| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9878213802435725, | |
| "eval_steps": 500, | |
| "global_step": 276, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010825439783491205, | |
| "grad_norm": 5.807920175447218, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 0.8567, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02165087956698241, | |
| "grad_norm": 5.811623230162752, | |
| "learning_rate": 1.1428571428571429e-05, | |
| "loss": 0.8604, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03247631935047361, | |
| "grad_norm": 4.253584558151055, | |
| "learning_rate": 1.7142857142857142e-05, | |
| "loss": 0.8173, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.04330175913396482, | |
| "grad_norm": 4.494099066603187, | |
| "learning_rate": 2.2857142857142858e-05, | |
| "loss": 0.8299, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.05412719891745602, | |
| "grad_norm": 4.509661541720279, | |
| "learning_rate": 2.8571428571428574e-05, | |
| "loss": 0.7651, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06495263870094722, | |
| "grad_norm": 3.951348852730613, | |
| "learning_rate": 3.4285714285714284e-05, | |
| "loss": 0.7664, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07577807848443843, | |
| "grad_norm": 2.62641811103538, | |
| "learning_rate": 4e-05, | |
| "loss": 0.7174, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.08660351826792964, | |
| "grad_norm": 2.5211356734881742, | |
| "learning_rate": 4.5714285714285716e-05, | |
| "loss": 0.7056, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.09742895805142084, | |
| "grad_norm": 2.2539730546361585, | |
| "learning_rate": 5.1428571428571436e-05, | |
| "loss": 0.6663, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.10825439783491204, | |
| "grad_norm": 2.856988089650331, | |
| "learning_rate": 5.714285714285715e-05, | |
| "loss": 0.6637, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.11907983761840325, | |
| "grad_norm": 1.8247182509862854, | |
| "learning_rate": 6.285714285714286e-05, | |
| "loss": 0.6377, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.12990527740189445, | |
| "grad_norm": 2.5491735246742953, | |
| "learning_rate": 6.857142857142857e-05, | |
| "loss": 0.6408, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.14073071718538566, | |
| "grad_norm": 2.464416812125527, | |
| "learning_rate": 7.42857142857143e-05, | |
| "loss": 0.6248, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.15155615696887687, | |
| "grad_norm": 1.8774381374385238, | |
| "learning_rate": 8e-05, | |
| "loss": 0.632, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.16238159675236807, | |
| "grad_norm": 2.186597255470316, | |
| "learning_rate": 8.571428571428571e-05, | |
| "loss": 0.6201, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.17320703653585928, | |
| "grad_norm": 2.367228905837931, | |
| "learning_rate": 9.142857142857143e-05, | |
| "loss": 0.6084, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.18403247631935046, | |
| "grad_norm": 1.961289391436032, | |
| "learning_rate": 9.714285714285714e-05, | |
| "loss": 0.5974, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.19485791610284167, | |
| "grad_norm": 1.6062347832492823, | |
| "learning_rate": 0.00010285714285714287, | |
| "loss": 0.5851, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.20568335588633288, | |
| "grad_norm": 2.5785754238732115, | |
| "learning_rate": 0.00010857142857142859, | |
| "loss": 0.5899, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2165087956698241, | |
| "grad_norm": 2.4973397443279985, | |
| "learning_rate": 0.0001142857142857143, | |
| "loss": 0.59, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2273342354533153, | |
| "grad_norm": 1.7261400197953252, | |
| "learning_rate": 0.00012000000000000002, | |
| "loss": 0.5817, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2381596752368065, | |
| "grad_norm": 2.8676046622672464, | |
| "learning_rate": 0.00012571428571428572, | |
| "loss": 0.5856, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2489851150202977, | |
| "grad_norm": 1.1982706010037665, | |
| "learning_rate": 0.00013142857142857143, | |
| "loss": 0.5718, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2598105548037889, | |
| "grad_norm": 2.6334206150571764, | |
| "learning_rate": 0.00013714285714285713, | |
| "loss": 0.6143, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2706359945872801, | |
| "grad_norm": 4.1727861440371905, | |
| "learning_rate": 0.00014285714285714287, | |
| "loss": 0.5957, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2814614343707713, | |
| "grad_norm": 1.726889837548816, | |
| "learning_rate": 0.0001485714285714286, | |
| "loss": 0.5778, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2922868741542625, | |
| "grad_norm": 1.9491289608648266, | |
| "learning_rate": 0.0001542857142857143, | |
| "loss": 0.5872, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.30311231393775373, | |
| "grad_norm": 2.3922212205589464, | |
| "learning_rate": 0.00016, | |
| "loss": 0.5856, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.31393775372124494, | |
| "grad_norm": 1.7812309864931277, | |
| "learning_rate": 0.0001599935812484263, | |
| "loss": 0.5814, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.32476319350473615, | |
| "grad_norm": 1.4649634825431583, | |
| "learning_rate": 0.00015997432602371444, | |
| "loss": 0.5663, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.33558863328822736, | |
| "grad_norm": 4.015544716712766, | |
| "learning_rate": 0.00015994223741572707, | |
| "loss": 0.5853, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.34641407307171856, | |
| "grad_norm": 11.281220320234869, | |
| "learning_rate": 0.00015989732057368423, | |
| "loss": 0.6967, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3572395128552097, | |
| "grad_norm": 7690.086558155647, | |
| "learning_rate": 0.00015983958270533721, | |
| "loss": 4.6049, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3680649526387009, | |
| "grad_norm": 112.72682447862128, | |
| "learning_rate": 0.0001597690330758118, | |
| "loss": 2.8115, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.37889039242219213, | |
| "grad_norm": 4.025212340328087, | |
| "learning_rate": 0.0001596856830061217, | |
| "loss": 0.7115, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.38971583220568334, | |
| "grad_norm": 2.4843639071881074, | |
| "learning_rate": 0.0001595895458713516, | |
| "loss": 0.6295, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.40054127198917455, | |
| "grad_norm": 1.7939204684367258, | |
| "learning_rate": 0.00015948063709851125, | |
| "loss": 0.6154, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.41136671177266576, | |
| "grad_norm": 2.459210039805435, | |
| "learning_rate": 0.00015935897416405944, | |
| "loss": 0.6166, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.42219215155615697, | |
| "grad_norm": 1.11090456282376, | |
| "learning_rate": 0.00015922457659110005, | |
| "loss": 0.6007, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.4330175913396482, | |
| "grad_norm": 84.92016735622583, | |
| "learning_rate": 0.00015907746594624893, | |
| "loss": 0.767, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4438430311231394, | |
| "grad_norm": 8.660965154683476, | |
| "learning_rate": 0.0001589176658361731, | |
| "loss": 0.8712, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.4546684709066306, | |
| "grad_norm": 3.4936052679031158, | |
| "learning_rate": 0.00015874520190380274, | |
| "loss": 0.7434, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4654939106901218, | |
| "grad_norm": 1.7045480694962383, | |
| "learning_rate": 0.00015856010182421635, | |
| "loss": 0.6308, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.476319350473613, | |
| "grad_norm": 14.30049648802306, | |
| "learning_rate": 0.00015836239530019958, | |
| "loss": 0.6371, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.4871447902571042, | |
| "grad_norm": 4.053341450365789, | |
| "learning_rate": 0.00015815211405747896, | |
| "loss": 0.6844, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4979702300405954, | |
| "grad_norm": 1.7771803961776749, | |
| "learning_rate": 0.00015792929183963098, | |
| "loss": 0.6299, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.5087956698240866, | |
| "grad_norm": 1.8665193904465571, | |
| "learning_rate": 0.00015769396440266714, | |
| "loss": 0.6056, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.5196211096075778, | |
| "grad_norm": 1.9789963866244402, | |
| "learning_rate": 0.00015744616950929635, | |
| "loss": 0.6017, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.530446549391069, | |
| "grad_norm": 2.0461657503603115, | |
| "learning_rate": 0.00015718594692286516, | |
| "loss": 0.5884, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.5412719891745602, | |
| "grad_norm": 1.216761740673657, | |
| "learning_rate": 0.000156913338400977, | |
| "loss": 0.5744, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5520974289580515, | |
| "grad_norm": 2.0980055824287938, | |
| "learning_rate": 0.0001566283876887913, | |
| "loss": 0.5937, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.5629228687415426, | |
| "grad_norm": 1.168756354199333, | |
| "learning_rate": 0.00015633114051200392, | |
| "loss": 0.5665, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5737483085250338, | |
| "grad_norm": 1.486001286642052, | |
| "learning_rate": 0.0001560216445695094, | |
| "loss": 0.561, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.584573748308525, | |
| "grad_norm": 1.385583557087864, | |
| "learning_rate": 0.00015569994952574698, | |
| "loss": 0.5679, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5953991880920162, | |
| "grad_norm": 1.1211177929259362, | |
| "learning_rate": 0.0001553661070027308, | |
| "loss": 0.5481, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.6062246278755075, | |
| "grad_norm": 1.338185754695372, | |
| "learning_rate": 0.00015502017057176644, | |
| "loss": 0.551, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.6170500676589986, | |
| "grad_norm": 0.9631327673448306, | |
| "learning_rate": 0.00015466219574485415, | |
| "loss": 0.5439, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.6278755074424899, | |
| "grad_norm": 1.296708217652255, | |
| "learning_rate": 0.000154292239965781, | |
| "loss": 0.5444, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.638700947225981, | |
| "grad_norm": 1.01308925336574, | |
| "learning_rate": 0.00015391036260090294, | |
| "loss": 0.529, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6495263870094723, | |
| "grad_norm": 1.1113699538858748, | |
| "learning_rate": 0.00015351662492961847, | |
| "loss": 0.5353, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6603518267929634, | |
| "grad_norm": 1.229664456155683, | |
| "learning_rate": 0.00015311109013453495, | |
| "loss": 0.5391, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.6711772665764547, | |
| "grad_norm": 0.7696540318430937, | |
| "learning_rate": 0.00015269382329133, | |
| "loss": 0.5236, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6820027063599459, | |
| "grad_norm": 1.0452145349007034, | |
| "learning_rate": 0.0001522648913583088, | |
| "loss": 0.5333, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6928281461434371, | |
| "grad_norm": 1.0076590725145298, | |
| "learning_rate": 0.00015182436316565936, | |
| "loss": 0.5239, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.7036535859269283, | |
| "grad_norm": 1.2124343568222422, | |
| "learning_rate": 0.00015137230940440738, | |
| "loss": 0.5236, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.7144790257104194, | |
| "grad_norm": 1.2765073901945772, | |
| "learning_rate": 0.00015090880261507268, | |
| "loss": 0.5307, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.7253044654939107, | |
| "grad_norm": 0.627724711812233, | |
| "learning_rate": 0.0001504339171760286, | |
| "loss": 0.5162, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.7361299052774019, | |
| "grad_norm": 0.648664232932664, | |
| "learning_rate": 0.00014994772929156658, | |
| "loss": 0.525, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.7469553450608931, | |
| "grad_norm": 0.853676037408326, | |
| "learning_rate": 0.00014945031697966788, | |
| "loss": 0.5147, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7577807848443843, | |
| "grad_norm": 0.7593724841438597, | |
| "learning_rate": 0.00014894176005948397, | |
| "loss": 0.4967, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7686062246278755, | |
| "grad_norm": 0.5321774839143892, | |
| "learning_rate": 0.00014842214013852817, | |
| "loss": 0.4977, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.7794316644113667, | |
| "grad_norm": 0.6449157547696196, | |
| "learning_rate": 0.00014789154059958009, | |
| "loss": 0.51, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.790257104194858, | |
| "grad_norm": 0.527820608461116, | |
| "learning_rate": 0.00014735004658730536, | |
| "loss": 0.5134, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.8010825439783491, | |
| "grad_norm": 0.7495799991859693, | |
| "learning_rate": 0.0001467977449945926, | |
| "loss": 0.5122, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.8119079837618404, | |
| "grad_norm": 1.038331830644402, | |
| "learning_rate": 0.00014623472444860974, | |
| "loss": 0.5134, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.8227334235453315, | |
| "grad_norm": 1.4042907578050945, | |
| "learning_rate": 0.00014566107529658212, | |
| "loss": 0.5083, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.8335588633288228, | |
| "grad_norm": 0.7986676831500767, | |
| "learning_rate": 0.00014507688959129473, | |
| "loss": 0.4975, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.8443843031123139, | |
| "grad_norm": 1.0273680785889077, | |
| "learning_rate": 0.00014448226107632042, | |
| "loss": 0.5049, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.8552097428958051, | |
| "grad_norm": 1.064561031536024, | |
| "learning_rate": 0.0001438772851709771, | |
| "loss": 0.5053, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.8660351826792964, | |
| "grad_norm": 1.0710244196033907, | |
| "learning_rate": 0.00014326205895501588, | |
| "loss": 0.4977, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8768606224627875, | |
| "grad_norm": 0.8475945967427466, | |
| "learning_rate": 0.00014263668115304285, | |
| "loss": 0.5053, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.8876860622462788, | |
| "grad_norm": 0.6798466426655683, | |
| "learning_rate": 0.0001420012521186767, | |
| "loss": 0.4898, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.8985115020297699, | |
| "grad_norm": 0.720780663324741, | |
| "learning_rate": 0.00014135587381844535, | |
| "loss": 0.4932, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.9093369418132612, | |
| "grad_norm": 0.6169019560016734, | |
| "learning_rate": 0.0001407006498154233, | |
| "loss": 0.4916, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.9201623815967523, | |
| "grad_norm": 0.48074951462504467, | |
| "learning_rate": 0.00014003568525261303, | |
| "loss": 0.4864, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.9309878213802436, | |
| "grad_norm": 0.5675353244905835, | |
| "learning_rate": 0.00013936108683607287, | |
| "loss": 0.4902, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.9418132611637348, | |
| "grad_norm": 0.5142478444309492, | |
| "learning_rate": 0.0001386769628177941, | |
| "loss": 0.4839, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.952638700947226, | |
| "grad_norm": 0.4166415283022488, | |
| "learning_rate": 0.0001379834229783296, | |
| "loss": 0.4701, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.9634641407307172, | |
| "grad_norm": 0.49915561344121656, | |
| "learning_rate": 0.00013728057860917784, | |
| "loss": 0.4842, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.9742895805142084, | |
| "grad_norm": 0.6649725562299833, | |
| "learning_rate": 0.00013656854249492382, | |
| "loss": 0.4837, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.9851150202976996, | |
| "grad_norm": 1.1671140064989096, | |
| "learning_rate": 0.00013584742889514073, | |
| "loss": 0.486, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.9959404600811907, | |
| "grad_norm": 1.368072672593922, | |
| "learning_rate": 0.00013511735352605495, | |
| "loss": 0.4884, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.006765899864682, | |
| "grad_norm": 1.0201145301236891, | |
| "learning_rate": 0.00013437843354197705, | |
| "loss": 0.7512, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.0175913396481733, | |
| "grad_norm": 1.8000094250512286, | |
| "learning_rate": 0.0001336307875165023, | |
| "loss": 0.4768, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.0284167794316643, | |
| "grad_norm": 0.5472576337231695, | |
| "learning_rate": 0.00013287453542348328, | |
| "loss": 0.4613, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.0392422192151556, | |
| "grad_norm": 1.1828047434659983, | |
| "learning_rate": 0.0001321097986177778, | |
| "loss": 0.4647, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0500676589986468, | |
| "grad_norm": 0.9813473966873533, | |
| "learning_rate": 0.0001313366998157752, | |
| "loss": 0.4528, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.060893098782138, | |
| "grad_norm": 0.7578769026059632, | |
| "learning_rate": 0.00013055536307570432, | |
| "loss": 0.4603, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.0717185385656292, | |
| "grad_norm": 0.5725242057599156, | |
| "learning_rate": 0.0001297659137777259, | |
| "loss": 0.4527, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.0825439783491204, | |
| "grad_norm": 0.575770129842184, | |
| "learning_rate": 0.00012896847860381305, | |
| "loss": 0.4501, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0933694181326117, | |
| "grad_norm": 0.6026831960506396, | |
| "learning_rate": 0.00012816318551742264, | |
| "loss": 0.448, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.104194857916103, | |
| "grad_norm": 0.4305530355604545, | |
| "learning_rate": 0.0001273501637429614, | |
| "loss": 0.4462, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.115020297699594, | |
| "grad_norm": 0.4454666291048703, | |
| "learning_rate": 0.0001265295437450491, | |
| "loss": 0.4511, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.1258457374830853, | |
| "grad_norm": 0.4678487064825796, | |
| "learning_rate": 0.0001257014572075834, | |
| "loss": 0.4411, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.1366711772665765, | |
| "grad_norm": 0.37950754053596286, | |
| "learning_rate": 0.00012486603701260837, | |
| "loss": 0.4434, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.1474966170500678, | |
| "grad_norm": 0.43449589812234013, | |
| "learning_rate": 0.0001240234172189913, | |
| "loss": 0.4477, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.1583220568335588, | |
| "grad_norm": 0.43939536219651565, | |
| "learning_rate": 0.0001231737330409104, | |
| "loss": 0.4431, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.16914749661705, | |
| "grad_norm": 0.3882093042316694, | |
| "learning_rate": 0.000122317120826157, | |
| "loss": 0.4364, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.1799729364005414, | |
| "grad_norm": 0.4351026316208757, | |
| "learning_rate": 0.00012145371803425616, | |
| "loss": 0.4431, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.1907983761840324, | |
| "grad_norm": 0.5501651171341266, | |
| "learning_rate": 0.0001205836632144086, | |
| "loss": 0.447, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.2016238159675237, | |
| "grad_norm": 0.4174336327490401, | |
| "learning_rate": 0.00011970709598325793, | |
| "loss": 0.4452, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.212449255751015, | |
| "grad_norm": 0.29631678479891477, | |
| "learning_rate": 0.0001188241570024865, | |
| "loss": 0.4398, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.2232746955345062, | |
| "grad_norm": 0.339334632156724, | |
| "learning_rate": 0.00011793498795624362, | |
| "loss": 0.4349, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.2341001353179972, | |
| "grad_norm": 0.30928533747901593, | |
| "learning_rate": 0.00011703973152840967, | |
| "loss": 0.4421, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.2449255751014885, | |
| "grad_norm": 0.39017965923706066, | |
| "learning_rate": 0.00011613853137969977, | |
| "loss": 0.4501, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.2557510148849798, | |
| "grad_norm": 0.6145727467731127, | |
| "learning_rate": 0.00011523153212461075, | |
| "loss": 0.4406, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.266576454668471, | |
| "grad_norm": 1.0187083339327228, | |
| "learning_rate": 0.00011431887930821501, | |
| "loss": 0.4402, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.277401894451962, | |
| "grad_norm": 1.2911461014630221, | |
| "learning_rate": 0.00011340071938280507, | |
| "loss": 0.4424, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.2882273342354533, | |
| "grad_norm": 0.35131816146651274, | |
| "learning_rate": 0.00011247719968439257, | |
| "loss": 0.4418, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.2990527740189446, | |
| "grad_norm": 0.7870620433304094, | |
| "learning_rate": 0.00011154846840906549, | |
| "loss": 0.4359, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.3098782138024356, | |
| "grad_norm": 1.168118392838753, | |
| "learning_rate": 0.00011061467458920719, | |
| "loss": 0.4439, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.320703653585927, | |
| "grad_norm": 0.46362753411050434, | |
| "learning_rate": 0.00010967596806958144, | |
| "loss": 0.4355, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.3315290933694182, | |
| "grad_norm": 0.631341811837205, | |
| "learning_rate": 0.00010873249948328698, | |
| "loss": 0.4417, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.3423545331529092, | |
| "grad_norm": 0.8838449406798384, | |
| "learning_rate": 0.00010778442022758562, | |
| "loss": 0.439, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.3531799729364005, | |
| "grad_norm": 0.44202149476361463, | |
| "learning_rate": 0.00010683188243960779, | |
| "loss": 0.436, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.3640054127198917, | |
| "grad_norm": 0.5464115885263443, | |
| "learning_rate": 0.00010587503897193905, | |
| "loss": 0.4365, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.374830852503383, | |
| "grad_norm": 0.7244992161712334, | |
| "learning_rate": 0.00010491404336809233, | |
| "loss": 0.4292, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.3856562922868743, | |
| "grad_norm": 0.3629285827544841, | |
| "learning_rate": 0.00010394904983786867, | |
| "loss": 0.4298, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.3964817320703653, | |
| "grad_norm": 0.5706457571028368, | |
| "learning_rate": 0.0001029802132326116, | |
| "loss": 0.4306, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.4073071718538566, | |
| "grad_norm": 0.5219581868931918, | |
| "learning_rate": 0.00010200768902035825, | |
| "loss": 0.4345, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.4181326116373478, | |
| "grad_norm": 0.3128700720370998, | |
| "learning_rate": 0.00010103163326089159, | |
| "loss": 0.4352, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.4289580514208389, | |
| "grad_norm": 0.5860532868715919, | |
| "learning_rate": 0.00010005220258069766, | |
| "loss": 0.4379, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.4397834912043301, | |
| "grad_norm": 0.299205037891474, | |
| "learning_rate": 9.906955414783193e-05, | |
| "loss": 0.4286, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.4506089309878214, | |
| "grad_norm": 0.47146400660400606, | |
| "learning_rate": 9.80838456466988e-05, | |
| "loss": 0.4324, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.4614343707713127, | |
| "grad_norm": 0.5772068931303819, | |
| "learning_rate": 9.709523525274806e-05, | |
| "loss": 0.4326, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.472259810554804, | |
| "grad_norm": 0.3805133128090234, | |
| "learning_rate": 9.610388160709281e-05, | |
| "loss": 0.4309, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.483085250338295, | |
| "grad_norm": 0.4271236662578828, | |
| "learning_rate": 9.510994379105235e-05, | |
| "loss": 0.428, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.4939106901217862, | |
| "grad_norm": 0.3913062302891241, | |
| "learning_rate": 9.41135813006247e-05, | |
| "loss": 0.4336, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.5047361299052775, | |
| "grad_norm": 0.4959549867268792, | |
| "learning_rate": 9.311495402089241e-05, | |
| "loss": 0.4324, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.5155615696887685, | |
| "grad_norm": 0.47173284252778047, | |
| "learning_rate": 9.211422220036615e-05, | |
| "loss": 0.4322, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.5263870094722598, | |
| "grad_norm": 0.4218219584473093, | |
| "learning_rate": 9.111154642526954e-05, | |
| "loss": 0.4285, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.537212449255751, | |
| "grad_norm": 0.4956469938375666, | |
| "learning_rate": 9.010708759377035e-05, | |
| "loss": 0.4289, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.548037889039242, | |
| "grad_norm": 0.4107706925787467, | |
| "learning_rate": 8.910100689016121e-05, | |
| "loss": 0.4352, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.5588633288227334, | |
| "grad_norm": 0.4458931818501929, | |
| "learning_rate": 8.809346575899459e-05, | |
| "loss": 0.4268, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.5696887686062246, | |
| "grad_norm": 0.4721100077989787, | |
| "learning_rate": 8.708462587917602e-05, | |
| "loss": 0.4245, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.5805142083897157, | |
| "grad_norm": 0.3742345022103872, | |
| "learning_rate": 8.607464913801971e-05, | |
| "loss": 0.4244, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.5913396481732072, | |
| "grad_norm": 0.33389508877981666, | |
| "learning_rate": 8.506369760527053e-05, | |
| "loss": 0.4216, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.6021650879566982, | |
| "grad_norm": 0.2725464341226312, | |
| "learning_rate": 8.405193350709702e-05, | |
| "loss": 0.4396, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.6129905277401895, | |
| "grad_norm": 0.2561974791876615, | |
| "learning_rate": 8.303951920005916e-05, | |
| "loss": 0.4226, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.6238159675236807, | |
| "grad_norm": 0.2312801621695805, | |
| "learning_rate": 8.202661714505503e-05, | |
| "loss": 0.4296, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.6346414073071718, | |
| "grad_norm": 0.2539035606451629, | |
| "learning_rate": 8.101338988125122e-05, | |
| "loss": 0.4297, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.645466847090663, | |
| "grad_norm": 0.2795943345506495, | |
| "learning_rate": 8e-05, | |
| "loss": 0.4232, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.6562922868741543, | |
| "grad_norm": 0.305802751614139, | |
| "learning_rate": 7.898661011874881e-05, | |
| "loss": 0.435, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.6671177266576453, | |
| "grad_norm": 0.29309245201661993, | |
| "learning_rate": 7.797338285494499e-05, | |
| "loss": 0.4276, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.6779431664411368, | |
| "grad_norm": 0.329501466273318, | |
| "learning_rate": 7.696048079994087e-05, | |
| "loss": 0.421, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.6887686062246279, | |
| "grad_norm": 0.3142714545795239, | |
| "learning_rate": 7.594806649290299e-05, | |
| "loss": 0.4295, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.699594046008119, | |
| "grad_norm": 0.33315198074843505, | |
| "learning_rate": 7.49363023947295e-05, | |
| "loss": 0.4155, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.7104194857916104, | |
| "grad_norm": 0.34360494433039884, | |
| "learning_rate": 7.392535086198032e-05, | |
| "loss": 0.4269, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.7212449255751014, | |
| "grad_norm": 0.28517772111690337, | |
| "learning_rate": 7.2915374120824e-05, | |
| "loss": 0.4337, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.7320703653585927, | |
| "grad_norm": 0.31951375045950287, | |
| "learning_rate": 7.190653424100544e-05, | |
| "loss": 0.4267, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.742895805142084, | |
| "grad_norm": 0.34023285855014257, | |
| "learning_rate": 7.089899310983883e-05, | |
| "loss": 0.4256, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.753721244925575, | |
| "grad_norm": 0.2949489362370772, | |
| "learning_rate": 6.989291240622969e-05, | |
| "loss": 0.4296, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.7645466847090663, | |
| "grad_norm": 0.26157765123078286, | |
| "learning_rate": 6.88884535747305e-05, | |
| "loss": 0.4234, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.7753721244925575, | |
| "grad_norm": 0.2795081501666077, | |
| "learning_rate": 6.78857777996339e-05, | |
| "loss": 0.4205, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.7861975642760486, | |
| "grad_norm": 0.19331675781027863, | |
| "learning_rate": 6.68850459791076e-05, | |
| "loss": 0.4275, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.79702300405954, | |
| "grad_norm": 0.21357341104162766, | |
| "learning_rate": 6.588641869937535e-05, | |
| "loss": 0.4203, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.8078484438430311, | |
| "grad_norm": 0.2167571809187653, | |
| "learning_rate": 6.489005620894769e-05, | |
| "loss": 0.4254, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.8186738836265224, | |
| "grad_norm": 0.24601105327924508, | |
| "learning_rate": 6.389611839290719e-05, | |
| "loss": 0.4223, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.8294993234100136, | |
| "grad_norm": 0.3019894658869235, | |
| "learning_rate": 6.290476474725193e-05, | |
| "loss": 0.4307, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.8403247631935047, | |
| "grad_norm": 0.2435718708672937, | |
| "learning_rate": 6.191615435330121e-05, | |
| "loss": 0.4213, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.851150202976996, | |
| "grad_norm": 0.17781892789685452, | |
| "learning_rate": 6.093044585216807e-05, | |
| "loss": 0.4177, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.8619756427604872, | |
| "grad_norm": 0.19371420909940787, | |
| "learning_rate": 5.994779741930236e-05, | |
| "loss": 0.4186, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.8728010825439783, | |
| "grad_norm": 0.20962600669395126, | |
| "learning_rate": 5.896836673910842e-05, | |
| "loss": 0.4126, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.8836265223274695, | |
| "grad_norm": 0.26385576587957305, | |
| "learning_rate": 5.799231097964176e-05, | |
| "loss": 0.4251, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.8944519621109608, | |
| "grad_norm": 0.186706517180332, | |
| "learning_rate": 5.7019786767388426e-05, | |
| "loss": 0.4198, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.9052774018944518, | |
| "grad_norm": 0.20177197005120157, | |
| "learning_rate": 5.605095016213137e-05, | |
| "loss": 0.4142, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.9161028416779433, | |
| "grad_norm": 0.22218528258050108, | |
| "learning_rate": 5.50859566319077e-05, | |
| "loss": 0.4159, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.9269282814614344, | |
| "grad_norm": 0.25615151629013416, | |
| "learning_rate": 5.4124961028060956e-05, | |
| "loss": 0.4256, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.9377537212449256, | |
| "grad_norm": 0.23647599450178192, | |
| "learning_rate": 5.316811756039224e-05, | |
| "loss": 0.4194, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.9485791610284169, | |
| "grad_norm": 0.2353513525938554, | |
| "learning_rate": 5.221557977241439e-05, | |
| "loss": 0.4217, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.959404600811908, | |
| "grad_norm": 0.1652695325427714, | |
| "learning_rate": 5.126750051671304e-05, | |
| "loss": 0.4197, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.9702300405953992, | |
| "grad_norm": 0.23356895899115995, | |
| "learning_rate": 5.032403193041859e-05, | |
| "loss": 0.4159, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.9810554803788905, | |
| "grad_norm": 0.1988738455239412, | |
| "learning_rate": 4.9385325410792824e-05, | |
| "loss": 0.4247, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.9918809201623815, | |
| "grad_norm": 0.20607159435477782, | |
| "learning_rate": 4.8451531590934535e-05, | |
| "loss": 0.4254, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.002706359945873, | |
| "grad_norm": 0.33625717644811787, | |
| "learning_rate": 4.752280031560744e-05, | |
| "loss": 0.6634, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.013531799729364, | |
| "grad_norm": 0.3504909382927573, | |
| "learning_rate": 4.659928061719497e-05, | |
| "loss": 0.3805, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.024357239512855, | |
| "grad_norm": 0.2417890196803602, | |
| "learning_rate": 4.5681120691785035e-05, | |
| "loss": 0.3785, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.0351826792963466, | |
| "grad_norm": 0.3195492118572409, | |
| "learning_rate": 4.476846787538925e-05, | |
| "loss": 0.3796, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.0460081190798376, | |
| "grad_norm": 0.28392833843993076, | |
| "learning_rate": 4.386146862030023e-05, | |
| "loss": 0.3727, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.0568335588633286, | |
| "grad_norm": 0.253698837225003, | |
| "learning_rate": 4.2960268471590345e-05, | |
| "loss": 0.3774, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.06765899864682, | |
| "grad_norm": 0.2795134398632703, | |
| "learning_rate": 4.20650120437564e-05, | |
| "loss": 0.383, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.078484438430311, | |
| "grad_norm": 0.23404884349599717, | |
| "learning_rate": 4.1175842997513535e-05, | |
| "loss": 0.3862, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.089309878213802, | |
| "grad_norm": 0.2623191416209601, | |
| "learning_rate": 4.02929040167421e-05, | |
| "loss": 0.3784, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.1001353179972937, | |
| "grad_norm": 0.22037288605612937, | |
| "learning_rate": 3.9416336785591436e-05, | |
| "loss": 0.3705, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.1109607577807847, | |
| "grad_norm": 0.21836908464552335, | |
| "learning_rate": 3.854628196574387e-05, | |
| "loss": 0.3805, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.121786197564276, | |
| "grad_norm": 0.2557311206491114, | |
| "learning_rate": 3.768287917384303e-05, | |
| "loss": 0.378, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.1326116373477673, | |
| "grad_norm": 0.18179411448982316, | |
| "learning_rate": 3.682626695908964e-05, | |
| "loss": 0.3729, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.1434370771312583, | |
| "grad_norm": 0.2687321585152136, | |
| "learning_rate": 3.5976582781008695e-05, | |
| "loss": 0.3726, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.15426251691475, | |
| "grad_norm": 0.16496012357687265, | |
| "learning_rate": 3.5133962987391655e-05, | |
| "loss": 0.3791, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.165087956698241, | |
| "grad_norm": 0.201693140671645, | |
| "learning_rate": 3.42985427924166e-05, | |
| "loss": 0.3815, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.175913396481732, | |
| "grad_norm": 0.18160977177829019, | |
| "learning_rate": 3.347045625495088e-05, | |
| "loss": 0.3753, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.1867388362652234, | |
| "grad_norm": 0.17670354726721407, | |
| "learning_rate": 3.264983625703861e-05, | |
| "loss": 0.3731, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.1975642760487144, | |
| "grad_norm": 0.18581677779566666, | |
| "learning_rate": 3.1836814482577355e-05, | |
| "loss": 0.3803, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.208389715832206, | |
| "grad_norm": 0.17123638159588167, | |
| "learning_rate": 3.103152139618697e-05, | |
| "loss": 0.3779, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.219215155615697, | |
| "grad_norm": 0.20118541787477454, | |
| "learning_rate": 3.02340862222741e-05, | |
| "loss": 0.3764, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.230040595399188, | |
| "grad_norm": 0.1623232875171876, | |
| "learning_rate": 2.944463692429569e-05, | |
| "loss": 0.3772, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.2408660351826795, | |
| "grad_norm": 0.20276172261445644, | |
| "learning_rate": 2.866330018422481e-05, | |
| "loss": 0.3808, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.2516914749661705, | |
| "grad_norm": 0.16912420408288342, | |
| "learning_rate": 2.789020138222224e-05, | |
| "loss": 0.3714, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.2625169147496615, | |
| "grad_norm": 0.15940767575093126, | |
| "learning_rate": 2.7125464576516754e-05, | |
| "loss": 0.3762, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.273342354533153, | |
| "grad_norm": 0.15384987091038757, | |
| "learning_rate": 2.6369212483497714e-05, | |
| "loss": 0.372, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.284167794316644, | |
| "grad_norm": 0.13258610865988332, | |
| "learning_rate": 2.5621566458022972e-05, | |
| "loss": 0.3735, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.2949932341001356, | |
| "grad_norm": 0.15584495134484205, | |
| "learning_rate": 2.4882646473945072e-05, | |
| "loss": 0.3741, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.3058186738836266, | |
| "grad_norm": 0.13063314218715893, | |
| "learning_rate": 2.4152571104859284e-05, | |
| "loss": 0.3819, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.3166441136671176, | |
| "grad_norm": 0.15183169354187515, | |
| "learning_rate": 2.3431457505076205e-05, | |
| "loss": 0.3761, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.3274695534506087, | |
| "grad_norm": 0.15612284126375733, | |
| "learning_rate": 2.2719421390822173e-05, | |
| "loss": 0.3768, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.3382949932341, | |
| "grad_norm": 0.14653578705168283, | |
| "learning_rate": 2.201657702167042e-05, | |
| "loss": 0.3824, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.349120433017591, | |
| "grad_norm": 0.15086974405889184, | |
| "learning_rate": 2.1323037182205945e-05, | |
| "loss": 0.376, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.3599458728010827, | |
| "grad_norm": 0.1290378904007159, | |
| "learning_rate": 2.0638913163927156e-05, | |
| "loss": 0.3834, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.3707713125845737, | |
| "grad_norm": 0.13318235861496663, | |
| "learning_rate": 1.9964314747387005e-05, | |
| "loss": 0.3803, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.381596752368065, | |
| "grad_norm": 0.13330711346211635, | |
| "learning_rate": 1.9299350184576733e-05, | |
| "loss": 0.3691, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.3924221921515563, | |
| "grad_norm": 0.1279862282948995, | |
| "learning_rate": 1.8644126181554663e-05, | |
| "loss": 0.3757, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.4032476319350473, | |
| "grad_norm": 0.11677628138557435, | |
| "learning_rate": 1.79987478813233e-05, | |
| "loss": 0.3806, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.414073071718539, | |
| "grad_norm": 0.12619918683231618, | |
| "learning_rate": 1.7363318846957175e-05, | |
| "loss": 0.3803, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.42489851150203, | |
| "grad_norm": 0.1380173906206575, | |
| "learning_rate": 1.6737941044984127e-05, | |
| "loss": 0.3784, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.435723951285521, | |
| "grad_norm": 0.11282616881280712, | |
| "learning_rate": 1.612271482902292e-05, | |
| "loss": 0.3755, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.4465493910690124, | |
| "grad_norm": 0.13236199767020781, | |
| "learning_rate": 1.5517738923679616e-05, | |
| "loss": 0.379, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.4573748308525034, | |
| "grad_norm": 0.12480115347036357, | |
| "learning_rate": 1.492311040870531e-05, | |
| "loss": 0.3721, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.4682002706359945, | |
| "grad_norm": 0.11687086053065535, | |
| "learning_rate": 1.4338924703417916e-05, | |
| "loss": 0.3744, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.479025710419486, | |
| "grad_norm": 0.12453487485947455, | |
| "learning_rate": 1.3765275551390293e-05, | |
| "loss": 0.3718, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.489851150202977, | |
| "grad_norm": 0.12306865218763892, | |
| "learning_rate": 1.320225500540742e-05, | |
| "loss": 0.3724, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.500676589986468, | |
| "grad_norm": 0.10843672768522547, | |
| "learning_rate": 1.2649953412694634e-05, | |
| "loss": 0.3819, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.5115020297699595, | |
| "grad_norm": 0.11114457199744829, | |
| "learning_rate": 1.2108459400419919e-05, | |
| "loss": 0.3778, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.5223274695534506, | |
| "grad_norm": 0.1166727854863552, | |
| "learning_rate": 1.1577859861471833e-05, | |
| "loss": 0.3741, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.533152909336942, | |
| "grad_norm": 0.10791002146328788, | |
| "learning_rate": 1.1058239940516028e-05, | |
| "loss": 0.3773, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.543978349120433, | |
| "grad_norm": 0.10154469628935157, | |
| "learning_rate": 1.0549683020332125e-05, | |
| "loss": 0.3693, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.554803788903924, | |
| "grad_norm": 0.09966044957220346, | |
| "learning_rate": 1.0052270708433434e-05, | |
| "loss": 0.3696, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.565629228687415, | |
| "grad_norm": 0.09113621422642913, | |
| "learning_rate": 9.566082823971432e-06, | |
| "loss": 0.3759, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.5764546684709067, | |
| "grad_norm": 0.09597646217579717, | |
| "learning_rate": 9.09119738492735e-06, | |
| "loss": 0.375, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.5872801082543977, | |
| "grad_norm": 0.10190240288320754, | |
| "learning_rate": 8.627690595592653e-06, | |
| "loss": 0.3725, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.598105548037889, | |
| "grad_norm": 0.09510018971201405, | |
| "learning_rate": 8.175636834340674e-06, | |
| "loss": 0.3784, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.60893098782138, | |
| "grad_norm": 0.09762377519980237, | |
| "learning_rate": 7.735108641691202e-06, | |
| "loss": 0.3814, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.6197564276048713, | |
| "grad_norm": 0.09742308222941193, | |
| "learning_rate": 7.3061767086700075e-06, | |
| "loss": 0.381, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.6305818673883627, | |
| "grad_norm": 0.107032573161411, | |
| "learning_rate": 6.888909865465065e-06, | |
| "loss": 0.3724, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.641407307171854, | |
| "grad_norm": 0.10638149049322657, | |
| "learning_rate": 6.483375070381552e-06, | |
| "loss": 0.3715, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.6522327469553453, | |
| "grad_norm": 0.09269800135085551, | |
| "learning_rate": 6.0896373990970614e-06, | |
| "loss": 0.3721, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.6630581867388363, | |
| "grad_norm": 0.09447916428148233, | |
| "learning_rate": 5.707760034219032e-06, | |
| "loss": 0.3737, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.6738836265223274, | |
| "grad_norm": 0.10758712743393119, | |
| "learning_rate": 5.337804255145873e-06, | |
| "loss": 0.3779, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.6847090663058184, | |
| "grad_norm": 0.11068434930689386, | |
| "learning_rate": 4.979829428233576e-06, | |
| "loss": 0.3751, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.69553450608931, | |
| "grad_norm": 0.10188815124559192, | |
| "learning_rate": 4.63389299726921e-06, | |
| "loss": 0.3696, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.706359945872801, | |
| "grad_norm": 0.0846895734709407, | |
| "learning_rate": 4.3000504742530505e-06, | |
| "loss": 0.3809, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.7171853856562924, | |
| "grad_norm": 0.09582049879963728, | |
| "learning_rate": 3.978355430490614e-06, | |
| "loss": 0.3765, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.7280108254397835, | |
| "grad_norm": 0.08128736985428159, | |
| "learning_rate": 3.668859487996095e-06, | |
| "loss": 0.3799, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.7388362652232745, | |
| "grad_norm": 0.09410442627327706, | |
| "learning_rate": 3.371612311208692e-06, | |
| "loss": 0.3787, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.749661705006766, | |
| "grad_norm": 0.08704554678240813, | |
| "learning_rate": 3.0866615990230085e-06, | |
| "loss": 0.3726, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.760487144790257, | |
| "grad_norm": 0.08454007173317693, | |
| "learning_rate": 2.814053077134835e-06, | |
| "loss": 0.3689, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.7713125845737485, | |
| "grad_norm": 0.0842299187052216, | |
| "learning_rate": 2.553830490703666e-06, | |
| "loss": 0.3677, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.7821380243572396, | |
| "grad_norm": 0.08055057722300664, | |
| "learning_rate": 2.306035597332885e-06, | |
| "loss": 0.3796, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.7929634641407306, | |
| "grad_norm": 0.07811234003228873, | |
| "learning_rate": 2.0707081603690458e-06, | |
| "loss": 0.3733, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.803788903924222, | |
| "grad_norm": 0.07814276948421515, | |
| "learning_rate": 1.8478859425210548e-06, | |
| "loss": 0.3724, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.814614343707713, | |
| "grad_norm": 0.07992325685583533, | |
| "learning_rate": 1.6376046998004413e-06, | |
| "loss": 0.3784, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.825439783491204, | |
| "grad_norm": 0.08281854715683096, | |
| "learning_rate": 1.4398981757836628e-06, | |
| "loss": 0.3675, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.8362652232746957, | |
| "grad_norm": 0.07865994423809364, | |
| "learning_rate": 1.2547980961972629e-06, | |
| "loss": 0.3722, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.8470906630581867, | |
| "grad_norm": 0.0785799626146214, | |
| "learning_rate": 1.0823341638269303e-06, | |
| "loss": 0.377, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.8579161028416777, | |
| "grad_norm": 0.08336944108556697, | |
| "learning_rate": 9.225340537510897e-07, | |
| "loss": 0.3724, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.8687415426251692, | |
| "grad_norm": 0.07735010497815362, | |
| "learning_rate": 7.754234088999379e-07, | |
| "loss": 0.3779, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.8795669824086603, | |
| "grad_norm": 0.07582962822504996, | |
| "learning_rate": 6.410258359405585e-07, | |
| "loss": 0.3749, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.8903924221921518, | |
| "grad_norm": 0.07634846186806633, | |
| "learning_rate": 5.19362901488778e-07, | |
| "loss": 0.3713, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.901217861975643, | |
| "grad_norm": 0.07627711053895611, | |
| "learning_rate": 4.1045412864838765e-07, | |
| "loss": 0.3722, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.912043301759134, | |
| "grad_norm": 0.07557853802708582, | |
| "learning_rate": 3.143169938783253e-07, | |
| "loss": 0.3744, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.9228687415426253, | |
| "grad_norm": 0.08096730663606065, | |
| "learning_rate": 2.3096692418820733e-07, | |
| "loss": 0.377, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.9336941813261164, | |
| "grad_norm": 0.0767483032746884, | |
| "learning_rate": 1.6041729466280686e-07, | |
| "loss": 0.3725, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.944519621109608, | |
| "grad_norm": 0.07637424113539698, | |
| "learning_rate": 1.0267942631577932e-07, | |
| "loss": 0.3786, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.955345060893099, | |
| "grad_norm": 0.07515546978566537, | |
| "learning_rate": 5.7762584272937903e-08, | |
| "loss": 0.3765, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.96617050067659, | |
| "grad_norm": 0.07848788057525896, | |
| "learning_rate": 2.5673976285558455e-08, | |
| "loss": 0.3746, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.976995940460081, | |
| "grad_norm": 0.07564833351349101, | |
| "learning_rate": 6.418751573713522e-09, | |
| "loss": 0.3818, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.9878213802435725, | |
| "grad_norm": 0.07278613437346892, | |
| "learning_rate": 0.0, | |
| "loss": 0.37, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.9878213802435725, | |
| "step": 276, | |
| "total_flos": 6.629121904995205e+18, | |
| "train_loss": 0.4911867194417594, | |
| "train_runtime": 27032.019, | |
| "train_samples_per_second": 5.248, | |
| "train_steps_per_second": 0.01 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 276, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.629121904995205e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |