| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 2702, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 1.4760147601476015e-06, | |
| "loss": 2.6485, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 7.380073800738008e-06, | |
| "loss": 2.6486, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 1.4760147601476015e-05, | |
| "loss": 2.6449, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 2.2140221402214025e-05, | |
| "loss": 2.6688, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 2.952029520295203e-05, | |
| "loss": 2.6268, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 3.690036900369004e-05, | |
| "loss": 2.6019, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.25, | |
| "learning_rate": 4.428044280442805e-05, | |
| "loss": 2.5532, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 5.166051660516605e-05, | |
| "loss": 2.508, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 5.904059040590406e-05, | |
| "loss": 2.4591, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 6.642066420664207e-05, | |
| "loss": 2.4235, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 7.380073800738008e-05, | |
| "loss": 2.3654, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 8.118081180811809e-05, | |
| "loss": 2.3604, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 8.85608856088561e-05, | |
| "loss": 2.3221, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 9.59409594095941e-05, | |
| "loss": 2.3041, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 0.0001033210332103321, | |
| "loss": 2.3051, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.109375, | |
| "learning_rate": 0.00011070110701107013, | |
| "loss": 2.2795, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.0986328125, | |
| "learning_rate": 0.00011808118081180812, | |
| "loss": 2.2926, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 0.00012546125461254613, | |
| "loss": 2.2594, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 0.00013284132841328414, | |
| "loss": 2.2639, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 0.00014022140221402215, | |
| "loss": 2.2496, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 0.00014760147601476016, | |
| "loss": 2.2428, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.1044921875, | |
| "learning_rate": 0.00015498154981549817, | |
| "loss": 2.2572, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.1025390625, | |
| "learning_rate": 0.00016236162361623618, | |
| "loss": 2.2447, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.10888671875, | |
| "learning_rate": 0.0001697416974169742, | |
| "loss": 2.2258, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 0.0001771217712177122, | |
| "loss": 2.2284, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.109375, | |
| "learning_rate": 0.0001845018450184502, | |
| "loss": 2.2346, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.10595703125, | |
| "learning_rate": 0.0001918819188191882, | |
| "loss": 2.2365, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.11083984375, | |
| "learning_rate": 0.00019926199261992622, | |
| "loss": 2.2068, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.10595703125, | |
| "learning_rate": 0.0002066420664206642, | |
| "loss": 2.2316, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.10986328125, | |
| "learning_rate": 0.00021402214022140222, | |
| "loss": 2.2427, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 0.00022140221402214025, | |
| "loss": 2.2405, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 0.00022878228782287826, | |
| "loss": 2.2197, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.10693359375, | |
| "learning_rate": 0.00023616236162361624, | |
| "loss": 2.2411, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.1181640625, | |
| "learning_rate": 0.00024354243542435425, | |
| "loss": 2.2195, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.10693359375, | |
| "learning_rate": 0.00025092250922509226, | |
| "loss": 2.2147, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.10888671875, | |
| "learning_rate": 0.00025830258302583027, | |
| "loss": 2.2341, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 0.0002656826568265683, | |
| "loss": 2.2116, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 0.0002730627306273063, | |
| "loss": 2.1973, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.10693359375, | |
| "learning_rate": 0.0002804428044280443, | |
| "loss": 2.2137, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.10302734375, | |
| "learning_rate": 0.0002878228782287823, | |
| "loss": 2.2388, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.10595703125, | |
| "learning_rate": 0.0002952029520295203, | |
| "loss": 2.2111, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.10205078125, | |
| "learning_rate": 0.00030258302583025833, | |
| "loss": 2.2361, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.10986328125, | |
| "learning_rate": 0.00030996309963099634, | |
| "loss": 2.1882, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.1044921875, | |
| "learning_rate": 0.00031734317343173435, | |
| "loss": 2.2399, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.10546875, | |
| "learning_rate": 0.00032472324723247236, | |
| "loss": 2.2062, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.10546875, | |
| "learning_rate": 0.0003321033210332103, | |
| "loss": 2.2068, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.10498046875, | |
| "learning_rate": 0.0003394833948339484, | |
| "loss": 2.2058, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.10791015625, | |
| "learning_rate": 0.0003468634686346864, | |
| "loss": 2.2137, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 0.0003542435424354244, | |
| "loss": 2.2127, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.09716796875, | |
| "learning_rate": 0.00036162361623616235, | |
| "loss": 2.2093, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.1015625, | |
| "learning_rate": 0.0003690036900369004, | |
| "loss": 2.2063, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 0.0003763837638376384, | |
| "loss": 2.2347, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 0.0003837638376383764, | |
| "loss": 2.1907, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 0.0003911439114391144, | |
| "loss": 2.2137, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.0986328125, | |
| "learning_rate": 0.00039852398523985245, | |
| "loss": 2.1975, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.09619140625, | |
| "learning_rate": 0.00039999732792377247, | |
| "loss": 2.203, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.09765625, | |
| "learning_rate": 0.00039998647273646793, | |
| "loss": 2.2062, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 0.00039996726788618994, | |
| "loss": 2.2064, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.09619140625, | |
| "learning_rate": 0.00039993971417476293, | |
| "loss": 2.1835, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 0.0003999038127525856, | |
| "loss": 2.1989, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.09716796875, | |
| "learning_rate": 0.00039985956511858335, | |
| "loss": 2.2137, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.09619140625, | |
| "learning_rate": 0.00039980697312014523, | |
| "loss": 2.1926, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.095703125, | |
| "learning_rate": 0.00039974603895304704, | |
| "loss": 2.1829, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.1005859375, | |
| "learning_rate": 0.00039967676516135974, | |
| "loss": 2.1946, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.099609375, | |
| "learning_rate": 0.00039959915463734285, | |
| "loss": 2.1814, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.099609375, | |
| "learning_rate": 0.00039951321062132425, | |
| "loss": 2.187, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 0.00039941893670156453, | |
| "loss": 2.2039, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.09423828125, | |
| "learning_rate": 0.0003993163368141071, | |
| "loss": 2.1839, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.099609375, | |
| "learning_rate": 0.0003992054152426141, | |
| "loss": 2.2013, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.09912109375, | |
| "learning_rate": 0.0003990861766181874, | |
| "loss": 2.1924, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.09619140625, | |
| "learning_rate": 0.0003989586259191755, | |
| "loss": 2.1955, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.0966796875, | |
| "learning_rate": 0.0003988227684709653, | |
| "loss": 2.1965, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.095703125, | |
| "learning_rate": 0.00039867860994575994, | |
| "loss": 2.1946, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.09765625, | |
| "learning_rate": 0.000398526156362342, | |
| "loss": 2.1955, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 0.0003983654140858221, | |
| "loss": 2.1897, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.09765625, | |
| "learning_rate": 0.00039819638982737353, | |
| "loss": 2.189, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.09423828125, | |
| "learning_rate": 0.0003980190906439514, | |
| "loss": 2.1812, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.0986328125, | |
| "learning_rate": 0.00039783352393799856, | |
| "loss": 2.188, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.0927734375, | |
| "learning_rate": 0.00039763969745713635, | |
| "loss": 2.1859, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.095703125, | |
| "learning_rate": 0.0003974376192938411, | |
| "loss": 2.1697, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.09765625, | |
| "learning_rate": 0.0003972272978851061, | |
| "loss": 2.2024, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.09375, | |
| "learning_rate": 0.00039700874201208976, | |
| "loss": 2.1922, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.095703125, | |
| "learning_rate": 0.00039678196079974865, | |
| "loss": 2.1752, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.09375, | |
| "learning_rate": 0.00039654696371645663, | |
| "loss": 2.1947, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.09619140625, | |
| "learning_rate": 0.0003963037605736096, | |
| "loss": 2.194, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.09375, | |
| "learning_rate": 0.0003960523615252156, | |
| "loss": 2.2065, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.09619140625, | |
| "learning_rate": 0.00039579277706747125, | |
| "loss": 2.1707, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.0947265625, | |
| "learning_rate": 0.00039552501803832336, | |
| "loss": 2.1889, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.09521484375, | |
| "learning_rate": 0.0003952490956170161, | |
| "loss": 2.1876, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.09375, | |
| "learning_rate": 0.00039496502132362494, | |
| "loss": 2.1877, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.09765625, | |
| "learning_rate": 0.000394672807018575, | |
| "loss": 2.177, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 0.0003943724649021464, | |
| "loss": 2.2009, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.09423828125, | |
| "learning_rate": 0.00039406400751396445, | |
| "loss": 2.1802, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.10205078125, | |
| "learning_rate": 0.0003937474477324764, | |
| "loss": 2.1789, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.09521484375, | |
| "learning_rate": 0.00039342279877441357, | |
| "loss": 2.1829, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 0.00039309007419423964, | |
| "loss": 2.1871, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.09716796875, | |
| "learning_rate": 0.00039274928788358477, | |
| "loss": 2.2183, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.0966796875, | |
| "learning_rate": 0.00039240045407066556, | |
| "loss": 2.1888, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.09423828125, | |
| "learning_rate": 0.00039204358731969083, | |
| "loss": 2.1962, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.1005859375, | |
| "learning_rate": 0.0003916787025302538, | |
| "loss": 2.1746, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.0986328125, | |
| "learning_rate": 0.0003913058149367101, | |
| "loss": 2.1784, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 0.0003909249401075413, | |
| "loss": 2.1903, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.09423828125, | |
| "learning_rate": 0.0003905360939447052, | |
| "loss": 2.1772, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 0.00039013929268297195, | |
| "loss": 2.2023, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.099609375, | |
| "learning_rate": 0.00038973455288924614, | |
| "loss": 2.1887, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.09765625, | |
| "learning_rate": 0.0003893218914618749, | |
| "loss": 2.1764, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 0.00038890132562994286, | |
| "loss": 2.1911, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 0.0003884728729525525, | |
| "loss": 2.2008, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.0966796875, | |
| "learning_rate": 0.0003880365513180908, | |
| "loss": 2.1832, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.09619140625, | |
| "learning_rate": 0.00038759237894348306, | |
| "loss": 2.1775, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.1005859375, | |
| "learning_rate": 0.0003871403743734316, | |
| "loss": 2.1768, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.095703125, | |
| "learning_rate": 0.0003866805564796421, | |
| "loss": 2.1825, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.09521484375, | |
| "learning_rate": 0.000386212944460035, | |
| "loss": 2.191, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.0966796875, | |
| "learning_rate": 0.0003857375578379449, | |
| "loss": 2.168, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 0.0003852544164613043, | |
| "loss": 2.1844, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.09765625, | |
| "learning_rate": 0.0003847635405018162, | |
| "loss": 2.1926, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.099609375, | |
| "learning_rate": 0.00038426495045411064, | |
| "loss": 2.1935, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.0966796875, | |
| "learning_rate": 0.0003837586671348901, | |
| "loss": 2.1654, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.0986328125, | |
| "learning_rate": 0.00038324471168205945, | |
| "loss": 2.1816, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.095703125, | |
| "learning_rate": 0.0003827231055538443, | |
| "loss": 2.1697, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.095703125, | |
| "learning_rate": 0.0003821938705278944, | |
| "loss": 2.1581, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.0966796875, | |
| "learning_rate": 0.00038165702870037485, | |
| "loss": 2.1588, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.0986328125, | |
| "learning_rate": 0.0003811126024850432, | |
| "loss": 2.1842, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.09912109375, | |
| "learning_rate": 0.0003805606146123139, | |
| "loss": 2.169, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.09765625, | |
| "learning_rate": 0.0003800010881283093, | |
| "loss": 2.17, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 0.0003794340463938972, | |
| "loss": 2.1613, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.0966796875, | |
| "learning_rate": 0.0003788595130837157, | |
| "loss": 2.1806, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 0.00037827751218518494, | |
| "loss": 2.1841, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.0966796875, | |
| "learning_rate": 0.000377688067997505, | |
| "loss": 2.1827, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 0.000377091205130642, | |
| "loss": 2.1846, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.1005859375, | |
| "learning_rate": 0.0003764869485043003, | |
| "loss": 2.1877, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.09765625, | |
| "learning_rate": 0.0003758753233468823, | |
| "loss": 2.1702, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.0966796875, | |
| "learning_rate": 0.00037525635519443466, | |
| "loss": 2.1755, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.099609375, | |
| "learning_rate": 0.00037463006988958266, | |
| "loss": 2.1691, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.1005859375, | |
| "learning_rate": 0.0003739964935804509, | |
| "loss": 2.1713, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.095703125, | |
| "learning_rate": 0.0003733556527195719, | |
| "loss": 2.1749, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.0986328125, | |
| "learning_rate": 0.00037270757406278126, | |
| "loss": 2.1685, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.0986328125, | |
| "learning_rate": 0.00037205228466810094, | |
| "loss": 2.1818, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.0986328125, | |
| "learning_rate": 0.00037138981189460945, | |
| "loss": 2.1661, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 0.00037072018340129936, | |
| "loss": 2.1733, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 0.0003700434271459229, | |
| "loss": 2.1749, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 0.0003693595713838243, | |
| "loss": 2.1902, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.099609375, | |
| "learning_rate": 0.0003686686446667605, | |
| "loss": 2.1807, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.1025390625, | |
| "learning_rate": 0.0003679706758417087, | |
| "loss": 2.1834, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.09912109375, | |
| "learning_rate": 0.0003672656940496621, | |
| "loss": 2.1786, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.0986328125, | |
| "learning_rate": 0.0003665537287244134, | |
| "loss": 2.1742, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.09765625, | |
| "learning_rate": 0.00036583480959132567, | |
| "loss": 2.1676, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 0.0003651089666660914, | |
| "loss": 2.1633, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.09765625, | |
| "learning_rate": 0.0003643762302534792, | |
| "loss": 2.1699, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.1005859375, | |
| "learning_rate": 0.0003636366309460688, | |
| "loss": 2.1772, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.1005859375, | |
| "learning_rate": 0.00036289019962297347, | |
| "loss": 2.1602, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.1015625, | |
| "learning_rate": 0.000362136967448551, | |
| "loss": 2.1592, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.099609375, | |
| "learning_rate": 0.00036137696587110234, | |
| "loss": 2.1497, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 0.0003606102266215589, | |
| "loss": 2.1816, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 0.0003598367817121574, | |
| "loss": 2.1681, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.0986328125, | |
| "learning_rate": 0.0003590566634351036, | |
| "loss": 2.179, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.09912109375, | |
| "learning_rate": 0.000358269904361224, | |
| "loss": 2.186, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.09765625, | |
| "learning_rate": 0.00035747653733860576, | |
| "loss": 2.1635, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 0.0003566765954912256, | |
| "loss": 2.1863, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.099609375, | |
| "learning_rate": 0.0003558701122175666, | |
| "loss": 2.1639, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 0.0003550571211892238, | |
| "loss": 2.1795, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 0.00035423765634949844, | |
| "loss": 2.1634, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 0.0003534117519119807, | |
| "loss": 2.1673, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.1005859375, | |
| "learning_rate": 0.0003525794423591214, | |
| "loss": 2.1722, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.09765625, | |
| "learning_rate": 0.00035174076244079216, | |
| "loss": 2.1687, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.10498046875, | |
| "learning_rate": 0.00035089574717283466, | |
| "loss": 2.1736, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 0.0003500444318355986, | |
| "loss": 2.1784, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.1005859375, | |
| "learning_rate": 0.0003491868519724688, | |
| "loss": 2.1783, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 0.000348323043388381, | |
| "loss": 2.1674, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.09912109375, | |
| "learning_rate": 0.00034745304214832726, | |
| "loss": 2.1791, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 0.0003465768845758502, | |
| "loss": 2.1573, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 0.00034569460725152615, | |
| "loss": 2.1842, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.09765625, | |
| "learning_rate": 0.00034480624701143807, | |
| "loss": 2.1624, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 0.00034391184094563764, | |
| "loss": 2.1892, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 0.00034301142639659663, | |
| "loss": 2.1788, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.1005859375, | |
| "learning_rate": 0.0003421050409576478, | |
| "loss": 2.174, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 0.00034119272247141536, | |
| "loss": 2.171, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.1015625, | |
| "learning_rate": 0.0003402745090282351, | |
| "loss": 2.1571, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.10498046875, | |
| "learning_rate": 0.00033935043896456384, | |
| "loss": 2.1716, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.09912109375, | |
| "learning_rate": 0.000338420550861379, | |
| "loss": 2.1631, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.09912109375, | |
| "learning_rate": 0.0003374848835425679, | |
| "loss": 2.1652, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.099609375, | |
| "learning_rate": 0.00033654347607330656, | |
| "loss": 2.1661, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.10205078125, | |
| "learning_rate": 0.0003355963677584288, | |
| "loss": 2.1697, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.1005859375, | |
| "learning_rate": 0.00033464359814078536, | |
| "loss": 2.1937, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.10205078125, | |
| "learning_rate": 0.0003336852069995927, | |
| "loss": 2.1646, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 0.0003327212343487725, | |
| "loss": 2.1635, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 0.0003317517204352804, | |
| "loss": 2.149, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.10302734375, | |
| "learning_rate": 0.0003307767057374266, | |
| "loss": 2.1678, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.1015625, | |
| "learning_rate": 0.00032979623096318487, | |
| "loss": 2.1643, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.1005859375, | |
| "learning_rate": 0.00032881033704849357, | |
| "loss": 2.1748, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 0.00032781906515554646, | |
| "loss": 2.1617, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.1015625, | |
| "learning_rate": 0.0003268224566710738, | |
| "loss": 2.1529, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.1025390625, | |
| "learning_rate": 0.00032582055320461465, | |
| "loss": 2.1781, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.10546875, | |
| "learning_rate": 0.0003248133965867798, | |
| "loss": 2.1429, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 0.00032380102886750493, | |
| "loss": 2.1629, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.1005859375, | |
| "learning_rate": 0.000322783492314295, | |
| "loss": 2.1655, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.10205078125, | |
| "learning_rate": 0.0003217608294104601, | |
| "loss": 2.1758, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.09912109375, | |
| "learning_rate": 0.00032073308285334085, | |
| "loss": 2.1565, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 0.0003197002955525264, | |
| "loss": 2.1806, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.099609375, | |
| "learning_rate": 0.00031866251062806267, | |
| "loss": 2.1701, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 0.00031761977140865207, | |
| "loss": 2.1675, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.1005859375, | |
| "learning_rate": 0.0003165721214298444, | |
| "loss": 2.1671, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 0.0003155196044322193, | |
| "loss": 2.1575, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.1015625, | |
| "learning_rate": 0.00031446226435956, | |
| "loss": 2.1381, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.1015625, | |
| "learning_rate": 0.0003134001453570186, | |
| "loss": 2.1626, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.10205078125, | |
| "learning_rate": 0.00031233329176927295, | |
| "loss": 2.1648, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.10205078125, | |
| "learning_rate": 0.0003112617481386752, | |
| "loss": 2.1671, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 0.0003101855592033922, | |
| "loss": 2.1648, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.1044921875, | |
| "learning_rate": 0.0003091047698955375, | |
| "loss": 2.1746, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.10302734375, | |
| "learning_rate": 0.00030801942533929545, | |
| "loss": 2.1603, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.1005859375, | |
| "learning_rate": 0.00030692957084903726, | |
| "loss": 2.172, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 0.00030583525192742897, | |
| "loss": 2.1643, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.1005859375, | |
| "learning_rate": 0.00030473651426353167, | |
| "loss": 2.1756, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 0.00030363340373089413, | |
| "loss": 2.1637, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 0.00030252596638563714, | |
| "loss": 2.1718, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.1015625, | |
| "learning_rate": 0.000301414248464531, | |
| "loss": 2.1552, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.1025390625, | |
| "learning_rate": 0.000300298296383065, | |
| "loss": 2.187, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.10205078125, | |
| "learning_rate": 0.00029917815673350935, | |
| "loss": 2.1747, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 0.0002980538762829698, | |
| "loss": 2.1677, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.10205078125, | |
| "learning_rate": 0.00029692550197143563, | |
| "loss": 2.1476, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 0.00029579308090981913, | |
| "loss": 2.163, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.1025390625, | |
| "learning_rate": 0.00029465666037798935, | |
| "loss": 2.1673, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 0.0002935162878227975, | |
| "loss": 2.1737, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 0.0002923720108560964, | |
| "loss": 2.1412, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.1044921875, | |
| "learning_rate": 0.00029122387725275244, | |
| "loss": 2.1455, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 0.00029007193494865103, | |
| "loss": 2.1477, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.1025390625, | |
| "learning_rate": 0.00028891623203869523, | |
| "loss": 2.1486, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 0.0002877568167747975, | |
| "loss": 2.1703, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.1044921875, | |
| "learning_rate": 0.0002865937375638654, | |
| "loss": 2.1819, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.10205078125, | |
| "learning_rate": 0.0002854270429657805, | |
| "loss": 2.1564, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.10302734375, | |
| "learning_rate": 0.0002842567816913708, | |
| "loss": 2.1723, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.10498046875, | |
| "learning_rate": 0.0002830830026003773, | |
| "loss": 2.1585, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 0.0002819057546994135, | |
| "loss": 2.1758, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.10302734375, | |
| "learning_rate": 0.00028072508713992007, | |
| "loss": 2.1637, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 0.00027954104921611194, | |
| "loss": 2.1478, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.10302734375, | |
| "learning_rate": 0.00027835369036292087, | |
| "loss": 2.1585, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.10302734375, | |
| "learning_rate": 0.000277163060153931, | |
| "loss": 2.1481, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.1015625, | |
| "learning_rate": 0.0002759692082993095, | |
| "loss": 2.1636, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.10205078125, | |
| "learning_rate": 0.00027477218464373076, | |
| "loss": 2.1594, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.1044921875, | |
| "learning_rate": 0.0002735720391642956, | |
| "loss": 2.153, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.1025390625, | |
| "learning_rate": 0.0002723688219684443, | |
| "loss": 2.1666, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.1015625, | |
| "learning_rate": 0.00027116258329186514, | |
| "loss": 2.1728, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 0.0002699533734963964, | |
| "loss": 2.158, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.10546875, | |
| "learning_rate": 0.000268741243067924, | |
| "loss": 2.1384, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 0.00026752624261427375, | |
| "loss": 2.1691, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.10302734375, | |
| "learning_rate": 0.0002663084228630982, | |
| "loss": 2.1531, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 0.0002650878346597586, | |
| "loss": 2.1576, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 0.0002638645289652025, | |
| "loss": 2.1653, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.109375, | |
| "learning_rate": 0.0002626385568538358, | |
| "loss": 2.15, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 0.0002614099695113901, | |
| "loss": 2.1738, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 0.00026017881823278607, | |
| "loss": 2.1608, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 0.00025894515441999156, | |
| "loss": 2.1914, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 0.00025770902957987556, | |
| "loss": 2.1659, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.10302734375, | |
| "learning_rate": 0.0002564704953220578, | |
| "loss": 2.1719, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 0.0002552296033567541, | |
| "loss": 2.1676, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 0.0002539864054926169, | |
| "loss": 2.165, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.1025390625, | |
| "learning_rate": 0.0002527409536345728, | |
| "loss": 2.1561, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.1025390625, | |
| "learning_rate": 0.00025149329978165516, | |
| "loss": 2.1635, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.1044921875, | |
| "learning_rate": 0.0002502434960248331, | |
| "loss": 2.1513, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.10205078125, | |
| "learning_rate": 0.00024899159454483665, | |
| "loss": 2.1599, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.10302734375, | |
| "learning_rate": 0.0002477376476099784, | |
| "loss": 2.1674, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.1025390625, | |
| "learning_rate": 0.00024648170757397055, | |
| "loss": 2.14, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 0.00024522382687374, | |
| "loss": 2.1519, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.10595703125, | |
| "learning_rate": 0.0002439640580272384, | |
| "loss": 2.1622, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.1015625, | |
| "learning_rate": 0.0002427024536312496, | |
| "loss": 2.1535, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.10302734375, | |
| "learning_rate": 0.00024143906635919383, | |
| "loss": 2.1383, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.1025390625, | |
| "learning_rate": 0.00024017394895892838, | |
| "loss": 2.1282, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.10595703125, | |
| "learning_rate": 0.00023890715425054545, | |
| "loss": 2.1554, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 0.0002376387351241666, | |
| "loss": 2.157, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 0.00023636874453773475, | |
| "loss": 2.1471, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.10546875, | |
| "learning_rate": 0.00023509723551480325, | |
| "loss": 2.1383, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.1541337966918945, | |
| "eval_runtime": 188.5948, | |
| "eval_samples_per_second": 25.785, | |
| "eval_steps_per_second": 3.224, | |
| "step": 1351 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 0.00023382426114232162, | |
| "loss": 2.1347, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.10595703125, | |
| "learning_rate": 0.00023254987456841956, | |
| "loss": 2.1403, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.107421875, | |
| "learning_rate": 0.00023127412900018782, | |
| "loss": 2.1468, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.10498046875, | |
| "learning_rate": 0.00022999707770145653, | |
| "loss": 2.1267, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.1064453125, | |
| "learning_rate": 0.0002287187739905717, | |
| "loss": 2.1406, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.10595703125, | |
| "learning_rate": 0.00022743927123816899, | |
| "loss": 2.1304, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 0.10791015625, | |
| "learning_rate": 0.00022615862286494537, | |
| "loss": 2.1199, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 0.109375, | |
| "learning_rate": 0.00022487688233942862, | |
| "loss": 2.1216, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 0.10693359375, | |
| "learning_rate": 0.00022359410317574548, | |
| "loss": 2.1188, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.10791015625, | |
| "learning_rate": 0.00022231033893138668, | |
| "loss": 2.123, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.1083984375, | |
| "learning_rate": 0.0002210256432049714, | |
| "loss": 2.1401, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.10888671875, | |
| "learning_rate": 0.0002197400696340091, | |
| "loss": 2.1212, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 0.10791015625, | |
| "learning_rate": 0.00021845367189266042, | |
| "loss": 2.1297, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 0.107421875, | |
| "learning_rate": 0.0002171665036894959, | |
| "loss": 2.1237, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 0.00021587861876525377, | |
| "loss": 2.1281, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 0.1083984375, | |
| "learning_rate": 0.00021459007089059625, | |
| "loss": 2.0898, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 0.10888671875, | |
| "learning_rate": 0.0002133009138638645, | |
| "loss": 2.1212, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 0.10693359375, | |
| "learning_rate": 0.00021201120150883234, | |
| "loss": 2.1285, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 0.10888671875, | |
| "learning_rate": 0.00021072098767245932, | |
| "loss": 2.1302, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 0.10888671875, | |
| "learning_rate": 0.00020943032622264238, | |
| "loss": 2.1079, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 0.00020813927104596666, | |
| "loss": 2.1353, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.10986328125, | |
| "learning_rate": 0.0002068478760454562, | |
| "loss": 2.1402, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.1064453125, | |
| "learning_rate": 0.0002055561951383227, | |
| "loss": 2.1276, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 0.109375, | |
| "learning_rate": 0.00020426428225371496, | |
| "loss": 2.1241, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 0.10791015625, | |
| "learning_rate": 0.00020297219133046714, | |
| "loss": 2.1255, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 0.0002016799763148467, | |
| "loss": 2.1191, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.1083984375, | |
| "learning_rate": 0.00020038769115830198, | |
| "loss": 2.1308, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 0.00019909538981521002, | |
| "loss": 2.1422, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 0.1083984375, | |
| "learning_rate": 0.00019780312624062326, | |
| "loss": 2.1358, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 0.11083984375, | |
| "learning_rate": 0.00019651095438801775, | |
| "loss": 2.1292, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 0.11083984375, | |
| "learning_rate": 0.00019521892820703975, | |
| "loss": 2.1308, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 0.0001939271016412536, | |
| "loss": 2.1339, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 0.00019263552862588948, | |
| "loss": 2.126, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 0.1103515625, | |
| "learning_rate": 0.00019134426308559162, | |
| "loss": 2.1325, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 0.10986328125, | |
| "learning_rate": 0.00019005335893216665, | |
| "loss": 2.1302, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 0.0001887628700623332, | |
| "loss": 2.125, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 0.11083984375, | |
| "learning_rate": 0.000187472850355471, | |
| "loss": 2.1217, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 0.1103515625, | |
| "learning_rate": 0.00018618335367137195, | |
| "loss": 2.147, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 0.0001848944338479909, | |
| "loss": 2.1437, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.10986328125, | |
| "learning_rate": 0.00018360614469919835, | |
| "loss": 2.117, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.1162109375, | |
| "learning_rate": 0.0001823185400125333, | |
| "loss": 2.1119, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.11083984375, | |
| "learning_rate": 0.00018103167354695756, | |
| "loss": 2.1242, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 0.00017974559903061172, | |
| "loss": 2.115, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 0.00017846037015857127, | |
| "loss": 2.1332, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 0.00017717604059060518, | |
| "loss": 2.1049, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 0.109375, | |
| "learning_rate": 0.0001758926639489354, | |
| "loss": 2.1246, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 0.00017461029381599832, | |
| "loss": 2.1229, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 0.11083984375, | |
| "learning_rate": 0.00017332898373220707, | |
| "loss": 2.1136, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 0.1103515625, | |
| "learning_rate": 0.000172048787193717, | |
| "loss": 2.1204, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 0.00017076975765019134, | |
| "loss": 2.1239, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 0.1103515625, | |
| "learning_rate": 0.00016949194850257002, | |
| "loss": 2.1308, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 0.00016821541310084006, | |
| "loss": 2.1219, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 0.00016694020474180814, | |
| "loss": 2.1299, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 0.00016566637666687547, | |
| "loss": 2.127, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 0.00016439398205981472, | |
| "loss": 2.1255, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 0.00016312307404454967, | |
| "loss": 2.1149, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 0.0001618537056829373, | |
| "loss": 2.1105, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 0.00016058592997255215, | |
| "loss": 2.1477, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 0.00015931979984447385, | |
| "loss": 2.13, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 0.00015805536816107703, | |
| "loss": 2.1188, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 0.00015679268771382428, | |
| "loss": 2.1194, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 0.00015553181122106234, | |
| "loss": 2.1449, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 0.00015427279132582055, | |
| "loss": 2.122, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 0.00015301568059361323, | |
| "loss": 2.117, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.1103515625, | |
| "learning_rate": 0.00015176053151024502, | |
| "loss": 2.1244, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 0.0001505073964796194, | |
| "loss": 2.1206, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 0.000149256327821551, | |
| "loss": 2.1345, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 0.00014800737776958097, | |
| "loss": 2.1259, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 0.00014676059846879615, | |
| "loss": 2.125, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 0.00014551604197365222, | |
| "loss": 2.124, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 0.0001442737602458001, | |
| "loss": 2.1176, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 0.0001430338051519165, | |
| "loss": 2.1208, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 0.1162109375, | |
| "learning_rate": 0.00014179622846153872, | |
| "loss": 2.1309, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 0.0001405610818449027, | |
| "loss": 2.1126, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 0.0001393284168707864, | |
| "loss": 2.1075, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 0.0001380982850043562, | |
| "loss": 2.1107, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 0.00013687073760501828, | |
| "loss": 2.1447, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 0.00013564582592427444, | |
| "loss": 2.1151, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 0.11669921875, | |
| "learning_rate": 0.00013442360110358224, | |
| "loss": 2.1212, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 0.00013320411417221974, | |
| "loss": 2.1249, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 0.0001319874160451551, | |
| "loss": 2.1384, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 0.00013077355752092061, | |
| "loss": 2.1235, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 0.00012956258927949196, | |
| "loss": 2.1457, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 0.00012835456188017226, | |
| "loss": 2.1232, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 0.00012714952575948102, | |
| "loss": 2.1135, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 0.00012594753122904858, | |
| "loss": 2.1268, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 0.00012474862847351527, | |
| "loss": 2.1204, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 0.00012355286754843654, | |
| "loss": 2.1317, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 0.00012236029837819264, | |
| "loss": 2.1233, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 0.00012117097075390449, | |
| "loss": 2.1264, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 0.00011998493433135474, | |
| "loss": 2.1372, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 0.00011880223862891462, | |
| "loss": 2.1296, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 0.00011762293302547649, | |
| "loss": 2.1252, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 0.00011644706675839232, | |
| "loss": 2.1107, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 0.00011527468892141785, | |
| "loss": 2.1107, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.1162109375, | |
| "learning_rate": 0.00011410584846266266, | |
| "loss": 2.1142, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 0.0001129405941825471, | |
| "loss": 2.1143, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 0.11572265625, | |
| "learning_rate": 0.00011177897473176413, | |
| "loss": 2.1129, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 0.00011062103860924873, | |
| "loss": 2.1186, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 0.00010946683416015264, | |
| "loss": 2.1213, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 0.00010831640957382601, | |
| "loss": 2.1192, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 0.00010716981288180526, | |
| "loss": 2.1373, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 0.11572265625, | |
| "learning_rate": 0.000106027091955808, | |
| "loss": 2.1416, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 0.00010488829450573434, | |
| "loss": 2.1381, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 0.0001037534680776744, | |
| "loss": 2.1228, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 0.00010262266005192399, | |
| "loss": 2.1245, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 0.00010149591764100586, | |
| "loss": 2.1272, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 0.00010037328788769884, | |
| "loss": 2.1265, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 9.925481766307341e-05, | |
| "loss": 2.1364, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.1162109375, | |
| "learning_rate": 9.814055366453523e-05, | |
| "loss": 2.1193, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 9.703054241387499e-05, | |
| "loss": 2.1262, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 9.592483025532652e-05, | |
| "loss": 2.144, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 9.48234633536316e-05, | |
| "loss": 2.1294, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 9.372648769211258e-05, | |
| "loss": 2.1113, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 9.263394907075244e-05, | |
| "loss": 2.1247, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 9.154589310428288e-05, | |
| "loss": 2.1161, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 9.046236522027938e-05, | |
| "loss": 2.13, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 8.938341065726508e-05, | |
| "loss": 2.1218, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 8.830907446282162e-05, | |
| "loss": 2.1196, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 8.723940149170853e-05, | |
| "loss": 2.132, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.1171875, | |
| "learning_rate": 8.617443640399056e-05, | |
| "loss": 2.12, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 8.51142236631727e-05, | |
| "loss": 2.1178, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 8.405880753434434e-05, | |
| "loss": 2.1231, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 8.300823208233062e-05, | |
| "loss": 2.1259, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 8.196254116985303e-05, | |
| "loss": 2.1061, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 8.0921778455698e-05, | |
| "loss": 2.1365, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 7.988598739289408e-05, | |
| "loss": 2.1183, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 7.885521122689753e-05, | |
| "loss": 2.1135, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 7.782949299378724e-05, | |
| "loss": 2.1114, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 7.68088755184673e-05, | |
| "loss": 2.1381, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.1162109375, | |
| "learning_rate": 7.579340141287965e-05, | |
| "loss": 2.1033, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 7.478311307422456e-05, | |
| "loss": 2.1169, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 7.377805268319076e-05, | |
| "loss": 2.1155, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.1162109375, | |
| "learning_rate": 7.27782622021939e-05, | |
| "loss": 2.1077, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 7.178378337362519e-05, | |
| "loss": 2.1162, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 7.079465771810828e-05, | |
| "loss": 2.1294, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 0.11083984375, | |
| "learning_rate": 6.981092653276547e-05, | |
| "loss": 2.1204, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 6.88326308894941e-05, | |
| "loss": 2.1197, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 6.78598116332513e-05, | |
| "loss": 2.1183, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 6.68925093803489e-05, | |
| "loss": 2.1259, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 6.593076451675734e-05, | |
| "loss": 2.1203, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 6.497461719642003e-05, | |
| "loss": 2.1051, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 6.402410733957627e-05, | |
| "loss": 2.1393, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 0.11572265625, | |
| "learning_rate": 6.307927463109504e-05, | |
| "loss": 2.131, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 6.214015851881788e-05, | |
| "loss": 2.1138, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 6.120679821191193e-05, | |
| "loss": 2.1361, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 6.027923267923279e-05, | |
| "loss": 2.1242, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 5.9357500647697786e-05, | |
| "loss": 2.1396, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 5.8441640600668924e-05, | |
| "loss": 2.1037, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 5.75316907763459e-05, | |
| "loss": 2.1335, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 5.6627689166170364e-05, | |
| "loss": 2.1064, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 5.5729673513238814e-05, | |
| "loss": 2.1039, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 5.4837681310727464e-05, | |
| "loss": 2.1427, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 5.395174980032645e-05, | |
| "loss": 2.1386, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 5.307191597068531e-05, | |
| "loss": 2.124, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 5.2198216555868206e-05, | |
| "loss": 2.1275, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 5.133068803382073e-05, | |
| "loss": 2.1182, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 5.046936662484658e-05, | |
| "loss": 2.1199, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 0.11572265625, | |
| "learning_rate": 4.9614288290095467e-05, | |
| "loss": 2.1488, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 4.8765488730061485e-05, | |
| "loss": 2.1171, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 4.792300338309288e-05, | |
| "loss": 2.1305, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 4.70868674239124e-05, | |
| "loss": 2.1236, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 4.625711576214831e-05, | |
| "loss": 2.1145, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 0.1162109375, | |
| "learning_rate": 4.543378304087746e-05, | |
| "loss": 2.1362, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 0.11572265625, | |
| "learning_rate": 4.461690363517857e-05, | |
| "loss": 2.1243, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 4.380651165069707e-05, | |
| "loss": 2.1126, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 4.3002640922221084e-05, | |
| "loss": 2.1095, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 4.220532501226902e-05, | |
| "loss": 2.1351, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 4.141459720968792e-05, | |
| "loss": 2.1225, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 4.0630490528264196e-05, | |
| "loss": 2.1168, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 3.985303770534459e-05, | |
| "loss": 2.1339, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 3.908227120046983e-05, | |
| "loss": 2.1018, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 3.831822319401916e-05, | |
| "loss": 2.1128, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 3.756092558586694e-05, | |
| "loss": 2.1124, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 3.681040999405079e-05, | |
| "loss": 2.1334, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 3.606670775345116e-05, | |
| "loss": 2.1247, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 3.532984991448356e-05, | |
| "loss": 2.1211, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 0.1162109375, | |
| "learning_rate": 3.459986724180189e-05, | |
| "loss": 2.1455, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 3.387679021301406e-05, | |
| "loss": 2.1305, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 3.316064901740934e-05, | |
| "loss": 2.1421, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 3.245147355469822e-05, | |
| "loss": 2.1317, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 3.174929343376374e-05, | |
| "loss": 2.1087, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 3.105413797142576e-05, | |
| "loss": 2.1278, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 3.0366036191216274e-05, | |
| "loss": 2.1105, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 2.9685016822168287e-05, | |
| "loss": 2.1215, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 2.9011108297615908e-05, | |
| "loss": 2.1419, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 2.834433875400755e-05, | |
| "loss": 2.1307, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 2.768473602973083e-05, | |
| "loss": 2.1006, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 2.7032327663950675e-05, | |
| "loss": 2.1167, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 2.6387140895459284e-05, | |
| "loss": 2.1192, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 2.5749202661538972e-05, | |
| "loss": 2.1193, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 2.511853959683752e-05, | |
| "loss": 2.119, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 2.4495178032255918e-05, | |
| "loss": 2.1334, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 2.3879143993849474e-05, | |
| "loss": 2.1016, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 2.3270463201740665e-05, | |
| "loss": 2.1279, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 2.2669161069045863e-05, | |
| "loss": 2.1051, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 2.2075262700813747e-05, | |
| "loss": 2.1403, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 2.148879289297756e-05, | |
| "loss": 2.1177, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 0.1181640625, | |
| "learning_rate": 2.0909776131319548e-05, | |
| "loss": 2.1135, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 2.0338236590448978e-05, | |
| "loss": 2.1412, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 1.9774198132792353e-05, | |
| "loss": 2.1297, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 1.9217684307597806e-05, | |
| "loss": 2.1109, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 1.866871834995112e-05, | |
| "loss": 2.1333, | |
| "step": 2365 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 1.8127323179806234e-05, | |
| "loss": 2.1076, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 1.7593521401027967e-05, | |
| "loss": 2.1145, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 1.7067335300448506e-05, | |
| "loss": 2.1214, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 0.11572265625, | |
| "learning_rate": 1.654878684693677e-05, | |
| "loss": 2.1398, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 1.6037897690481075e-05, | |
| "loss": 2.0997, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 0.11572265625, | |
| "learning_rate": 1.5534689161285643e-05, | |
| "loss": 2.1291, | |
| "step": 2395 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 1.5039182268879504e-05, | |
| "loss": 2.1358, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 1.4551397701239721e-05, | |
| "loss": 2.1091, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 1.4071355823927424e-05, | |
| "loss": 2.1098, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 1.3599076679237676e-05, | |
| "loss": 2.1321, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 1.3134579985362517e-05, | |
| "loss": 2.1004, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 1.2677885135567979e-05, | |
| "loss": 2.1175, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 1.2229011197384021e-05, | |
| "loss": 2.1235, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 1.1787976911808773e-05, | |
| "loss": 2.1341, | |
| "step": 2435 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 1.1354800692525835e-05, | |
| "loss": 2.1321, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 1.092950062513567e-05, | |
| "loss": 2.0928, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 1.0512094466400402e-05, | |
| "loss": 2.1177, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 1.0102599643502508e-05, | |
| "loss": 2.1335, | |
| "step": 2455 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 0.11572265625, | |
| "learning_rate": 9.70103325331717e-06, | |
| "loss": 2.111, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 9.307412061698428e-06, | |
| "loss": 2.1119, | |
| "step": 2465 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 8.92175250277929e-06, | |
| "loss": 2.1151, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 8.54407067828551e-06, | |
| "loss": 2.1228, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 8.174382356863365e-06, | |
| "loss": 2.1171, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 7.812702973421182e-06, | |
| "loss": 2.1102, | |
| "step": 2485 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 0.11572265625, | |
| "learning_rate": 7.4590476284852165e-06, | |
| "loss": 2.1014, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 7.113431087568745e-06, | |
| "loss": 2.1109, | |
| "step": 2495 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 6.775867780555989e-06, | |
| "loss": 2.1115, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 6.446371801099371e-06, | |
| "loss": 2.1103, | |
| "step": 2505 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 6.124956906031276e-06, | |
| "loss": 2.1203, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 0.1181640625, | |
| "learning_rate": 5.811636514789598e-06, | |
| "loss": 2.1227, | |
| "step": 2515 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 5.506423708857455e-06, | |
| "loss": 2.1251, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 5.209331231217052e-06, | |
| "loss": 2.1305, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 4.920371485817632e-06, | |
| "loss": 2.1124, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 4.639556537057677e-06, | |
| "loss": 2.11, | |
| "step": 2535 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 4.3668981092810365e-06, | |
| "loss": 2.1314, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 4.102407586287571e-06, | |
| "loss": 2.1307, | |
| "step": 2545 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 3.846096010857791e-06, | |
| "loss": 2.1042, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 3.5979740842918995e-06, | |
| "loss": 2.1353, | |
| "step": 2555 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 3.3580521659628106e-06, | |
| "loss": 2.1164, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.11572265625, | |
| "learning_rate": 3.126340272883899e-06, | |
| "loss": 2.1268, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 2.902848079290488e-06, | |
| "loss": 2.119, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 2.687584916236241e-06, | |
| "loss": 2.1196, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 0.1162109375, | |
| "learning_rate": 2.4805597712032946e-06, | |
| "loss": 2.1185, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 2.281781287727247e-06, | |
| "loss": 2.1372, | |
| "step": 2585 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 2.091257765036092e-06, | |
| "loss": 2.1092, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 1.908997157703918e-06, | |
| "loss": 2.1357, | |
| "step": 2595 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 1.7350070753186176e-06, | |
| "loss": 2.1145, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 1.5692947821642324e-06, | |
| "loss": 2.114, | |
| "step": 2605 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 1.4118671969177265e-06, | |
| "loss": 2.1391, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 1.2627308923600644e-06, | |
| "loss": 2.1302, | |
| "step": 2615 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 1.1218920951018064e-06, | |
| "loss": 2.135, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 9.893566853230951e-07, | |
| "loss": 2.1385, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 8.651301965282077e-07, | |
| "loss": 2.1302, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.11669921875, | |
| "learning_rate": 7.492178153145402e-07, | |
| "loss": 2.125, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.11572265625, | |
| "learning_rate": 6.416243811559808e-07, | |
| "loss": 2.1311, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 5.423543862009384e-07, | |
| "loss": 2.1087, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 4.5141197508475894e-07, | |
| "loss": 2.105, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 3.6880094475664204e-07, | |
| "loss": 2.1256, | |
| "step": 2655 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 2.945247443212118e-07, | |
| "loss": 2.1204, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 2.2858647489441e-07, | |
| "loss": 2.1264, | |
| "step": 2665 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 1.7098888947404412e-07, | |
| "loss": 2.1144, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 1.217343928249237e-07, | |
| "loss": 2.0979, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 8.082504137836288e-08, | |
| "loss": 2.136, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 4.8262543146382345e-08, | |
| "loss": 2.1116, | |
| "step": 2685 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 2.404825765034424e-08, | |
| "loss": 2.1157, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 0.12255859375, | |
| "learning_rate": 8.183195864264192e-09, | |
| "loss": 2.1259, | |
| "step": 2695 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 6.680201725117741e-10, | |
| "loss": 2.1135, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 2.14662766456604, | |
| "eval_runtime": 188.6485, | |
| "eval_samples_per_second": 25.778, | |
| "eval_steps_per_second": 3.223, | |
| "step": 2702 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 2702, | |
| "total_flos": 6.137045069867254e+17, | |
| "train_loss": 2.1601854102158, | |
| "train_runtime": 21984.2259, | |
| "train_samples_per_second": 7.864, | |
| "train_steps_per_second": 0.123 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 2702, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 100, | |
| "total_flos": 6.137045069867254e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |