.75, | |
| "learning_rate": 8.055555555555557e-06, | |
| "loss": 0.5329, | |
| "num_input_tokens_seen": 9504, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3888888888888889, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 9.444444444444445e-06, | |
| "loss": 0.7157, | |
| "num_input_tokens_seen": 11104, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.0833333333333334e-05, | |
| "loss": 0.8525, | |
| "num_input_tokens_seen": 12704, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.2222222222222222e-05, | |
| "loss": 0.6187, | |
| "num_input_tokens_seen": 14240, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 14.375, | |
| "learning_rate": 1.3611111111111111e-05, | |
| "loss": 0.6171, | |
| "num_input_tokens_seen": 15808, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6111111111111112, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.7308, | |
| "num_input_tokens_seen": 17344, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 1.638888888888889e-05, | |
| "loss": 0.5338, | |
| "num_input_tokens_seen": 18912, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7222222222222222, | |
| "grad_norm": 12.9375, | |
| "learning_rate": 1.777777777777778e-05, | |
| "loss": 0.6748, | |
| "num_input_tokens_seen": 20448, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 8.5, | |
| "learning_rate": 1.9166666666666667e-05, | |
| "loss": 0.5461, | |
| "num_input_tokens_seen": 21984, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 9.25, | |
| "learning_rate": 2.0555555555555555e-05, | |
| "loss": 0.687, | |
| "num_input_tokens_seen": 23552, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 2.1944444444444445e-05, | |
| "loss": 0.6023, | |
| "num_input_tokens_seen": 25120, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.9444444444444444, | |
| "grad_norm": 11.0, | |
| "learning_rate": 2.3333333333333336e-05, | |
| "loss": 0.5416, | |
| "num_input_tokens_seen": 26656, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 17.125, | |
| "learning_rate": 2.4722222222222223e-05, | |
| "loss": 0.6809, | |
| "num_input_tokens_seen": 28192, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.5333930850028992, | |
| "eval_runtime": 0.4854, | |
| "eval_samples_per_second": 82.398, | |
| "eval_steps_per_second": 20.6, | |
| "num_input_tokens_seen": 28192, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.0555555555555556, | |
| "grad_norm": 11.25, | |
| "learning_rate": 2.6111111111111114e-05, | |
| "loss": 0.5289, | |
| "num_input_tokens_seen": 29792, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.5997, | |
| "num_input_tokens_seen": 31328, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.1666666666666667, | |
| "grad_norm": 13.4375, | |
| "learning_rate": 2.8888888888888888e-05, | |
| "loss": 0.5872, | |
| "num_input_tokens_seen": 32832, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.2222222222222223, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 3.0277777777777776e-05, | |
| "loss": 0.5543, | |
| "num_input_tokens_seen": 34304, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.2777777777777777, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 3.1666666666666666e-05, | |
| "loss": 0.7101, | |
| "num_input_tokens_seen": 35840, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 3.3055555555555553e-05, | |
| "loss": 0.8676, | |
| "num_input_tokens_seen": 37376, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.3888888888888888, | |
| "grad_norm": 8.75, | |
| "learning_rate": 3.444444444444445e-05, | |
| "loss": 0.5599, | |
| "num_input_tokens_seen": 38944, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.4444444444444444, | |
| "grad_norm": 11.375, | |
| "learning_rate": 3.5833333333333335e-05, | |
| "loss": 0.7014, | |
| "num_input_tokens_seen": 40512, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 10.25, | |
| "learning_rate": 3.722222222222222e-05, | |
| "loss": 0.4826, | |
| "num_input_tokens_seen": 42080, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.5555555555555556, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 3.8611111111111116e-05, | |
| "loss": 0.7206, | |
| "num_input_tokens_seen": 43680, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.6111111111111112, | |
| "grad_norm": 9.5, | |
| "learning_rate": 4e-05, | |
| "loss": 0.5154, | |
| "num_input_tokens_seen": 45216, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 4.138888888888889e-05, | |
| "loss": 0.8157, | |
| "num_input_tokens_seen": 46720, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.7222222222222223, | |
| "grad_norm": 9.0, | |
| "learning_rate": 4.277777777777778e-05, | |
| "loss": 0.4187, | |
| "num_input_tokens_seen": 48288, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 13.9375, | |
| "learning_rate": 4.4166666666666665e-05, | |
| "loss": 0.5965, | |
| "num_input_tokens_seen": 49888, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.8333333333333335, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 4.555555555555556e-05, | |
| "loss": 0.4704, | |
| "num_input_tokens_seen": 51424, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.8888888888888888, | |
| "grad_norm": 12.25, | |
| "learning_rate": 4.6944444444444446e-05, | |
| "loss": 0.7356, | |
| "num_input_tokens_seen": 52992, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.9444444444444444, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 4.8333333333333334e-05, | |
| "loss": 0.5851, | |
| "num_input_tokens_seen": 54592, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 13.0625, | |
| "learning_rate": 4.972222222222223e-05, | |
| "loss": 0.6512, | |
| "num_input_tokens_seen": 56192, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.46258974075317383, | |
| "eval_runtime": 0.4905, | |
| "eval_samples_per_second": 81.548, | |
| "eval_steps_per_second": 20.387, | |
| "num_input_tokens_seen": 56192, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.0555555555555554, | |
| "grad_norm": 10.0, | |
| "learning_rate": 4.9999247861994194e-05, | |
| "loss": 0.5343, | |
| "num_input_tokens_seen": 57760, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.111111111111111, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 4.9996192378909786e-05, | |
| "loss": 0.5625, | |
| "num_input_tokens_seen": 59264, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.1666666666666665, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 4.999078682916774e-05, | |
| "loss": 0.5095, | |
| "num_input_tokens_seen": 60768, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 16.0, | |
| "learning_rate": 4.998303172098155e-05, | |
| "loss": 0.5655, | |
| "num_input_tokens_seen": 62272, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.2777777777777777, | |
| "grad_norm": 8.0, | |
| "learning_rate": 4.997292778346312e-05, | |
| "loss": 0.5422, | |
| "num_input_tokens_seen": 63808, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.3333333333333335, | |
| "grad_norm": 9.25, | |
| "learning_rate": 4.996047596655418e-05, | |
| "loss": 0.6897, | |
| "num_input_tokens_seen": 65344, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.388888888888889, | |
| "grad_norm": 8.375, | |
| "learning_rate": 4.994567744093703e-05, | |
| "loss": 0.465, | |
| "num_input_tokens_seen": 66912, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.4444444444444446, | |
| "grad_norm": 10.25, | |
| "learning_rate": 4.992853359792444e-05, | |
| "loss": 0.7158, | |
| "num_input_tokens_seen": 68480, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 4.9909046049328846e-05, | |
| "loss": 0.4735, | |
| "num_input_tokens_seen": 70048, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.5555555555555554, | |
| "grad_norm": 13.0625, | |
| "learning_rate": 4.988721662731083e-05, | |
| "loss": 0.5161, | |
| "num_input_tokens_seen": 71680, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.611111111111111, | |
| "grad_norm": 12.4375, | |
| "learning_rate": 4.9863047384206835e-05, | |
| "loss": 0.8189, | |
| "num_input_tokens_seen": 73280, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 10.375, | |
| "learning_rate": 4.983654059233626e-05, | |
| "loss": 0.5923, | |
| "num_input_tokens_seen": 74752, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.7222222222222223, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 4.9807698743787744e-05, | |
| "loss": 0.5109, | |
| "num_input_tokens_seen": 76288, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.7777777777777777, | |
| "grad_norm": 8.625, | |
| "learning_rate": 4.9776524550184965e-05, | |
| "loss": 0.5233, | |
| "num_input_tokens_seen": 77888, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.8333333333333335, | |
| "grad_norm": 7.0, | |
| "learning_rate": 4.974302094243164e-05, | |
| "loss": 0.3302, | |
| "num_input_tokens_seen": 79456, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.888888888888889, | |
| "grad_norm": 10.25, | |
| "learning_rate": 4.970719107043595e-05, | |
| "loss": 0.5286, | |
| "num_input_tokens_seen": 81088, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.9444444444444446, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 4.966903830281449e-05, | |
| "loss": 0.4274, | |
| "num_input_tokens_seen": 82592, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 4.962856622657541e-05, | |
| "loss": 0.608, | |
| "num_input_tokens_seen": 84192, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.4239669740200043, | |
| "eval_runtime": 0.4915, | |
| "eval_samples_per_second": 81.382, | |
| "eval_steps_per_second": 20.345, | |
| "num_input_tokens_seen": 84192, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.0555555555555554, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 4.9585778646781364e-05, | |
| "loss": 0.4894, | |
| "num_input_tokens_seen": 85696, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.111111111111111, | |
| "grad_norm": 11.125, | |
| "learning_rate": 4.9540679586191605e-05, | |
| "loss": 0.4343, | |
| "num_input_tokens_seen": 87296, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.1666666666666665, | |
| "grad_norm": 9.625, | |
| "learning_rate": 4.9493273284883854e-05, | |
| "loss": 0.4369, | |
| "num_input_tokens_seen": 88832, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.2222222222222223, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 4.9443564199855666e-05, | |
| "loss": 0.4493, | |
| "num_input_tokens_seen": 90368, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.2777777777777777, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 4.939155700460536e-05, | |
| "loss": 0.419, | |
| "num_input_tokens_seen": 91968, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 9.5, | |
| "learning_rate": 4.933725658869267e-05, | |
| "loss": 0.4637, | |
| "num_input_tokens_seen": 93568, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.388888888888889, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 4.9280668057279014e-05, | |
| "loss": 0.4865, | |
| "num_input_tokens_seen": 95168, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.4444444444444446, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 4.9221796730647516e-05, | |
| "loss": 0.4821, | |
| "num_input_tokens_seen": 96704, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 4.916064814370287e-05, | |
| "loss": 0.5254, | |
| "num_input_tokens_seen": 98304, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 6.375, | |
| "learning_rate": 4.9097228045450864e-05, | |
| "loss": 0.3374, | |
| "num_input_tokens_seen": 99840, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.611111111111111, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 4.9031542398457974e-05, | |
| "loss": 0.4323, | |
| "num_input_tokens_seen": 101376, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 3.6666666666666665, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 4.896359737829071e-05, | |
| "loss": 0.549, | |
| "num_input_tokens_seen": 102976, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 3.7222222222222223, | |
| "grad_norm": 10.5, | |
| "learning_rate": 4.889339937293508e-05, | |
| "loss": 0.5552, | |
| "num_input_tokens_seen": 104576, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 3.7777777777777777, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 4.8820954982195905e-05, | |
| "loss": 0.6452, | |
| "num_input_tokens_seen": 106112, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.8333333333333335, | |
| "grad_norm": 7.125, | |
| "learning_rate": 4.874627101707644e-05, | |
| "loss": 0.4973, | |
| "num_input_tokens_seen": 107712, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 3.888888888888889, | |
| "grad_norm": 6.875, | |
| "learning_rate": 4.8669354499137955e-05, | |
| "loss": 0.5412, | |
| "num_input_tokens_seen": 109344, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.9444444444444446, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 4.859021265983959e-05, | |
| "loss": 0.4439, | |
| "num_input_tokens_seen": 110944, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 8.0, | |
| "learning_rate": 4.850885293985853e-05, | |
| "loss": 0.5009, | |
| "num_input_tokens_seen": 112544, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.3563203811645508, | |
| "eval_runtime": 0.4941, | |
| "eval_samples_per_second": 80.961, | |
| "eval_steps_per_second": 20.24, | |
| "num_input_tokens_seen": 112544, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.055555555555555, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 4.8425282988390376e-05, | |
| "loss": 0.5249, | |
| "num_input_tokens_seen": 114144, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.111111111111111, | |
| "grad_norm": 11.25, | |
| "learning_rate": 4.8339510662430046e-05, | |
| "loss": 0.4523, | |
| "num_input_tokens_seen": 115712, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.166666666666667, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 4.825154402603308e-05, | |
| "loss": 0.6025, | |
| "num_input_tokens_seen": 117312, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.222222222222222, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 4.816139134955746e-05, | |
| "loss": 0.5733, | |
| "num_input_tokens_seen": 118848, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.277777777777778, | |
| "grad_norm": 8.25, | |
| "learning_rate": 4.806906110888606e-05, | |
| "loss": 0.4163, | |
| "num_input_tokens_seen": 120416, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.333333333333333, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 4.797456198462979e-05, | |
| "loss": 0.4261, | |
| "num_input_tokens_seen": 122016, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.388888888888889, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 4.7877902861311446e-05, | |
| "loss": 0.58, | |
| "num_input_tokens_seen": 123584, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 4.444444444444445, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 4.777909282653042e-05, | |
| "loss": 0.2823, | |
| "num_input_tokens_seen": 125216, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 4.7678141170108345e-05, | |
| "loss": 0.3767, | |
| "num_input_tokens_seen": 126784, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 4.555555555555555, | |
| "grad_norm": 3.703125, | |
| "learning_rate": 4.757505738321563e-05, | |
| "loss": 0.319, | |
| "num_input_tokens_seen": 128352, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 4.611111111111111, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 4.7469851157479177e-05, | |
| "loss": 0.242, | |
| "num_input_tokens_seen": 129920, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 4.666666666666667, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 4.736253238407119e-05, | |
| "loss": 0.3992, | |
| "num_input_tokens_seen": 131488, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 4.722222222222222, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 4.725311115277924e-05, | |
| "loss": 0.4625, | |
| "num_input_tokens_seen": 133056, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 4.777777777777778, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 4.714159775105765e-05, | |
| "loss": 0.3862, | |
| "num_input_tokens_seen": 134624, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 4.833333333333333, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 4.70280026630603e-05, | |
| "loss": 0.4036, | |
| "num_input_tokens_seen": 136192, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 4.888888888888889, | |
| "grad_norm": 6.5, | |
| "learning_rate": 4.6912336568654925e-05, | |
| "loss": 0.3366, | |
| "num_input_tokens_seen": 137824, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 4.944444444444445, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 4.679461034241906e-05, | |
| "loss": 0.4257, | |
| "num_input_tokens_seen": 139392, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 4.667483505261762e-05, | |
| "loss": 0.5363, | |
| "num_input_tokens_seen": 140960, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.3248513340950012, | |
| "eval_runtime": 0.4865, | |
| "eval_samples_per_second": 82.214, | |
| "eval_steps_per_second": 20.554, | |
| "num_input_tokens_seen": 140960, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 5.055555555555555, | |
| "grad_norm": 8.5, | |
| "learning_rate": 4.655302196016228e-05, | |
| "loss": 0.4614, | |
| "num_input_tokens_seen": 142560, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 5.111111111111111, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 4.642918251755281e-05, | |
| "loss": 0.4382, | |
| "num_input_tokens_seen": 144096, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 5.166666666666667, | |
| "grad_norm": 5.75, | |
| "learning_rate": 4.6303328367800284e-05, | |
| "loss": 0.2871, | |
| "num_input_tokens_seen": 145632, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 5.222222222222222, | |
| "grad_norm": 9.0, | |
| "learning_rate": 4.6175471343332485e-05, | |
| "loss": 0.5127, | |
| "num_input_tokens_seen": 147232, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 5.277777777777778, | |
| "grad_norm": 5.75, | |
| "learning_rate": 4.604562346488144e-05, | |
| "loss": 0.455, | |
| "num_input_tokens_seen": 148864, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 5.333333333333333, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 4.591379694035325e-05, | |
| "loss": 0.3908, | |
| "num_input_tokens_seen": 150432, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 5.388888888888889, | |
| "grad_norm": 8.5, | |
| "learning_rate": 4.5780004163680365e-05, | |
| "loss": 0.4341, | |
| "num_input_tokens_seen": 152032, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 5.444444444444445, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 4.5644257713656356e-05, | |
| "loss": 0.385, | |
| "num_input_tokens_seen": 153600, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 4.550657035275323e-05, | |
| "loss": 0.3744, | |
| "num_input_tokens_seen": 155200, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 5.555555555555555, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 4.536695502592162e-05, | |
| "loss": 0.4971, | |
| "num_input_tokens_seen": 156800, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 5.611111111111111, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 4.522542485937369e-05, | |
| "loss": 0.3377, | |
| "num_input_tokens_seen": 158336, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 5.666666666666667, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 4.5081993159349056e-05, | |
| "loss": 0.2584, | |
| "num_input_tokens_seen": 159904, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 5.722222222222222, | |
| "grad_norm": 15.8125, | |
| "learning_rate": 4.493667341086379e-05, | |
| "loss": 0.4332, | |
| "num_input_tokens_seen": 161504, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 5.777777777777778, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 4.478947927644258e-05, | |
| "loss": 0.3261, | |
| "num_input_tokens_seen": 163008, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 5.833333333333333, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 4.464042459483425e-05, | |
| "loss": 0.3614, | |
| "num_input_tokens_seen": 164512, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 5.888888888888889, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 4.448952337971064e-05, | |
| "loss": 0.6182, | |
| "num_input_tokens_seen": 166080, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 5.944444444444445, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 4.43367898183491e-05, | |
| "loss": 0.3672, | |
| "num_input_tokens_seen": 167680, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 7.0, | |
| "learning_rate": 4.418223827029867e-05, | |
| "loss": 0.2618, | |
| "num_input_tokens_seen": 169216, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.29478442668914795, | |
| "eval_runtime": 0.4948, | |
| "eval_samples_per_second": 80.845, | |
| "eval_steps_per_second": 20.211, | |
| "num_input_tokens_seen": 169216, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 6.055555555555555, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 4.402588326603002e-05, | |
| "loss": 0.2542, | |
| "num_input_tokens_seen": 170784, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 6.111111111111111, | |
| "grad_norm": 10.5, | |
| "learning_rate": 4.386773950556931e-05, | |
| "loss": 0.4418, | |
| "num_input_tokens_seen": 172352, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 6.166666666666667, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 4.3707821857116176e-05, | |
| "loss": 0.414, | |
| "num_input_tokens_seen": 173920, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 6.222222222222222, | |
| "grad_norm": 21.25, | |
| "learning_rate": 4.354614535564588e-05, | |
| "loss": 0.3713, | |
| "num_input_tokens_seen": 175488, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 6.277777777777778, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 4.3382725201495723e-05, | |
| "loss": 0.3619, | |
| "num_input_tokens_seen": 177024, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 6.333333333333333, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 4.321757675893596e-05, | |
| "loss": 0.4409, | |
| "num_input_tokens_seen": 178624, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 6.388888888888889, | |
| "grad_norm": 8.875, | |
| "learning_rate": 4.305071555472534e-05, | |
| "loss": 0.6276, | |
| "num_input_tokens_seen": 180096, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 6.444444444444445, | |
| "grad_norm": 9.25, | |
| "learning_rate": 4.288215727665129e-05, | |
| "loss": 0.3764, | |
| "num_input_tokens_seen": 181600, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 4.2711917772055e-05, | |
| "loss": 0.4464, | |
| "num_input_tokens_seen": 183232, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 6.555555555555555, | |
| "grad_norm": 10.625, | |
| "learning_rate": 4.254001304634151e-05, | |
| "loss": 0.4249, | |
| "num_input_tokens_seen": 184768, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 6.611111111111111, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 4.2366459261474933e-05, | |
| "loss": 0.3735, | |
| "num_input_tokens_seen": 186208, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 8.0, | |
| "learning_rate": 4.2191272734458955e-05, | |
| "loss": 0.3011, | |
| "num_input_tokens_seen": 187776, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 6.722222222222222, | |
| "grad_norm": 9.0, | |
| "learning_rate": 4.201446993580276e-05, | |
| "loss": 0.3692, | |
| "num_input_tokens_seen": 189344, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 6.777777777777778, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 4.183606748797251e-05, | |
| "loss": 0.3027, | |
| "num_input_tokens_seen": 190912, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 6.833333333333333, | |
| "grad_norm": 8.875, | |
| "learning_rate": 4.1656082163828566e-05, | |
| "loss": 0.3768, | |
| "num_input_tokens_seen": 192512, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 6.888888888888889, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 4.147453088504854e-05, | |
| "loss": 0.2615, | |
| "num_input_tokens_seen": 194112, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 6.944444444444445, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 4.129143072053638e-05, | |
| "loss": 0.4622, | |
| "num_input_tokens_seen": 195712, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 4.110679888481763e-05, | |
| "loss": 0.2119, | |
| "num_input_tokens_seen": 197248, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.28023624420166016, | |
| "eval_runtime": 0.5025, | |
| "eval_samples_per_second": 79.6, | |
| "eval_steps_per_second": 19.9, | |
| "num_input_tokens_seen": 197248, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 7.055555555555555, | |
| "grad_norm": 5.875, | |
| "learning_rate": 4.09206527364209e-05, | |
| "loss": 0.3463, | |
| "num_input_tokens_seen": 198784, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 7.111111111111111, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 4.073300977624594e-05, | |
| "loss": 0.3805, | |
| "num_input_tokens_seen": 200352, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 7.166666666666667, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 4.054388764591822e-05, | |
| "loss": 0.3768, | |
| "num_input_tokens_seen": 201856, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 7.222222222222222, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 4.035330412613035e-05, | |
| "loss": 0.3672, | |
| "num_input_tokens_seen": 203488, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 7.277777777777778, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 4.0161277134970345e-05, | |
| "loss": 0.279, | |
| "num_input_tokens_seen": 205056, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 7.333333333333333, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 3.996782472623705e-05, | |
| "loss": 0.4698, | |
| "num_input_tokens_seen": 206656, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 7.388888888888889, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 3.977296508774278e-05, | |
| "loss": 0.3624, | |
| "num_input_tokens_seen": 208288, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 7.444444444444445, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 3.957671653960337e-05, | |
| "loss": 0.433, | |
| "num_input_tokens_seen": 209856, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 3.9379097532515725e-05, | |
| "loss": 0.3233, | |
| "num_input_tokens_seen": 211424, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 7.555555555555555, | |
| "grad_norm": 8.875, | |
| "learning_rate": 3.918012664602317e-05, | |
| "loss": 0.2125, | |
| "num_input_tokens_seen": 212928, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 7.611111111111111, | |
| "grad_norm": 7.5, | |
| "learning_rate": 3.897982258676867e-05, | |
| "loss": 0.3247, | |
| "num_input_tokens_seen": 214496, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 7.666666666666667, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 3.8778204186736076e-05, | |
| "loss": 0.3252, | |
| "num_input_tokens_seen": 216064, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 7.722222222222222, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 3.8575290401479586e-05, | |
| "loss": 0.4832, | |
| "num_input_tokens_seen": 217600, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 7.777777777777778, | |
| "grad_norm": 10.625, | |
| "learning_rate": 3.837110030834161e-05, | |
| "loss": 0.3107, | |
| "num_input_tokens_seen": 219136, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 7.833333333333333, | |
| "grad_norm": 6.375, | |
| "learning_rate": 3.8165653104659185e-05, | |
| "loss": 0.2935, | |
| "num_input_tokens_seen": 220704, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 7.888888888888889, | |
| "grad_norm": 7.375, | |
| "learning_rate": 3.79589681059591e-05, | |
| "loss": 0.438, | |
| "num_input_tokens_seen": 222272, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 7.944444444444445, | |
| "grad_norm": 5.15625, | |
| "learning_rate": 3.775106474414188e-05, | |
| "loss": 0.1976, | |
| "num_input_tokens_seen": 223872, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 3.75419625656549e-05, | |
| "loss": 0.621, | |
| "num_input_tokens_seen": 225440, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.27384695410728455, | |
| "eval_runtime": 0.49, | |
| "eval_samples_per_second": 81.636, | |
| "eval_steps_per_second": 20.409, | |
| "num_input_tokens_seen": 225440, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 8.055555555555555, | |
| "grad_norm": 14.0, | |
| "learning_rate": 3.7331681229654635e-05, | |
| "loss": 0.4381, | |
| "num_input_tokens_seen": 226912, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 8.11111111111111, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 3.712024050615843e-05, | |
| "loss": 0.3431, | |
| "num_input_tokens_seen": 228512, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 8.166666666666666, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 3.690766027418573e-05, | |
| "loss": 0.4304, | |
| "num_input_tokens_seen": 230080, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 8.222222222222221, | |
| "grad_norm": 20.125, | |
| "learning_rate": 3.6693960519889106e-05, | |
| "loss": 0.31, | |
| "num_input_tokens_seen": 231616, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 8.277777777777779, | |
| "grad_norm": 4.71875, | |
| "learning_rate": 3.6479161334675296e-05, | |
| "loss": 0.3465, | |
| "num_input_tokens_seen": 233184, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 8.333333333333334, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 3.626328291331618e-05, | |
| "loss": 0.3556, | |
| "num_input_tokens_seen": 234688, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 8.38888888888889, | |
| "grad_norm": 8.625, | |
| "learning_rate": 3.60463455520502e-05, | |
| "loss": 0.4815, | |
| "num_input_tokens_seen": 236256, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 8.444444444444445, | |
| "grad_norm": 6.125, | |
| "learning_rate": 3.582836964667408e-05, | |
| "loss": 0.2465, | |
| "num_input_tokens_seen": 237856, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 3.560937569062538e-05, | |
| "loss": 0.2593, | |
| "num_input_tokens_seen": 239392, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 8.555555555555555, | |
| "grad_norm": 5.875, | |
| "learning_rate": 3.538938427305573e-05, | |
| "loss": 0.5183, | |
| "num_input_tokens_seen": 240960, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 8.61111111111111, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 3.516841607689501e-05, | |
| "loss": 0.347, | |
| "num_input_tokens_seen": 242496, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 8.666666666666666, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 3.494649187690695e-05, | |
| "loss": 0.4336, | |
| "num_input_tokens_seen": 244064, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 8.722222222222221, | |
| "grad_norm": 9.25, | |
| "learning_rate": 3.4723632537735846e-05, | |
| "loss": 0.303, | |
| "num_input_tokens_seen": 245664, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 8.777777777777779, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 3.449985901194498e-05, | |
| "loss": 0.3301, | |
| "num_input_tokens_seen": 247264, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 8.833333333333334, | |
| "grad_norm": 6.75, | |
| "learning_rate": 3.427519233804667e-05, | |
| "loss": 0.3121, | |
| "num_input_tokens_seen": 248832, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 8.88888888888889, | |
| "grad_norm": 5.15625, | |
| "learning_rate": 3.404965363852437e-05, | |
| "loss": 0.4487, | |
| "num_input_tokens_seen": 250432, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 8.944444444444445, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 3.382326411784672e-05, | |
| "loss": 0.2848, | |
| "num_input_tokens_seen": 252064, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 3.359604506047403e-05, | |
| "loss": 0.235, | |
| "num_input_tokens_seen": 253632, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 0.26062294840812683, | |
| "eval_runtime": 0.4918, | |
| "eval_samples_per_second": 81.336, | |
| "eval_steps_per_second": 20.334, | |
| "num_input_tokens_seen": 253632, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 9.055555555555555, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 3.336801782885712e-05, | |
| "loss": 0.5006, | |
| "num_input_tokens_seen": 255232, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 9.11111111111111, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 3.313920386142892e-05, | |
| "loss": 0.3224, | |
| "num_input_tokens_seen": 256800, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 9.166666666666666, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 3.290962467058891e-05, | |
| "loss": 0.4505, | |
| "num_input_tokens_seen": 258368, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 9.222222222222221, | |
| "grad_norm": 7.0, | |
| "learning_rate": 3.267930184068057e-05, | |
| "loss": 0.2689, | |
| "num_input_tokens_seen": 259872, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 9.277777777777779, | |
| "grad_norm": 8.75, | |
| "learning_rate": 3.244825702596205e-05, | |
| "loss": 0.2065, | |
| "num_input_tokens_seen": 261376, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 9.333333333333334, | |
| "grad_norm": 6.75, | |
| "learning_rate": 3.2216511948570374e-05, | |
| "loss": 0.3132, | |
| "num_input_tokens_seen": 262944, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 9.38888888888889, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 3.198408839647911e-05, | |
| "loss": 0.4963, | |
| "num_input_tokens_seen": 264544, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 9.444444444444445, | |
| "grad_norm": 6.75, | |
| "learning_rate": 3.1751008221450025e-05, | |
| "loss": 0.2322, | |
| "num_input_tokens_seen": 266080, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "grad_norm": 6.03125, | |
| "learning_rate": 3.151729333697854e-05, | |
| "loss": 0.2957, | |
| "num_input_tokens_seen": 267680, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 9.555555555555555, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 3.1282965716233594e-05, | |
| "loss": 0.5915, | |
| "num_input_tokens_seen": 269280, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 9.61111111111111, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 3.104804738999169e-05, | |
| "loss": 0.2449, | |
| "num_input_tokens_seen": 270912, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 9.666666666666666, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 3.0812560444565745e-05, | |
| "loss": 0.3233, | |
| "num_input_tokens_seen": 272512, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 9.722222222222221, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 3.057652701972848e-05, | |
| "loss": 0.3581, | |
| "num_input_tokens_seen": 274048, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 9.777777777777779, | |
| "grad_norm": 10.0, | |
| "learning_rate": 3.0339969306631005e-05, | |
| "loss": 0.4252, | |
| "num_input_tokens_seen": 275648, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 9.833333333333334, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 3.0102909545716396e-05, | |
| "loss": 0.3948, | |
| "num_input_tokens_seen": 277248, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 9.88888888888889, | |
| "grad_norm": 8.25, | |
| "learning_rate": 2.9865370024628775e-05, | |
| "loss": 0.3315, | |
| "num_input_tokens_seen": 278816, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 9.944444444444445, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 2.9627373076117863e-05, | |
| "loss": 0.2975, | |
| "num_input_tokens_seen": 280416, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 9.125, | |
| "learning_rate": 2.9388941075939334e-05, | |
| "loss": 0.3074, | |
| "num_input_tokens_seen": 281984, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.26525378227233887, | |
| "eval_runtime": 0.4922, | |
| "eval_samples_per_second": 81.262, | |
| "eval_steps_per_second": 20.315, | |
| "num_input_tokens_seen": 281984, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 10.055555555555555, | |
| "grad_norm": 12.75, | |
| "learning_rate": 2.9150096440751107e-05, | |
| "loss": 0.5148, | |
| "num_input_tokens_seen": 283552, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 10.11111111111111, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 2.8910861626005776e-05, | |
| "loss": 0.3101, | |
| "num_input_tokens_seen": 285120, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 10.166666666666666, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 2.8671259123839472e-05, | |
| "loss": 0.3843, | |
| "num_input_tokens_seen": 286720, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 10.222222222222221, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 2.843131146095719e-05, | |
| "loss": 0.245, | |
| "num_input_tokens_seen": 288288, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 10.277777777777779, | |
| "grad_norm": 7.0, | |
| "learning_rate": 2.8191041196514873e-05, | |
| "loss": 0.3618, | |
| "num_input_tokens_seen": 289856, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 10.333333333333334, | |
| "grad_norm": 10.625, | |
| "learning_rate": 2.795047091999849e-05, | |
| "loss": 0.6491, | |
| "num_input_tokens_seen": 291392, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 10.38888888888889, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 2.770962324910027e-05, | |
| "loss": 0.2309, | |
| "num_input_tokens_seen": 292864, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 10.444444444444445, | |
| "grad_norm": 5.875, | |
| "learning_rate": 2.7468520827592197e-05, | |
| "loss": 0.387, | |
| "num_input_tokens_seen": 294496, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 10.5, | |
| "grad_norm": 9.25, | |
| "learning_rate": 2.7227186323197162e-05, | |
| "loss": 0.4132, | |
| "num_input_tokens_seen": 296096, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 10.555555555555555, | |
| "grad_norm": 5.53125, | |
| "learning_rate": 2.6985642425457757e-05, | |
| "loss": 0.3229, | |
| "num_input_tokens_seen": 297632, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 10.61111111111111, | |
| "grad_norm": 8.625, | |
| "learning_rate": 2.674391184360313e-05, | |
| "loss": 0.3197, | |
| "num_input_tokens_seen": 299168, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 10.666666666666666, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 2.650201730441392e-05, | |
| "loss": 0.4312, | |
| "num_input_tokens_seen": 300736, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 10.722222222222221, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 2.6259981550085504e-05, | |
| "loss": 0.3807, | |
| "num_input_tokens_seen": 302304, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 10.777777777777779, | |
| "grad_norm": 8.125, | |
| "learning_rate": 2.60178273360899e-05, | |
| "loss": 0.2715, | |
| "num_input_tokens_seen": 303936, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 10.833333333333334, | |
| "grad_norm": 11.625, | |
| "learning_rate": 2.5775577429036345e-05, | |
| "loss": 0.2901, | |
| "num_input_tokens_seen": 305504, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 10.88888888888889, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 2.553325460453086e-05, | |
| "loss": 0.2219, | |
| "num_input_tokens_seen": 307040, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 10.944444444444445, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 2.5290881645034932e-05, | |
| "loss": 0.3929, | |
| "num_input_tokens_seen": 308608, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 2.504848133772358e-05, | |
| "loss": 0.2951, | |
| "num_input_tokens_seen": 310176, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 0.27018141746520996, | |
| "eval_runtime": 0.4943, | |
| "eval_samples_per_second": 80.919, | |
| "eval_steps_per_second": 20.23, | |
| "num_input_tokens_seen": 310176, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 11.055555555555555, | |
| "grad_norm": 11.75, | |
| "learning_rate": 2.4806076472342997e-05, | |
| "loss": 0.4875, | |
| "num_input_tokens_seen": 311712, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 11.11111111111111, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 2.4563689839067913e-05, | |
| "loss": 0.2882, | |
| "num_input_tokens_seen": 313280, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 11.166666666666666, | |
| "grad_norm": 8.75, | |
| "learning_rate": 2.432134422635893e-05, | |
| "loss": 0.4769, | |
| "num_input_tokens_seen": 314816, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 11.222222222222221, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 2.4079062418820002e-05, | |
| "loss": 0.2237, | |
| "num_input_tokens_seen": 316384, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 11.277777777777779, | |
| "grad_norm": 3.4375, | |
| "learning_rate": 2.3836867195056335e-05, | |
| "loss": 0.3421, | |
| "num_input_tokens_seen": 317984, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 11.333333333333334, | |
| "grad_norm": 6.5, | |
| "learning_rate": 2.3594781325532784e-05, | |
| "loss": 0.5096, | |
| "num_input_tokens_seen": 319552, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 11.38888888888889, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 2.3352827570433036e-05, | |
| "loss": 0.2577, | |
| "num_input_tokens_seen": 321184, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 11.444444444444445, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 2.3111028677519804e-05, | |
| "loss": 0.3533, | |
| "num_input_tokens_seen": 322688, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 11.5, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 2.2869407379996088e-05, | |
| "loss": 0.4808, | |
| "num_input_tokens_seen": 324224, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 11.555555555555555, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 2.2627986394367938e-05, | |
| "loss": 0.3517, | |
| "num_input_tokens_seen": 325760, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 11.61111111111111, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 2.238678841830867e-05, | |
| "loss": 0.3247, | |
| "num_input_tokens_seen": 327264, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 11.666666666666666, | |
| "grad_norm": 9.25, | |
| "learning_rate": 2.2145836128524902e-05, | |
| "loss": 0.2964, | |
| "num_input_tokens_seen": 328832, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 11.722222222222221, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 2.1905152178624595e-05, | |
| "loss": 0.3598, | |
| "num_input_tokens_seen": 330464, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 11.777777777777779, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 2.1664759196987182e-05, | |
| "loss": 0.2635, | |
| "num_input_tokens_seen": 332064, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 11.833333333333334, | |
| "grad_norm": 3.28125, | |
| "learning_rate": 2.1424679784636144e-05, | |
| "loss": 0.2413, | |
| "num_input_tokens_seen": 333664, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 11.88888888888889, | |
| "grad_norm": 6.5, | |
| "learning_rate": 2.118493651311413e-05, | |
| "loss": 0.4733, | |
| "num_input_tokens_seen": 335232, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 11.944444444444445, | |
| "grad_norm": 7.875, | |
| "learning_rate": 2.0945551922360818e-05, | |
| "loss": 0.2508, | |
| "num_input_tokens_seen": 336800, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 10.625, | |
| "learning_rate": 2.070654851859383e-05, | |
| "loss": 0.4453, | |
| "num_input_tokens_seen": 338400, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.2673038840293884, | |
| "eval_runtime": 0.4944, | |
| "eval_samples_per_second": 80.904, | |
| "eval_steps_per_second": 20.226, | |
| "num_input_tokens_seen": 338400, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 12.055555555555555, | |
| "grad_norm": 3.375, | |
| "learning_rate": 2.0467948772192713e-05, | |
| "loss": 0.5311, | |
| "num_input_tokens_seen": 339968, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 12.11111111111111, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 2.022977511558638e-05, | |
| "loss": 0.4152, | |
| "num_input_tokens_seen": 341536, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 12.166666666666666, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 1.9992049941144066e-05, | |
| "loss": 0.1981, | |
| "num_input_tokens_seen": 343104, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 12.222222222222221, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 1.9754795599070068e-05, | |
| "loss": 0.297, | |
| "num_input_tokens_seen": 344640, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 12.277777777777779, | |
| "grad_norm": 3.265625, | |
| "learning_rate": 1.9518034395302414e-05, | |
| "loss": 0.4067, | |
| "num_input_tokens_seen": 346272, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 12.333333333333334, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 1.9281788589415804e-05, | |
| "loss": 0.316, | |
| "num_input_tokens_seen": 347808, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 12.38888888888889, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 1.9046080392528735e-05, | |
| "loss": 0.3412, | |
| "num_input_tokens_seen": 349344, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 12.444444444444445, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 1.8810931965215356e-05, | |
| "loss": 0.2152, | |
| "num_input_tokens_seen": 350912, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 9.125, | |
| "learning_rate": 1.857636541542195e-05, | |
| "loss": 0.4586, | |
| "num_input_tokens_seen": 352480, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 12.555555555555555, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 1.8342402796388445e-05, | |
| "loss": 0.3069, | |
| "num_input_tokens_seen": 354048, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 12.61111111111111, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 1.8109066104575023e-05, | |
| "loss": 0.259, | |
| "num_input_tokens_seen": 355648, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 12.666666666666666, | |
| "grad_norm": 8.25, | |
| "learning_rate": 1.7876377277594053e-05, | |
| "loss": 0.2584, | |
| "num_input_tokens_seen": 357248, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 12.722222222222221, | |
| "grad_norm": 12.375, | |
| "learning_rate": 1.764435819214762e-05, | |
| "loss": 0.4422, | |
| "num_input_tokens_seen": 358752, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 12.777777777777779, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 1.7413030661970742e-05, | |
| "loss": 0.3268, | |
| "num_input_tokens_seen": 360288, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 12.833333333333334, | |
| "grad_norm": 9.125, | |
| "learning_rate": 1.7182416435780454e-05, | |
| "loss": 0.3052, | |
| "num_input_tokens_seen": 361920, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 12.88888888888889, | |
| "grad_norm": 5.75, | |
| "learning_rate": 1.695253719523115e-05, | |
| "loss": 0.5676, | |
| "num_input_tokens_seen": 363520, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 12.944444444444445, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 1.672341455287605e-05, | |
| "loss": 0.3776, | |
| "num_input_tokens_seen": 365088, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 7.375, | |
| "learning_rate": 1.649507005013532e-05, | |
| "loss": 0.3689, | |
| "num_input_tokens_seen": 366688, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_loss": 0.2658033072948456, | |
| "eval_runtime": 0.4934, | |
| "eval_samples_per_second": 81.065, | |
| "eval_steps_per_second": 20.266, | |
| "num_input_tokens_seen": 366688, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 13.055555555555555, | |
| "grad_norm": 8.5, | |
| "learning_rate": 1.6267525155270773e-05, | |
| "loss": 0.417, | |
| "num_input_tokens_seen": 368192, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 13.11111111111111, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 1.6040801261367493e-05, | |
| "loss": 0.4746, | |
| "num_input_tokens_seen": 369792, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 13.166666666666666, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 1.5814919684322545e-05, | |
| "loss": 0.3091, | |
| "num_input_tokens_seen": 371360, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 13.222222222222221, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.5589901660840896e-05, | |
| "loss": 0.4502, | |
| "num_input_tokens_seen": 372928, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 13.277777777777779, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 1.5365768346438797e-05, | |
| "loss": 0.2372, | |
| "num_input_tokens_seen": 374528, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 13.333333333333334, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 1.5142540813454836e-05, | |
| "loss": 0.2824, | |
| "num_input_tokens_seen": 376000, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 13.38888888888889, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 1.4920240049068748e-05, | |
| "loss": 0.288, | |
| "num_input_tokens_seen": 377600, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 13.444444444444445, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.4698886953328292e-05, | |
| "loss": 0.4595, | |
| "num_input_tokens_seen": 379200, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 13.5, | |
| "grad_norm": 5.375, | |
| "learning_rate": 1.4478502337184274e-05, | |
| "loss": 0.4339, | |
| "num_input_tokens_seen": 380704, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 13.555555555555555, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 1.4259106920533955e-05, | |
| "loss": 0.3858, | |
| "num_input_tokens_seen": 382272, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 13.61111111111111, | |
| "grad_norm": 19.25, | |
| "learning_rate": 1.4040721330273062e-05, | |
| "loss": 0.4033, | |
| "num_input_tokens_seen": 383776, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 13.666666666666666, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 1.3823366098356487e-05, | |
| "loss": 0.3113, | |
| "num_input_tokens_seen": 385344, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 13.722222222222221, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.3607061659867892e-05, | |
| "loss": 0.2684, | |
| "num_input_tokens_seen": 386912, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 13.777777777777779, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 1.3391828351098578e-05, | |
| "loss": 0.4218, | |
| "num_input_tokens_seen": 388480, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 13.833333333333334, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 1.3177686407635417e-05, | |
| "loss": 0.3828, | |
| "num_input_tokens_seen": 390048, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 13.88888888888889, | |
| "grad_norm": 5.53125, | |
| "learning_rate": 1.29646559624584e-05, | |
| "loss": 0.2402, | |
| "num_input_tokens_seen": 391584, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 13.944444444444445, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 1.2752757044047827e-05, | |
| "loss": 0.3431, | |
| "num_input_tokens_seen": 393184, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.2542009574501246e-05, | |
| "loss": 0.3106, | |
| "num_input_tokens_seen": 394752, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 0.2602008879184723, | |
| "eval_runtime": 0.4976, | |
| "eval_samples_per_second": 80.382, | |
| "eval_steps_per_second": 20.096, | |
| "num_input_tokens_seen": 394752, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 14.055555555555555, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 1.2332433367660442e-05, | |
| "loss": 0.3819, | |
| "num_input_tokens_seen": 396320, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 14.11111111111111, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 1.2124048127248644e-05, | |
| "loss": 0.3088, | |
| "num_input_tokens_seen": 397920, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 14.166666666666666, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 1.1916873445017982e-05, | |
| "loss": 0.3087, | |
| "num_input_tokens_seen": 399552, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 14.222222222222221, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 1.1710928798907556e-05, | |
| "loss": 0.2781, | |
| "num_input_tokens_seen": 401120, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 14.277777777777779, | |
| "grad_norm": 10.625, | |
| "learning_rate": 1.1506233551212186e-05, | |
| "loss": 0.3501, | |
| "num_input_tokens_seen": 402720, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 14.333333333333334, | |
| "grad_norm": 9.0, | |
| "learning_rate": 1.1302806946762004e-05, | |
| "loss": 0.3561, | |
| "num_input_tokens_seen": 404320, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 14.38888888888889, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 1.1100668111113166e-05, | |
| "loss": 0.2586, | |
| "num_input_tokens_seen": 405824, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 14.444444444444445, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 1.0899836048749645e-05, | |
| "loss": 0.4233, | |
| "num_input_tokens_seen": 407360, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 14.5, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 1.0700329641296541e-05, | |
| "loss": 0.5562, | |
| "num_input_tokens_seen": 408928, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 14.555555555555555, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 1.0502167645744895e-05, | |
| "loss": 0.3228, | |
| "num_input_tokens_seen": 410496, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 14.61111111111111, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 1.0305368692688174e-05, | |
| "loss": 0.229, | |
| "num_input_tokens_seen": 412032, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 14.666666666666666, | |
| "grad_norm": 8.5, | |
| "learning_rate": 1.01099512845707e-05, | |
| "loss": 0.3049, | |
| "num_input_tokens_seen": 413536, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 14.722222222222221, | |
| "grad_norm": 8.125, | |
| "learning_rate": 9.91593379394811e-06, | |
| "loss": 0.3829, | |
| "num_input_tokens_seen": 415136, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 14.777777777777779, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 9.723334461760006e-06, | |
| "loss": 0.6535, | |
| "num_input_tokens_seen": 416704, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 14.833333333333334, | |
| "grad_norm": 6.75, | |
| "learning_rate": 9.532171395615036e-06, | |
| "loss": 0.2201, | |
| "num_input_tokens_seen": 418272, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 14.88888888888889, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 9.342462568088416e-06, | |
| "loss": 0.4891, | |
| "num_input_tokens_seen": 419808, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 14.944444444444445, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 9.154225815032242e-06, | |
| "loss": 0.3325, | |
| "num_input_tokens_seen": 421344, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 8.967478833898612e-06, | |
| "loss": 0.2445, | |
| "num_input_tokens_seen": 422912, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_loss": 0.2658233046531677, | |
| "eval_runtime": 0.4947, | |
| "eval_samples_per_second": 80.851, | |
| "eval_steps_per_second": 20.213, | |
| "num_input_tokens_seen": 422912, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 15.055555555555555, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 8.78223918207575e-06, | |
| "loss": 0.38, | |
| "num_input_tokens_seen": 424480, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 15.11111111111111, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 8.598524275237322e-06, | |
| "loss": 0.3808, | |
| "num_input_tokens_seen": 426048, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 15.166666666666666, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 8.41635138570507e-06, | |
| "loss": 0.3339, | |
| "num_input_tokens_seen": 427648, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 15.222222222222221, | |
| "grad_norm": 12.5, | |
| "learning_rate": 8.235737640824908e-06, | |
| "loss": 0.5404, | |
| "num_input_tokens_seen": 429216, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 15.277777777777779, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 8.056700021356694e-06, | |
| "loss": 0.2567, | |
| "num_input_tokens_seen": 430720, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 15.333333333333334, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 7.879255359877705e-06, | |
| "loss": 0.3756, | |
| "num_input_tokens_seen": 432320, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 15.38888888888889, | |
| "grad_norm": 11.875, | |
| "learning_rate": 7.703420339200101e-06, | |
| "loss": 0.4089, | |
| "num_input_tokens_seen": 433888, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 15.444444444444445, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 7.529211490802498e-06, | |
| "loss": 0.4601, | |
| "num_input_tokens_seen": 435456, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 15.5, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 7.3566451932756744e-06, | |
| "loss": 0.2816, | |
| "num_input_tokens_seen": 437024, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 15.555555555555555, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 7.185737670782727e-06, | |
| "loss": 0.1451, | |
| "num_input_tokens_seen": 438528, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 15.61111111111111, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 7.016504991533726e-06, | |
| "loss": 0.2137, | |
| "num_input_tokens_seen": 440096, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 15.666666666666666, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 6.848963066275027e-06, | |
| "loss": 0.2693, | |
| "num_input_tokens_seen": 441632, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 15.722222222222221, | |
| "grad_norm": 8.375, | |
| "learning_rate": 6.683127646793411e-06, | |
| "loss": 0.4805, | |
| "num_input_tokens_seen": 443136, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 15.777777777777779, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 6.519014324435102e-06, | |
| "loss": 0.4068, | |
| "num_input_tokens_seen": 444736, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 15.833333333333334, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 6.356638528639955e-06, | |
| "loss": 0.2475, | |
| "num_input_tokens_seen": 446336, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 15.88888888888889, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 6.196015525490825e-06, | |
| "loss": 0.548, | |
| "num_input_tokens_seen": 447936, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 15.944444444444445, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 6.037160416278278e-06, | |
| "loss": 0.4648, | |
| "num_input_tokens_seen": 449440, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 5.880088136080814e-06, | |
| "loss": 0.261, | |
| "num_input_tokens_seen": 451008, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 0.2752898335456848, | |
| "eval_runtime": 0.4953, | |
| "eval_samples_per_second": 80.758, | |
| "eval_steps_per_second": 20.19, | |
| "num_input_tokens_seen": 451008, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 16.055555555555557, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 5.724813452360736e-06, | |
| "loss": 0.3175, | |
| "num_input_tokens_seen": 452576, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 16.11111111111111, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 5.571350963575728e-06, | |
| "loss": 0.4173, | |
| "num_input_tokens_seen": 454112, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 16.166666666666668, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 5.4197150978063965e-06, | |
| "loss": 0.3359, | |
| "num_input_tokens_seen": 455680, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 16.22222222222222, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 5.269920111399732e-06, | |
| "loss": 0.1756, | |
| "num_input_tokens_seen": 457280, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 16.27777777777778, | |
| "grad_norm": 8.125, | |
| "learning_rate": 5.121980087628803e-06, | |
| "loss": 0.2221, | |
| "num_input_tokens_seen": 458848, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 16.333333333333332, | |
| "grad_norm": 3.859375, | |
| "learning_rate": 4.975908935368701e-06, | |
| "loss": 0.3035, | |
| "num_input_tokens_seen": 460416, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 16.38888888888889, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 4.831720387788827e-06, | |
| "loss": 0.3419, | |
| "num_input_tokens_seen": 462048, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 16.444444444444443, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 4.689428001061774e-06, | |
| "loss": 0.3395, | |
| "num_input_tokens_seen": 463584, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 16.5, | |
| "grad_norm": 6.125, | |
| "learning_rate": 4.549045153088813e-06, | |
| "loss": 0.5784, | |
| "num_input_tokens_seen": 465152, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 16.555555555555557, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 4.410585042242124e-06, | |
| "loss": 0.471, | |
| "num_input_tokens_seen": 466720, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 16.61111111111111, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 4.274060686123959e-06, | |
| "loss": 0.2842, | |
| "num_input_tokens_seen": 468256, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 16.666666666666668, | |
| "grad_norm": 10.25, | |
| "learning_rate": 4.1394849203427284e-06, | |
| "loss": 0.4824, | |
| "num_input_tokens_seen": 469792, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 16.72222222222222, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 4.006870397306256e-06, | |
| "loss": 0.1715, | |
| "num_input_tokens_seen": 471360, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 16.77777777777778, | |
| "grad_norm": 12.4375, | |
| "learning_rate": 3.876229585032245e-06, | |
| "loss": 0.2982, | |
| "num_input_tokens_seen": 472896, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 16.833333333333332, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 3.7475747659760502e-06, | |
| "loss": 0.5391, | |
| "num_input_tokens_seen": 474496, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 16.88888888888889, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 3.6209180358759394e-06, | |
| "loss": 0.361, | |
| "num_input_tokens_seen": 476064, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 16.944444444444443, | |
| "grad_norm": 4.5, | |
| "learning_rate": 3.4962713026158694e-06, | |
| "loss": 0.2989, | |
| "num_input_tokens_seen": 477600, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 6.625, | |
| "learning_rate": 3.373646285105958e-06, | |
| "loss": 0.5208, | |
| "num_input_tokens_seen": 479104, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_loss": 0.26678234338760376, | |
| "eval_runtime": 0.4964, | |
| "eval_samples_per_second": 80.576, | |
| "eval_steps_per_second": 20.144, | |
| "num_input_tokens_seen": 479104, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 17.055555555555557, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 3.2530545121807145e-06, | |
| "loss": 0.3504, | |
| "num_input_tokens_seen": 480736, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 17.11111111111111, | |
| "grad_norm": 5.875, | |
| "learning_rate": 3.1345073215151066e-06, | |
| "loss": 0.366, | |
| "num_input_tokens_seen": 482304, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 17.166666666666668, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 3.0180158585586397e-06, | |
| "loss": 0.3029, | |
| "num_input_tokens_seen": 483872, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 17.22222222222222, | |
| "grad_norm": 14.375, | |
| "learning_rate": 2.9035910754875136e-06, | |
| "loss": 0.4225, | |
| "num_input_tokens_seen": 485408, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 17.27777777777778, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 2.7912437301749026e-06, | |
| "loss": 0.4289, | |
| "num_input_tokens_seen": 486944, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 17.333333333333332, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 2.6809843851795357e-06, | |
| "loss": 0.3148, | |
| "num_input_tokens_seen": 488544, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 17.38888888888889, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 2.57282340675267e-06, | |
| "loss": 0.2946, | |
| "num_input_tokens_seen": 490144, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 17.444444444444443, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 2.4667709638634434e-06, | |
| "loss": 0.319, | |
| "num_input_tokens_seen": 491680, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 17.5, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 2.3628370272428564e-06, | |
| "loss": 0.3226, | |
| "num_input_tokens_seen": 493248, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 17.555555555555557, | |
| "grad_norm": 8.625, | |
| "learning_rate": 2.2610313684463177e-06, | |
| "loss": 0.3729, | |
| "num_input_tokens_seen": 494816, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 17.61111111111111, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 2.1613635589349756e-06, | |
| "loss": 0.4399, | |
| "num_input_tokens_seen": 496416, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 17.666666666666668, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 2.063842969175847e-06, | |
| "loss": 0.3715, | |
| "num_input_tokens_seen": 497952, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 17.72222222222222, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 1.968478767760812e-06, | |
| "loss": 0.3388, | |
| "num_input_tokens_seen": 499520, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 17.77777777777778, | |
| "grad_norm": 10.875, | |
| "learning_rate": 1.8752799205445982e-06, | |
| "loss": 0.6584, | |
| "num_input_tokens_seen": 501120, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 17.833333333333332, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 1.784255189801895e-06, | |
| "loss": 0.3057, | |
| "num_input_tokens_seen": 502688, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 17.88888888888889, | |
| "grad_norm": 4.75, | |
| "learning_rate": 1.6954131334034922e-06, | |
| "loss": 0.2274, | |
| "num_input_tokens_seen": 504224, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 17.944444444444443, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 1.6087621040117157e-06, | |
| "loss": 0.3742, | |
| "num_input_tokens_seen": 505824, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.524310248295152e-06, | |
| "loss": 0.2563, | |
| "num_input_tokens_seen": 507392, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_loss": 0.2767786383628845, | |
| "eval_runtime": 0.4945, | |
| "eval_samples_per_second": 80.882, | |
| "eval_steps_per_second": 20.221, | |
| "num_input_tokens_seen": 507392, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 18.055555555555557, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 1.4420655061626932e-06, | |
| "loss": 0.3343, | |
| "num_input_tokens_seen": 508928, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 18.11111111111111, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 1.362035610017079e-06, | |
| "loss": 0.3041, | |
| "num_input_tokens_seen": 510496, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 18.166666666666668, | |
| "grad_norm": 15.3125, | |
| "learning_rate": 1.2842280840278997e-06, | |
| "loss": 0.4172, | |
| "num_input_tokens_seen": 512032, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 18.22222222222222, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 1.2086502434241865e-06, | |
| "loss": 0.2408, | |
| "num_input_tokens_seen": 513504, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 18.27777777777778, | |
| "grad_norm": 7.625, | |
| "learning_rate": 1.1353091938067023e-06, | |
| "loss": 0.2925, | |
| "num_input_tokens_seen": 515040, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 18.333333333333332, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 1.0642118304798442e-06, | |
| "loss": 0.3155, | |
| "num_input_tokens_seen": 516640, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 18.38888888888889, | |
| "grad_norm": 9.375, | |
| "learning_rate": 9.95364837803392e-07, | |
| "loss": 0.3967, | |
| "num_input_tokens_seen": 518208, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 18.444444444444443, | |
| "grad_norm": 9.0, | |
| "learning_rate": 9.287746885640603e-07, | |
| "loss": 0.3127, | |
| "num_input_tokens_seen": 519776, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 18.5, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 8.64447643366953e-07, | |
| "loss": 0.2802, | |
| "num_input_tokens_seen": 521344, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 18.555555555555557, | |
| "grad_norm": 2.890625, | |
| "learning_rate": 8.023897500469391e-07, | |
| "loss": 0.5487, | |
| "num_input_tokens_seen": 522976, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 18.61111111111111, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 7.426068431000882e-07, | |
| "loss": 0.3772, | |
| "num_input_tokens_seen": 524544, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 18.666666666666668, | |
| "grad_norm": 8.125, | |
| "learning_rate": 6.851045431350927e-07, | |
| "loss": 0.2568, | |
| "num_input_tokens_seen": 526112, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 18.72222222222222, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 6.298882563448599e-07, | |
| "loss": 0.3091, | |
| "num_input_tokens_seen": 527712, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 18.77777777777778, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 5.769631739982267e-07, | |
| "loss": 0.4354, | |
| "num_input_tokens_seen": 529280, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 18.833333333333332, | |
| "grad_norm": 9.75, | |
| "learning_rate": 5.263342719518921e-07, | |
| "loss": 0.421, | |
| "num_input_tokens_seen": 530816, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 18.88888888888889, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 4.780063101826132e-07, | |
| "loss": 0.3758, | |
| "num_input_tokens_seen": 532448, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 18.944444444444443, | |
| "grad_norm": 12.5, | |
| "learning_rate": 4.319838323396691e-07, | |
| "loss": 0.4763, | |
| "num_input_tokens_seen": 534016, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 3.88271165317694e-07, | |
| "loss": 0.4139, | |
| "num_input_tokens_seen": 535584, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_loss": 0.2729262709617615, | |
| "eval_runtime": 0.4948, | |
| "eval_samples_per_second": 80.84, | |
| "eval_steps_per_second": 20.21, | |
| "num_input_tokens_seen": 535584, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 19.055555555555557, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 3.468724188498751e-07, | |
| "loss": 0.5006, | |
| "num_input_tokens_seen": 537152, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 19.11111111111111, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 3.077914851215585e-07, | |
| "loss": 0.263, | |
| "num_input_tokens_seen": 538720, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 19.166666666666668, | |
| "grad_norm": 9.0, | |
| "learning_rate": 2.71032038404323e-07, | |
| "loss": 0.3506, | |
| "num_input_tokens_seen": 540256, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 19.22222222222222, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 2.365975347105448e-07, | |
| "loss": 0.3691, | |
| "num_input_tokens_seen": 541824, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 19.27777777777778, | |
| "grad_norm": 5.53125, | |
| "learning_rate": 2.0449121146845774e-07, | |
| "loss": 0.2407, | |
| "num_input_tokens_seen": 543424, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 19.333333333333332, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 1.747160872177883e-07, | |
| "loss": 0.3171, | |
| "num_input_tokens_seen": 545024, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 19.38888888888889, | |
| "grad_norm": 6.75, | |
| "learning_rate": 1.472749613259661e-07, | |
| "loss": 0.3309, | |
| "num_input_tokens_seen": 546560, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 19.444444444444443, | |
| "grad_norm": 8.375, | |
| "learning_rate": 1.22170413724923e-07, | |
| "loss": 0.3624, | |
| "num_input_tokens_seen": 548096, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 19.5, | |
| "grad_norm": 10.5, | |
| "learning_rate": 9.940480466855417e-08, | |
| "loss": 0.4127, | |
| "num_input_tokens_seen": 549664, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 19.555555555555557, | |
| "grad_norm": 3.875, | |
| "learning_rate": 7.898027451078982e-08, | |
| "loss": 0.2348, | |
| "num_input_tokens_seen": 551264, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 19.61111111111111, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 6.089874350439506e-08, | |
| "loss": 0.4732, | |
| "num_input_tokens_seen": 552800, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 19.666666666666668, | |
| "grad_norm": 12.4375, | |
| "learning_rate": 4.516191162040051e-08, | |
| "loss": 0.4527, | |
| "num_input_tokens_seen": 554336, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 19.72222222222222, | |
| "grad_norm": 5.5, | |
| "learning_rate": 3.177125838830786e-08, | |
| "loss": 0.2406, | |
| "num_input_tokens_seen": 555872, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 19.77777777777778, | |
| "grad_norm": 10.375, | |
| "learning_rate": 2.0728042756967824e-08, | |
| "loss": 0.3651, | |
| "num_input_tokens_seen": 557504, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 19.833333333333332, | |
| "grad_norm": 7.25, | |
| "learning_rate": 1.2033302976222071e-08, | |
| "loss": 0.3143, | |
| "num_input_tokens_seen": 559040, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 19.88888888888889, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 5.687856499297928e-09, | |
| "loss": 0.4014, | |
| "num_input_tokens_seen": 560640, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 19.944444444444443, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.692299905944883e-09, | |
| "loss": 0.3499, | |
| "num_input_tokens_seen": 562176, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 5.875, | |
| "learning_rate": 4.700884634611935e-11, | |
| "loss": 0.3828, | |
| "num_input_tokens_seen": 563744, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_loss": 0.26922377943992615, | |
| "eval_runtime": 0.4933, | |
| "eval_samples_per_second": 81.086, | |
| "eval_steps_per_second": 20.272, | |
| "num_input_tokens_seen": 563744, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "num_input_tokens_seen": 563744, | |
| "step": 1800, | |
| "total_flos": 2.538603807296717e+16, | |
| "train_loss": 0.4076899822221862, | |
| "train_runtime": 254.7052, | |
| "train_samples_per_second": 28.268, | |
| "train_steps_per_second": 7.067 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1800, | |
| "num_input_tokens_seen": 563744, | |
| "num_train_epochs": 20, | |
| "save_steps": 90, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.538603807296717e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |