| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 784, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0012755102040816326, | |
| "grad_norm": 226.4242401123047, | |
| "learning_rate": 3.7974683544303794e-06, | |
| "loss": 48.9032, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006377551020408163, | |
| "grad_norm": 208.31333923339844, | |
| "learning_rate": 1.89873417721519e-05, | |
| "loss": 48.5429, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.012755102040816327, | |
| "grad_norm": 113.34983825683594, | |
| "learning_rate": 3.79746835443038e-05, | |
| "loss": 39.6906, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01913265306122449, | |
| "grad_norm": 62.39444351196289, | |
| "learning_rate": 5.696202531645569e-05, | |
| "loss": 31.8222, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.025510204081632654, | |
| "grad_norm": 16.312618255615234, | |
| "learning_rate": 7.59493670886076e-05, | |
| "loss": 24.5018, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03188775510204082, | |
| "grad_norm": 11.158844947814941, | |
| "learning_rate": 9.493670886075949e-05, | |
| "loss": 21.551, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.03826530612244898, | |
| "grad_norm": 7.093152046203613, | |
| "learning_rate": 0.00011392405063291139, | |
| "loss": 20.5881, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.044642857142857144, | |
| "grad_norm": 4.10010290145874, | |
| "learning_rate": 0.00013291139240506327, | |
| "loss": 19.0191, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.05102040816326531, | |
| "grad_norm": 5.54206657409668, | |
| "learning_rate": 0.0001518987341772152, | |
| "loss": 18.1907, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.05739795918367347, | |
| "grad_norm": 10.086193084716797, | |
| "learning_rate": 0.00017088607594936708, | |
| "loss": 17.5312, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.06377551020408163, | |
| "grad_norm": 19.117507934570312, | |
| "learning_rate": 0.00018987341772151899, | |
| "loss": 15.1473, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07015306122448979, | |
| "grad_norm": 30.6151180267334, | |
| "learning_rate": 0.0002088607594936709, | |
| "loss": 11.3497, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.07653061224489796, | |
| "grad_norm": 21.005615234375, | |
| "learning_rate": 0.00022784810126582277, | |
| "loss": 6.1773, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.08290816326530612, | |
| "grad_norm": 11.67768383026123, | |
| "learning_rate": 0.0002468354430379747, | |
| "loss": 3.0207, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.08928571428571429, | |
| "grad_norm": 6.371170520782471, | |
| "learning_rate": 0.00026582278481012653, | |
| "loss": 2.478, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.09566326530612244, | |
| "grad_norm": 2.4198906421661377, | |
| "learning_rate": 0.00028481012658227844, | |
| "loss": 2.1055, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.10204081632653061, | |
| "grad_norm": 1.157726764678955, | |
| "learning_rate": 0.00029999851070045734, | |
| "loss": 1.8032, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.10841836734693877, | |
| "grad_norm": 1.53150475025177, | |
| "learning_rate": 0.00029994638832161753, | |
| "loss": 1.6379, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.11479591836734694, | |
| "grad_norm": 4.24367618560791, | |
| "learning_rate": 0.0002998198305365145, | |
| "loss": 1.5582, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1211734693877551, | |
| "grad_norm": 0.8776500821113586, | |
| "learning_rate": 0.000299618900170137, | |
| "loss": 1.5178, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.12755102040816327, | |
| "grad_norm": 1.7082425355911255, | |
| "learning_rate": 0.00029934369696702433, | |
| "loss": 1.4593, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.13392857142857142, | |
| "grad_norm": 2.545419931411743, | |
| "learning_rate": 0.0002989943575417525, | |
| "loss": 1.4022, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.14030612244897958, | |
| "grad_norm": 0.9073415994644165, | |
| "learning_rate": 0.0002985710553111161, | |
| "loss": 1.391, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.14668367346938777, | |
| "grad_norm": 0.7289413809776306, | |
| "learning_rate": 0.0002980740004080424, | |
| "loss": 1.3611, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.15306122448979592, | |
| "grad_norm": 0.6918051838874817, | |
| "learning_rate": 0.0002975034395772784, | |
| "loss": 1.3474, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.15943877551020408, | |
| "grad_norm": 1.0997295379638672, | |
| "learning_rate": 0.00029685965605290337, | |
| "loss": 1.3001, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.16581632653061223, | |
| "grad_norm": 4.529347896575928, | |
| "learning_rate": 0.0002961429694177276, | |
| "loss": 1.3039, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.17219387755102042, | |
| "grad_norm": 2.3405444622039795, | |
| "learning_rate": 0.0002953537354446474, | |
| "loss": 1.2888, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.17857142857142858, | |
| "grad_norm": 2.904238224029541, | |
| "learning_rate": 0.00029449234592003436, | |
| "loss": 1.2796, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.18494897959183673, | |
| "grad_norm": 1.2537798881530762, | |
| "learning_rate": 0.0002935592284492476, | |
| "loss": 1.2941, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.1913265306122449, | |
| "grad_norm": 1.00596284866333, | |
| "learning_rate": 0.0002925548462443644, | |
| "loss": 1.2674, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.19770408163265307, | |
| "grad_norm": 1.229261040687561, | |
| "learning_rate": 0.0002914796978942358, | |
| "loss": 1.2493, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.20408163265306123, | |
| "grad_norm": 5.567780494689941, | |
| "learning_rate": 0.0002903343171169803, | |
| "loss": 1.2494, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.21045918367346939, | |
| "grad_norm": 2.096792221069336, | |
| "learning_rate": 0.00028911927249503864, | |
| "loss": 1.2534, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.21683673469387754, | |
| "grad_norm": 2.6790685653686523, | |
| "learning_rate": 0.0002878351671929224, | |
| "loss": 1.2544, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.22321428571428573, | |
| "grad_norm": 1.817202091217041, | |
| "learning_rate": 0.0002864826386577947, | |
| "loss": 1.2283, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.22959183673469388, | |
| "grad_norm": 1.452048420906067, | |
| "learning_rate": 0.00028506235830303274, | |
| "loss": 1.2416, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.23596938775510204, | |
| "grad_norm": 2.810974359512329, | |
| "learning_rate": 0.00028357503117492944, | |
| "loss": 1.2204, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.2423469387755102, | |
| "grad_norm": 0.8295219540596008, | |
| "learning_rate": 0.0002820213956026987, | |
| "loss": 1.2267, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.24872448979591838, | |
| "grad_norm": 1.347506046295166, | |
| "learning_rate": 0.0002804022228319591, | |
| "loss": 1.2054, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.25510204081632654, | |
| "grad_norm": 1.5803207159042358, | |
| "learning_rate": 0.000278718316641877, | |
| "loss": 1.1688, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2614795918367347, | |
| "grad_norm": 0.6454965472221375, | |
| "learning_rate": 0.00027697051294615964, | |
| "loss": 1.2095, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.26785714285714285, | |
| "grad_norm": 0.8256592154502869, | |
| "learning_rate": 0.0002751596793780968, | |
| "loss": 1.1808, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.274234693877551, | |
| "grad_norm": 1.8195074796676636, | |
| "learning_rate": 0.00027328671485985515, | |
| "loss": 1.1963, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.28061224489795916, | |
| "grad_norm": 1.3873289823532104, | |
| "learning_rate": 0.0002713525491562421, | |
| "loss": 1.1795, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.2869897959183674, | |
| "grad_norm": 2.0349373817443848, | |
| "learning_rate": 0.00026935814241315736, | |
| "loss": 1.1802, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.29336734693877553, | |
| "grad_norm": 1.6414597034454346, | |
| "learning_rate": 0.00026730448468096377, | |
| "loss": 1.173, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.2997448979591837, | |
| "grad_norm": 1.2609707117080688, | |
| "learning_rate": 0.00026519259542301317, | |
| "loss": 1.1561, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.30612244897959184, | |
| "grad_norm": 4.516489028930664, | |
| "learning_rate": 0.0002630235230095708, | |
| "loss": 1.1573, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3125, | |
| "grad_norm": 1.8223940134048462, | |
| "learning_rate": 0.0002607983441973903, | |
| "loss": 1.1619, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.31887755102040816, | |
| "grad_norm": 1.7611159086227417, | |
| "learning_rate": 0.0002585181635951971, | |
| "loss": 1.1654, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.3252551020408163, | |
| "grad_norm": 1.7671295404434204, | |
| "learning_rate": 0.0002561841131153459, | |
| "loss": 1.1701, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.33163265306122447, | |
| "grad_norm": 1.1895421743392944, | |
| "learning_rate": 0.00025379735141192397, | |
| "loss": 1.1517, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.3380102040816326, | |
| "grad_norm": 2.26123046875, | |
| "learning_rate": 0.0002513590633055798, | |
| "loss": 1.1304, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.34438775510204084, | |
| "grad_norm": 2.0483951568603516, | |
| "learning_rate": 0.00024887045919536207, | |
| "loss": 1.1403, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.350765306122449, | |
| "grad_norm": 1.611738681793213, | |
| "learning_rate": 0.00024633277445786143, | |
| "loss": 1.1499, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 2.4385955333709717, | |
| "learning_rate": 0.00024374726883395273, | |
| "loss": 1.1424, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.3635204081632653, | |
| "grad_norm": 1.5102245807647705, | |
| "learning_rate": 0.00024111522580344265, | |
| "loss": 1.1135, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.36989795918367346, | |
| "grad_norm": 0.8719022870063782, | |
| "learning_rate": 0.00023843795194793336, | |
| "loss": 1.1236, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3762755102040816, | |
| "grad_norm": 0.9534187316894531, | |
| "learning_rate": 0.0002357167763022175, | |
| "loss": 1.1253, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.3826530612244898, | |
| "grad_norm": 1.8618340492248535, | |
| "learning_rate": 0.00023295304969452748, | |
| "loss": 1.1296, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.38903061224489793, | |
| "grad_norm": 0.7363711595535278, | |
| "learning_rate": 0.00023014814407596628, | |
| "loss": 1.1298, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.39540816326530615, | |
| "grad_norm": 8.200729370117188, | |
| "learning_rate": 0.00022730345183945248, | |
| "loss": 1.1155, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.4017857142857143, | |
| "grad_norm": 1.8119416236877441, | |
| "learning_rate": 0.0002244203851285173, | |
| "loss": 1.1161, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.40816326530612246, | |
| "grad_norm": 3.4057934284210205, | |
| "learning_rate": 0.000221500375136298, | |
| "loss": 1.1257, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.4145408163265306, | |
| "grad_norm": 0.9975801706314087, | |
| "learning_rate": 0.00021854487139507398, | |
| "loss": 1.1356, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.42091836734693877, | |
| "grad_norm": 1.9435389041900635, | |
| "learning_rate": 0.00021555534105669976, | |
| "loss": 1.1142, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.4272959183673469, | |
| "grad_norm": 1.204506516456604, | |
| "learning_rate": 0.00021253326816429076, | |
| "loss": 1.1174, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.4336734693877551, | |
| "grad_norm": 2.5877513885498047, | |
| "learning_rate": 0.0002094801529155244, | |
| "loss": 1.0873, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.44005102040816324, | |
| "grad_norm": 0.7226624488830566, | |
| "learning_rate": 0.00020639751091792188, | |
| "loss": 1.1018, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.44642857142857145, | |
| "grad_norm": 1.762408971786499, | |
| "learning_rate": 0.00020328687243648, | |
| "loss": 1.1056, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.4528061224489796, | |
| "grad_norm": 4.760770320892334, | |
| "learning_rate": 0.00020014978163402725, | |
| "loss": 1.0828, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.45918367346938777, | |
| "grad_norm": 0.723813533782959, | |
| "learning_rate": 0.0001969877958046804, | |
| "loss": 1.1015, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.4655612244897959, | |
| "grad_norm": 1.1869585514068604, | |
| "learning_rate": 0.0001938024846007828, | |
| "loss": 1.0955, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.4719387755102041, | |
| "grad_norm": 0.7096392512321472, | |
| "learning_rate": 0.00019059542925370783, | |
| "loss": 1.0949, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.47831632653061223, | |
| "grad_norm": 2.3162615299224854, | |
| "learning_rate": 0.0001873682217889143, | |
| "loss": 1.0943, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.4846938775510204, | |
| "grad_norm": 0.7215905785560608, | |
| "learning_rate": 0.0001841224642356436, | |
| "loss": 1.0957, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.49107142857142855, | |
| "grad_norm": 1.3136285543441772, | |
| "learning_rate": 0.0001808597678316509, | |
| "loss": 1.0865, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.49744897959183676, | |
| "grad_norm": 2.323671817779541, | |
| "learning_rate": 0.00017758175222336496, | |
| "loss": 1.0752, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5038265306122449, | |
| "grad_norm": 1.1084301471710205, | |
| "learning_rate": 0.00017429004466187377, | |
| "loss": 1.0892, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.5102040816326531, | |
| "grad_norm": 2.052607297897339, | |
| "learning_rate": 0.00017098627919513528, | |
| "loss": 1.1036, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5165816326530612, | |
| "grad_norm": 1.0028424263000488, | |
| "learning_rate": 0.00016767209585681394, | |
| "loss": 1.0913, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.5229591836734694, | |
| "grad_norm": 0.9611196517944336, | |
| "learning_rate": 0.00016434913985214597, | |
| "loss": 1.0825, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.5293367346938775, | |
| "grad_norm": 2.9150185585021973, | |
| "learning_rate": 0.0001610190607412373, | |
| "loss": 1.085, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.5357142857142857, | |
| "grad_norm": 2.879117250442505, | |
| "learning_rate": 0.00015768351162019975, | |
| "loss": 1.0889, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.5420918367346939, | |
| "grad_norm": 1.0958130359649658, | |
| "learning_rate": 0.0001543441483005318, | |
| "loss": 1.0833, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.548469387755102, | |
| "grad_norm": 0.9718190431594849, | |
| "learning_rate": 0.00015100262848715133, | |
| "loss": 1.0813, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.5548469387755102, | |
| "grad_norm": 1.2847260236740112, | |
| "learning_rate": 0.000147660610955489, | |
| "loss": 1.0684, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.5612244897959183, | |
| "grad_norm": 1.039225459098816, | |
| "learning_rate": 0.00014431975472804936, | |
| "loss": 1.0797, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.5676020408163265, | |
| "grad_norm": 1.2738494873046875, | |
| "learning_rate": 0.00014098171825084995, | |
| "loss": 1.0839, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.5739795918367347, | |
| "grad_norm": 1.06371009349823, | |
| "learning_rate": 0.00013764815857014636, | |
| "loss": 1.0649, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.5803571428571429, | |
| "grad_norm": 1.394904375076294, | |
| "learning_rate": 0.000134320730509852, | |
| "loss": 1.0798, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.5867346938775511, | |
| "grad_norm": 0.6322318315505981, | |
| "learning_rate": 0.00013100108585006146, | |
| "loss": 1.0702, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.5931122448979592, | |
| "grad_norm": 1.5734529495239258, | |
| "learning_rate": 0.0001276908725070845, | |
| "loss": 1.0571, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.5994897959183674, | |
| "grad_norm": 1.5619932413101196, | |
| "learning_rate": 0.00012439173371539818, | |
| "loss": 1.0714, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.6058673469387755, | |
| "grad_norm": 1.8657833337783813, | |
| "learning_rate": 0.00012110530721192319, | |
| "loss": 1.0727, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.6122448979591837, | |
| "grad_norm": 2.461625337600708, | |
| "learning_rate": 0.00011783322442302915, | |
| "loss": 1.0472, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.6186224489795918, | |
| "grad_norm": 1.1668888330459595, | |
| "learning_rate": 0.00011457710965467266, | |
| "loss": 1.0482, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 1.11457359790802, | |
| "learning_rate": 0.00011133857928606992, | |
| "loss": 1.0618, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.6313775510204082, | |
| "grad_norm": 1.9202951192855835, | |
| "learning_rate": 0.00010811924096730416, | |
| "loss": 1.0494, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.6377551020408163, | |
| "grad_norm": 0.7689357995986938, | |
| "learning_rate": 0.00010492069282126685, | |
| "loss": 1.0443, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6441326530612245, | |
| "grad_norm": 2.5137813091278076, | |
| "learning_rate": 0.00010174452265032769, | |
| "loss": 1.0689, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.6505102040816326, | |
| "grad_norm": 0.6587682962417603, | |
| "learning_rate": 9.859230714812812e-05, | |
| "loss": 1.0482, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.6568877551020408, | |
| "grad_norm": 1.251999020576477, | |
| "learning_rate": 9.546561111688958e-05, | |
| "loss": 1.0439, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.6632653061224489, | |
| "grad_norm": 0.9404116272926331, | |
| "learning_rate": 9.236598669062443e-05, | |
| "loss": 1.042, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.6696428571428571, | |
| "grad_norm": 2.4428858757019043, | |
| "learning_rate": 8.929497256463586e-05, | |
| "loss": 1.0399, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.6760204081632653, | |
| "grad_norm": 0.9080251455307007, | |
| "learning_rate": 8.625409323168858e-05, | |
| "loss": 1.0655, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.6823979591836735, | |
| "grad_norm": 0.9034568071365356, | |
| "learning_rate": 8.32448582252305e-05, | |
| "loss": 1.0342, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.6887755102040817, | |
| "grad_norm": 0.7829791307449341, | |
| "learning_rate": 8.026876137003956e-05, | |
| "loss": 1.0468, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.6951530612244898, | |
| "grad_norm": 0.9003449082374573, | |
| "learning_rate": 7.732728004066927e-05, | |
| "loss": 1.0516, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.701530612244898, | |
| "grad_norm": 1.5128637552261353, | |
| "learning_rate": 7.442187442805997e-05, | |
| "loss": 1.0486, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.7079081632653061, | |
| "grad_norm": 0.843595027923584, | |
| "learning_rate": 7.155398681468041e-05, | |
| "loss": 1.035, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 1.120501160621643, | |
| "learning_rate": 6.872504085855947e-05, | |
| "loss": 1.0468, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.7206632653061225, | |
| "grad_norm": 0.6691803932189941, | |
| "learning_rate": 6.593644088656277e-05, | |
| "loss": 1.032, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.7270408163265306, | |
| "grad_norm": 1.066981315612793, | |
| "learning_rate": 6.318957119726636e-05, | |
| "loss": 1.0366, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.7334183673469388, | |
| "grad_norm": 0.9314639568328857, | |
| "learning_rate": 6.048579537377189e-05, | |
| "loss": 1.0258, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.7397959183673469, | |
| "grad_norm": 2.091320753097534, | |
| "learning_rate": 5.782645560680571e-05, | |
| "loss": 1.0328, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.7461734693877551, | |
| "grad_norm": 0.6987316012382507, | |
| "learning_rate": 5.5212872028437564e-05, | |
| "loss": 1.0588, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.7525510204081632, | |
| "grad_norm": 0.9516999125480652, | |
| "learning_rate": 5.264634205674884e-05, | |
| "loss": 1.036, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.7589285714285714, | |
| "grad_norm": 0.9213479161262512, | |
| "learning_rate": 5.012813975177708e-05, | |
| "loss": 1.0275, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.7653061224489796, | |
| "grad_norm": 1.1120542287826538, | |
| "learning_rate": 4.765951518305524e-05, | |
| "loss": 1.0286, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.7716836734693877, | |
| "grad_norm": 1.146470308303833, | |
| "learning_rate": 4.524169380906042e-05, | |
| "loss": 1.0306, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.7780612244897959, | |
| "grad_norm": 0.7507238388061523, | |
| "learning_rate": 4.287587586887991e-05, | |
| "loss": 1.0358, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.7844387755102041, | |
| "grad_norm": 0.7958881258964539, | |
| "learning_rate": 4.056323578639599e-05, | |
| "loss": 1.0522, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.7908163265306123, | |
| "grad_norm": 1.0853825807571411, | |
| "learning_rate": 3.83049215872865e-05, | |
| "loss": 1.0318, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.7971938775510204, | |
| "grad_norm": 0.6897711157798767, | |
| "learning_rate": 3.610205432912914e-05, | |
| "loss": 1.0253, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.8035714285714286, | |
| "grad_norm": 1.08404541015625, | |
| "learning_rate": 3.395572754489349e-05, | |
| "loss": 1.0536, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.8099489795918368, | |
| "grad_norm": 0.9480798244476318, | |
| "learning_rate": 3.1867006700096625e-05, | |
| "loss": 1.0363, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.8163265306122449, | |
| "grad_norm": 1.013339877128601, | |
| "learning_rate": 2.9836928663891517e-05, | |
| "loss": 1.0286, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.8227040816326531, | |
| "grad_norm": 0.8233193755149841, | |
| "learning_rate": 2.7866501194351253e-05, | |
| "loss": 1.047, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.8290816326530612, | |
| "grad_norm": 0.8080177903175354, | |
| "learning_rate": 2.595670243820461e-05, | |
| "loss": 1.0383, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.8354591836734694, | |
| "grad_norm": 0.6761140823364258, | |
| "learning_rate": 2.4108480445270346e-05, | |
| "loss": 1.032, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.8418367346938775, | |
| "grad_norm": 1.0132781267166138, | |
| "learning_rate": 2.2322752697833045e-05, | |
| "loss": 1.0381, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.8482142857142857, | |
| "grad_norm": 2.3679418563842773, | |
| "learning_rate": 2.0600405655192166e-05, | |
| "loss": 1.035, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.8545918367346939, | |
| "grad_norm": 1.0122100114822388, | |
| "learning_rate": 1.8942294313611974e-05, | |
| "loss": 1.0523, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.860969387755102, | |
| "grad_norm": 0.797031819820404, | |
| "learning_rate": 1.7349241781889654e-05, | |
| "loss": 1.0225, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.8673469387755102, | |
| "grad_norm": 0.6979250907897949, | |
| "learning_rate": 1.5822038872753166e-05, | |
| "loss": 1.023, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.8737244897959183, | |
| "grad_norm": 0.5788841843605042, | |
| "learning_rate": 1.4361443710291149e-05, | |
| "loss": 1.0388, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.8801020408163265, | |
| "grad_norm": 0.7020506858825684, | |
| "learning_rate": 1.2968181353609852e-05, | |
| "loss": 1.0148, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.8864795918367347, | |
| "grad_norm": 0.6345797181129456, | |
| "learning_rate": 1.1642943436904111e-05, | |
| "loss": 1.0269, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.8928571428571429, | |
| "grad_norm": 0.7500460147857666, | |
| "learning_rate": 1.038638782612073e-05, | |
| "loss": 1.0271, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.8992346938775511, | |
| "grad_norm": 0.6654278635978699, | |
| "learning_rate": 9.199138292385006e-06, | |
| "loss": 1.0241, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.9056122448979592, | |
| "grad_norm": 0.6786679029464722, | |
| "learning_rate": 8.081784202352404e-06, | |
| "loss": 1.0301, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.9119897959183674, | |
| "grad_norm": 0.712835967540741, | |
| "learning_rate": 7.034880225638878e-06, | |
| "loss": 1.0355, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.9183673469387755, | |
| "grad_norm": 0.691426694393158, | |
| "learning_rate": 6.058946059475595e-06, | |
| "loss": 1.0374, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.9247448979591837, | |
| "grad_norm": 0.6497547626495361, | |
| "learning_rate": 5.154466170724031e-06, | |
| "loss": 1.03, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.9311224489795918, | |
| "grad_norm": 0.7832013964653015, | |
| "learning_rate": 4.32188955538022e-06, | |
| "loss": 1.0222, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.9375, | |
| "grad_norm": 1.1313321590423584, | |
| "learning_rate": 3.5616295156870754e-06, | |
| "loss": 1.0005, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.9438775510204082, | |
| "grad_norm": 0.7574149966239929, | |
| "learning_rate": 2.874063454965558e-06, | |
| "loss": 1.0293, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.9502551020408163, | |
| "grad_norm": 0.6553170680999756, | |
| "learning_rate": 2.2595326902666412e-06, | |
| "loss": 1.027, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.9566326530612245, | |
| "grad_norm": 0.7936610579490662, | |
| "learning_rate": 1.7183422829368409e-06, | |
| "loss": 1.0301, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.9630102040816326, | |
| "grad_norm": 0.6665956974029541, | |
| "learning_rate": 1.2507608871817566e-06, | |
| "loss": 1.0299, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.9693877551020408, | |
| "grad_norm": 0.6638586521148682, | |
| "learning_rate": 8.570206167024807e-07, | |
| "loss": 1.0108, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.9757653061224489, | |
| "grad_norm": 0.6911438703536987, | |
| "learning_rate": 5.37316929471121e-07, | |
| "loss": 1.0219, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.9821428571428571, | |
| "grad_norm": 0.6120124459266663, | |
| "learning_rate": 2.9180853070303044e-07, | |
| "loss": 1.0279, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.9885204081632653, | |
| "grad_norm": 0.6324964165687561, | |
| "learning_rate": 1.2061729407324416e-07, | |
| "loss": 1.0428, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.9948979591836735, | |
| "grad_norm": 0.5991183519363403, | |
| "learning_rate": 2.3828201216774533e-08, | |
| "loss": 1.01, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.388667106628418, | |
| "eval_runtime": 1.6706, | |
| "eval_samples_per_second": 5.986, | |
| "eval_steps_per_second": 1.197, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 784, | |
| "total_flos": 5.976388653306348e+17, | |
| "train_loss": 2.8103741881798725, | |
| "train_runtime": 4265.0133, | |
| "train_samples_per_second": 2.94, | |
| "train_steps_per_second": 0.184 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 784, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.976388653306348e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |