| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 3080, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.032467532467532464, | |
| "grad_norm": 0.56201171875, | |
| "learning_rate": 0.00019999479806942977, | |
| "loss": 0.8091, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06493506493506493, | |
| "grad_norm": 0.406005859375, | |
| "learning_rate": 0.00019997919281892067, | |
| "loss": 0.7613, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09740259740259741, | |
| "grad_norm": 0.4560546875, | |
| "learning_rate": 0.0001999531858720213, | |
| "loss": 0.797, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12987012987012986, | |
| "grad_norm": 0.424560546875, | |
| "learning_rate": 0.0001999167799344583, | |
| "loss": 0.7912, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.16233766233766234, | |
| "grad_norm": 0.468994140625, | |
| "learning_rate": 0.00019986997879385487, | |
| "loss": 0.7866, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.19480519480519481, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 0.0001998127873193367, | |
| "loss": 0.7504, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.22727272727272727, | |
| "grad_norm": 0.389892578125, | |
| "learning_rate": 0.00019974521146102537, | |
| "loss": 0.8146, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2597402597402597, | |
| "grad_norm": 0.39599609375, | |
| "learning_rate": 0.00019966725824941932, | |
| "loss": 0.7641, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2922077922077922, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.00019957893579466252, | |
| "loss": 0.7902, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3246753246753247, | |
| "grad_norm": 0.3681640625, | |
| "learning_rate": 0.00019948025328570042, | |
| "loss": 0.7831, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 0.4013671875, | |
| "learning_rate": 0.00019937122098932428, | |
| "loss": 0.7268, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.38961038961038963, | |
| "grad_norm": 0.41748046875, | |
| "learning_rate": 0.00019925185024910277, | |
| "loss": 0.7673, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.42207792207792205, | |
| "grad_norm": 0.423583984375, | |
| "learning_rate": 0.000199122153484202, | |
| "loss": 0.7652, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 0.428955078125, | |
| "learning_rate": 0.0001989821441880933, | |
| "loss": 0.7806, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.487012987012987, | |
| "grad_norm": 0.387939453125, | |
| "learning_rate": 0.00019883183692714936, | |
| "loss": 0.7474, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "grad_norm": 0.428955078125, | |
| "learning_rate": 0.0001986712473391289, | |
| "loss": 0.752, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.551948051948052, | |
| "grad_norm": 0.38330078125, | |
| "learning_rate": 0.00019850039213154973, | |
| "loss": 0.785, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5844155844155844, | |
| "grad_norm": 0.44677734375, | |
| "learning_rate": 0.0001983192890799503, | |
| "loss": 0.7612, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6168831168831169, | |
| "grad_norm": 0.406982421875, | |
| "learning_rate": 0.00019812795702604073, | |
| "loss": 0.784, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6493506493506493, | |
| "grad_norm": 0.410400390625, | |
| "learning_rate": 0.00019792641587574212, | |
| "loss": 0.7649, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6818181818181818, | |
| "grad_norm": 0.4072265625, | |
| "learning_rate": 0.00019771468659711595, | |
| "loss": 0.7543, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 0.37744140625, | |
| "learning_rate": 0.00019749279121818235, | |
| "loss": 0.7209, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7467532467532467, | |
| "grad_norm": 0.431396484375, | |
| "learning_rate": 0.00019726075282462845, | |
| "loss": 0.7532, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7792207792207793, | |
| "grad_norm": 0.3720703125, | |
| "learning_rate": 0.00019701859555740648, | |
| "loss": 0.7781, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8116883116883117, | |
| "grad_norm": 0.420654296875, | |
| "learning_rate": 0.00019676634461022224, | |
| "loss": 0.7613, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8441558441558441, | |
| "grad_norm": 0.350830078125, | |
| "learning_rate": 0.000196504026226914, | |
| "loss": 0.7439, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8766233766233766, | |
| "grad_norm": 0.40087890625, | |
| "learning_rate": 0.00019623166769872213, | |
| "loss": 0.7324, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.3779296875, | |
| "learning_rate": 0.00019594929736144976, | |
| "loss": 0.7481, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9415584415584416, | |
| "grad_norm": 0.3818359375, | |
| "learning_rate": 0.00019565694459251458, | |
| "loss": 0.7584, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.974025974025974, | |
| "grad_norm": 0.4423828125, | |
| "learning_rate": 0.00019535463980789277, | |
| "loss": 0.7364, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0064935064935066, | |
| "grad_norm": 0.42138671875, | |
| "learning_rate": 0.00019504241445895436, | |
| "loss": 0.7284, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0389610389610389, | |
| "grad_norm": 0.4599609375, | |
| "learning_rate": 0.000194720301029191, | |
| "loss": 0.7216, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.0714285714285714, | |
| "grad_norm": 0.450927734375, | |
| "learning_rate": 0.00019438833303083678, | |
| "loss": 0.6916, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.103896103896104, | |
| "grad_norm": 0.387451171875, | |
| "learning_rate": 0.00019404654500138117, | |
| "loss": 0.7093, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.1363636363636362, | |
| "grad_norm": 0.46435546875, | |
| "learning_rate": 0.0001936949724999762, | |
| "loss": 0.7334, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.1688311688311688, | |
| "grad_norm": 0.423583984375, | |
| "learning_rate": 0.0001933336521037367, | |
| "loss": 0.7308, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.2012987012987013, | |
| "grad_norm": 0.377197265625, | |
| "learning_rate": 0.00019296262140393498, | |
| "loss": 0.7008, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.2337662337662338, | |
| "grad_norm": 0.427001953125, | |
| "learning_rate": 0.0001925819190020898, | |
| "loss": 0.7344, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.2662337662337662, | |
| "grad_norm": 0.397216796875, | |
| "learning_rate": 0.0001921915845059505, | |
| "loss": 0.6947, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.2987012987012987, | |
| "grad_norm": 0.42626953125, | |
| "learning_rate": 0.00019179165852537596, | |
| "loss": 0.7058, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.3311688311688312, | |
| "grad_norm": 0.467529296875, | |
| "learning_rate": 0.00019138218266810986, | |
| "loss": 0.7244, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.3636363636363638, | |
| "grad_norm": 0.457275390625, | |
| "learning_rate": 0.00019096319953545185, | |
| "loss": 0.7003, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.396103896103896, | |
| "grad_norm": 0.4013671875, | |
| "learning_rate": 0.0001905347527178252, | |
| "loss": 0.6961, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.458740234375, | |
| "learning_rate": 0.0001900968867902419, | |
| "loss": 0.7052, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.4610389610389611, | |
| "grad_norm": 0.426025390625, | |
| "learning_rate": 0.00018964964730766508, | |
| "loss": 0.7059, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.4935064935064934, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 0.0001891930808002694, | |
| "loss": 0.6921, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.525974025974026, | |
| "grad_norm": 0.46044921875, | |
| "learning_rate": 0.00018872723476860034, | |
| "loss": 0.7038, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.5584415584415585, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 0.00018825215767863214, | |
| "loss": 0.7459, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.5909090909090908, | |
| "grad_norm": 0.46728515625, | |
| "learning_rate": 0.00018776789895672558, | |
| "loss": 0.726, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.6233766233766234, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.00018727450898448563, | |
| "loss": 0.7192, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.655844155844156, | |
| "grad_norm": 0.490966796875, | |
| "learning_rate": 0.00018677203909351988, | |
| "loss": 0.7186, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.6883116883116882, | |
| "grad_norm": 0.4443359375, | |
| "learning_rate": 0.00018626054156009806, | |
| "loss": 0.7177, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.7207792207792207, | |
| "grad_norm": 0.509765625, | |
| "learning_rate": 0.00018574006959971333, | |
| "loss": 0.6966, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.7532467532467533, | |
| "grad_norm": 0.467529296875, | |
| "learning_rate": 0.00018521067736154568, | |
| "loss": 0.6668, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "grad_norm": 0.5068359375, | |
| "learning_rate": 0.00018467241992282843, | |
| "loss": 0.6856, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.433837890625, | |
| "learning_rate": 0.00018412535328311814, | |
| "loss": 0.7126, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.8506493506493507, | |
| "grad_norm": 0.46826171875, | |
| "learning_rate": 0.0001835695343584683, | |
| "loss": 0.7255, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.883116883116883, | |
| "grad_norm": 0.452880859375, | |
| "learning_rate": 0.00018300502097550806, | |
| "loss": 0.6792, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.9155844155844157, | |
| "grad_norm": 0.46142578125, | |
| "learning_rate": 0.00018243187186542593, | |
| "loss": 0.6891, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.948051948051948, | |
| "grad_norm": 0.51904296875, | |
| "learning_rate": 0.00018185014665785936, | |
| "loss": 0.6959, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.9805194805194806, | |
| "grad_norm": 0.4833984375, | |
| "learning_rate": 0.00018125990587469123, | |
| "loss": 0.6975, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.012987012987013, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.000180661210923753, | |
| "loss": 0.6652, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.0454545454545454, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 0.00018005412409243606, | |
| "loss": 0.6559, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.0779220779220777, | |
| "grad_norm": 0.580078125, | |
| "learning_rate": 0.00017943870854121124, | |
| "loss": 0.6489, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.1103896103896105, | |
| "grad_norm": 0.469970703125, | |
| "learning_rate": 0.00017881502829705808, | |
| "loss": 0.6409, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 0.51611328125, | |
| "learning_rate": 0.000178183148246803, | |
| "loss": 0.6835, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.175324675324675, | |
| "grad_norm": 0.52099609375, | |
| "learning_rate": 0.00017754313413036906, | |
| "loss": 0.66, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.207792207792208, | |
| "grad_norm": 0.490966796875, | |
| "learning_rate": 0.0001768950525339362, | |
| "loss": 0.6767, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.24025974025974, | |
| "grad_norm": 0.5166015625, | |
| "learning_rate": 0.00017623897088301385, | |
| "loss": 0.6568, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.2727272727272725, | |
| "grad_norm": 0.64111328125, | |
| "learning_rate": 0.00017557495743542585, | |
| "loss": 0.6585, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.3051948051948052, | |
| "grad_norm": 0.52001953125, | |
| "learning_rate": 0.00017490308127420928, | |
| "loss": 0.6728, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.3376623376623376, | |
| "grad_norm": 0.55908203125, | |
| "learning_rate": 0.000174223412300427, | |
| "loss": 0.6574, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.3701298701298703, | |
| "grad_norm": 0.53076171875, | |
| "learning_rate": 0.00017353602122589527, | |
| "loss": 0.6486, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.4025974025974026, | |
| "grad_norm": 0.51513671875, | |
| "learning_rate": 0.00017284097956582692, | |
| "loss": 0.6571, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.435064935064935, | |
| "grad_norm": 0.482177734375, | |
| "learning_rate": 0.0001721383596313912, | |
| "loss": 0.666, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.4675324675324677, | |
| "grad_norm": 0.492919921875, | |
| "learning_rate": 0.00017142823452219038, | |
| "loss": 0.665, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.54541015625, | |
| "learning_rate": 0.00017071067811865476, | |
| "loss": 0.654, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.5324675324675323, | |
| "grad_norm": 0.5810546875, | |
| "learning_rate": 0.00016998576507435618, | |
| "loss": 0.6573, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.564935064935065, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00016925357080824118, | |
| "loss": 0.6627, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.5974025974025974, | |
| "grad_norm": 0.5185546875, | |
| "learning_rate": 0.00016851417149678444, | |
| "loss": 0.651, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.62987012987013, | |
| "grad_norm": 0.603515625, | |
| "learning_rate": 0.0001677676440660636, | |
| "loss": 0.6616, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.6623376623376624, | |
| "grad_norm": 1.826171875, | |
| "learning_rate": 0.00016701406618375596, | |
| "loss": 0.6668, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.6948051948051948, | |
| "grad_norm": 0.541015625, | |
| "learning_rate": 0.00016625351625105796, | |
| "loss": 0.6661, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 0.00016548607339452853, | |
| "loss": 0.6378, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.75974025974026, | |
| "grad_norm": 0.52001953125, | |
| "learning_rate": 0.00016471181745785672, | |
| "loss": 0.6672, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.792207792207792, | |
| "grad_norm": 0.5888671875, | |
| "learning_rate": 0.00016393082899355516, | |
| "loss": 0.6573, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.824675324675325, | |
| "grad_norm": 0.5673828125, | |
| "learning_rate": 0.0001631431892545791, | |
| "loss": 0.6471, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 0.69482421875, | |
| "learning_rate": 0.00016234898018587337, | |
| "loss": 0.6604, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.8896103896103895, | |
| "grad_norm": 0.5029296875, | |
| "learning_rate": 0.00016154828441584655, | |
| "loss": 0.6535, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.9220779220779223, | |
| "grad_norm": 0.49462890625, | |
| "learning_rate": 0.00016074118524777477, | |
| "loss": 0.647, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.9545454545454546, | |
| "grad_norm": 0.55322265625, | |
| "learning_rate": 0.0001599277666511347, | |
| "loss": 0.66, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.987012987012987, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 0.00015910811325286768, | |
| "loss": 0.6381, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.0194805194805197, | |
| "grad_norm": 0.509765625, | |
| "learning_rate": 0.00015828231032857503, | |
| "loss": 0.6559, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.051948051948052, | |
| "grad_norm": 0.55517578125, | |
| "learning_rate": 0.00015745044379364634, | |
| "loss": 0.5904, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.0844155844155843, | |
| "grad_norm": 0.58642578125, | |
| "learning_rate": 0.00015661260019432077, | |
| "loss": 0.6012, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.116883116883117, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 0.00015576886669868296, | |
| "loss": 0.6161, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.1493506493506493, | |
| "grad_norm": 0.521484375, | |
| "learning_rate": 0.0001549193310875942, | |
| "loss": 0.6019, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.1818181818181817, | |
| "grad_norm": 0.55224609375, | |
| "learning_rate": 0.00015406408174555976, | |
| "loss": 0.5906, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.2142857142857144, | |
| "grad_norm": 0.61865234375, | |
| "learning_rate": 0.00015320320765153367, | |
| "loss": 0.6128, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.2467532467532467, | |
| "grad_norm": 0.642578125, | |
| "learning_rate": 0.00015233679836966122, | |
| "loss": 0.6012, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.279220779220779, | |
| "grad_norm": 0.60107421875, | |
| "learning_rate": 0.000151464944039961, | |
| "loss": 0.6152, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.311688311688312, | |
| "grad_norm": 0.5966796875, | |
| "learning_rate": 0.00015058773536894685, | |
| "loss": 0.5867, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.344155844155844, | |
| "grad_norm": 0.5986328125, | |
| "learning_rate": 0.00014970526362019079, | |
| "loss": 0.6248, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.3766233766233764, | |
| "grad_norm": 0.60009765625, | |
| "learning_rate": 0.00014881762060482814, | |
| "loss": 0.6134, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.409090909090909, | |
| "grad_norm": 0.5986328125, | |
| "learning_rate": 0.0001479248986720057, | |
| "loss": 0.6078, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.4415584415584415, | |
| "grad_norm": 0.5732421875, | |
| "learning_rate": 0.0001470271906992737, | |
| "loss": 0.6028, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 3.474025974025974, | |
| "grad_norm": 0.63623046875, | |
| "learning_rate": 0.00014612459008292306, | |
| "loss": 0.6176, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 3.5064935064935066, | |
| "grad_norm": 0.6259765625, | |
| "learning_rate": 0.00014521719072826858, | |
| "loss": 0.6287, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 3.538961038961039, | |
| "grad_norm": 0.5595703125, | |
| "learning_rate": 0.00014430508703987914, | |
| "loss": 0.6007, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 0.00014338837391175582, | |
| "loss": 0.6112, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.603896103896104, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 0.00014246714671745965, | |
| "loss": 0.6273, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.64501953125, | |
| "learning_rate": 0.00014154150130018866, | |
| "loss": 0.5894, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 3.6688311688311686, | |
| "grad_norm": 0.5341796875, | |
| "learning_rate": 0.00014061153396280674, | |
| "loss": 0.6315, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 3.7012987012987013, | |
| "grad_norm": 0.7060546875, | |
| "learning_rate": 0.00013967734145782425, | |
| "loss": 0.6377, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 3.7337662337662336, | |
| "grad_norm": 0.626953125, | |
| "learning_rate": 0.0001387390209773323, | |
| "loss": 0.6212, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.7662337662337664, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 0.00013779667014289065, | |
| "loss": 0.6126, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 3.7987012987012987, | |
| "grad_norm": 0.67431640625, | |
| "learning_rate": 0.00013685038699537164, | |
| "loss": 0.6255, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.8311688311688314, | |
| "grad_norm": 0.6494140625, | |
| "learning_rate": 0.00013590026998475986, | |
| "loss": 0.6199, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 3.8636363636363638, | |
| "grad_norm": 0.6630859375, | |
| "learning_rate": 0.00013494641795990986, | |
| "loss": 0.6109, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 3.896103896103896, | |
| "grad_norm": 0.63525390625, | |
| "learning_rate": 0.00013398893015826167, | |
| "loss": 0.6359, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.928571428571429, | |
| "grad_norm": 0.5849609375, | |
| "learning_rate": 0.00013302790619551674, | |
| "loss": 0.6523, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 3.961038961038961, | |
| "grad_norm": 0.59033203125, | |
| "learning_rate": 0.00013206344605527355, | |
| "loss": 0.5958, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 3.9935064935064934, | |
| "grad_norm": 0.62060546875, | |
| "learning_rate": 0.00013109565007862596, | |
| "loss": 0.6284, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.025974025974026, | |
| "grad_norm": 0.76904296875, | |
| "learning_rate": 0.00013012461895372344, | |
| "loss": 0.5536, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 4.058441558441558, | |
| "grad_norm": 0.65966796875, | |
| "learning_rate": 0.00012915045370529585, | |
| "loss": 0.5586, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.090909090909091, | |
| "grad_norm": 0.64599609375, | |
| "learning_rate": 0.00012817325568414297, | |
| "loss": 0.581, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 4.123376623376624, | |
| "grad_norm": 0.681640625, | |
| "learning_rate": 0.00012719312655658994, | |
| "loss": 0.545, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 4.1558441558441555, | |
| "grad_norm": 0.6435546875, | |
| "learning_rate": 0.00012621016829391022, | |
| "loss": 0.5603, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 4.188311688311688, | |
| "grad_norm": 0.7587890625, | |
| "learning_rate": 0.0001252244831617165, | |
| "loss": 0.557, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 4.220779220779221, | |
| "grad_norm": 0.642578125, | |
| "learning_rate": 0.00012423617370932127, | |
| "loss": 0.5634, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.253246753246753, | |
| "grad_norm": 0.7373046875, | |
| "learning_rate": 0.00012324534275906748, | |
| "loss": 0.567, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 0.00012225209339563145, | |
| "loss": 0.554, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 4.318181818181818, | |
| "grad_norm": 0.68310546875, | |
| "learning_rate": 0.00012125652895529766, | |
| "loss": 0.5777, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 4.35064935064935, | |
| "grad_norm": 0.85205078125, | |
| "learning_rate": 0.0001202587530152081, | |
| "loss": 0.5824, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 4.383116883116883, | |
| "grad_norm": 0.70751953125, | |
| "learning_rate": 0.00011925886938258605, | |
| "loss": 0.5551, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 4.415584415584416, | |
| "grad_norm": 0.69140625, | |
| "learning_rate": 0.00011825698208393619, | |
| "loss": 0.574, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 4.448051948051948, | |
| "grad_norm": 0.708984375, | |
| "learning_rate": 0.00011725319535422188, | |
| "loss": 0.5919, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 4.48051948051948, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 0.00011624761362602061, | |
| "loss": 0.5592, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 4.512987012987013, | |
| "grad_norm": 0.66455078125, | |
| "learning_rate": 0.00011524034151865903, | |
| "loss": 0.5882, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 4.545454545454545, | |
| "grad_norm": 0.7119140625, | |
| "learning_rate": 0.00011423148382732853, | |
| "loss": 0.5741, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.577922077922078, | |
| "grad_norm": 0.66748046875, | |
| "learning_rate": 0.00011322114551218239, | |
| "loss": 0.5842, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 4.6103896103896105, | |
| "grad_norm": 0.69873046875, | |
| "learning_rate": 0.000112209431687416, | |
| "loss": 0.5626, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 4.642857142857143, | |
| "grad_norm": 0.6357421875, | |
| "learning_rate": 0.00011119644761033078, | |
| "loss": 0.5657, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 4.675324675324675, | |
| "grad_norm": 0.677734375, | |
| "learning_rate": 0.00011018229867038356, | |
| "loss": 0.5664, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 4.707792207792208, | |
| "grad_norm": 0.7998046875, | |
| "learning_rate": 0.00010916709037822173, | |
| "loss": 0.5758, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 4.740259740259741, | |
| "grad_norm": 0.68212890625, | |
| "learning_rate": 0.00010815092835470633, | |
| "loss": 0.5785, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 4.7727272727272725, | |
| "grad_norm": 0.75341796875, | |
| "learning_rate": 0.00010713391831992323, | |
| "loss": 0.5646, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 4.805194805194805, | |
| "grad_norm": 0.65478515625, | |
| "learning_rate": 0.00010611616608218429, | |
| "loss": 0.577, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 4.837662337662338, | |
| "grad_norm": 0.7626953125, | |
| "learning_rate": 0.00010509777752701899, | |
| "loss": 0.5878, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 4.87012987012987, | |
| "grad_norm": 0.88623046875, | |
| "learning_rate": 0.00010407885860615859, | |
| "loss": 0.5931, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.902597402597403, | |
| "grad_norm": 0.80419921875, | |
| "learning_rate": 0.00010305951532651261, | |
| "loss": 0.5946, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 4.935064935064935, | |
| "grad_norm": 0.71630859375, | |
| "learning_rate": 0.00010203985373914056, | |
| "loss": 0.5671, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 4.967532467532467, | |
| "grad_norm": 0.73779296875, | |
| "learning_rate": 0.00010101997992821797, | |
| "loss": 0.5942, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.67333984375, | |
| "learning_rate": 0.0001, | |
| "loss": 0.5654, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 5.032467532467533, | |
| "grad_norm": 0.7333984375, | |
| "learning_rate": 9.898002007178204e-05, | |
| "loss": 0.5399, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 5.064935064935065, | |
| "grad_norm": 0.77001953125, | |
| "learning_rate": 9.79601462608595e-05, | |
| "loss": 0.4992, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 5.097402597402597, | |
| "grad_norm": 0.74267578125, | |
| "learning_rate": 9.69404846734874e-05, | |
| "loss": 0.5386, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 5.12987012987013, | |
| "grad_norm": 0.87060546875, | |
| "learning_rate": 9.592114139384145e-05, | |
| "loss": 0.5219, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 5.162337662337662, | |
| "grad_norm": 0.8505859375, | |
| "learning_rate": 9.490222247298099e-05, | |
| "loss": 0.5218, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 5.194805194805195, | |
| "grad_norm": 0.7724609375, | |
| "learning_rate": 9.388383391781575e-05, | |
| "loss": 0.5193, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 5.2272727272727275, | |
| "grad_norm": 0.8701171875, | |
| "learning_rate": 9.286608168007678e-05, | |
| "loss": 0.5356, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 5.259740259740259, | |
| "grad_norm": 0.7861328125, | |
| "learning_rate": 9.184907164529368e-05, | |
| "loss": 0.5276, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 5.292207792207792, | |
| "grad_norm": 0.86083984375, | |
| "learning_rate": 9.083290962177828e-05, | |
| "loss": 0.5385, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 5.324675324675325, | |
| "grad_norm": 0.8212890625, | |
| "learning_rate": 8.981770132961649e-05, | |
| "loss": 0.5315, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 5.357142857142857, | |
| "grad_norm": 0.77001953125, | |
| "learning_rate": 8.880355238966923e-05, | |
| "loss": 0.5144, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 5.3896103896103895, | |
| "grad_norm": 0.8388671875, | |
| "learning_rate": 8.779056831258402e-05, | |
| "loss": 0.557, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 5.422077922077922, | |
| "grad_norm": 0.84326171875, | |
| "learning_rate": 8.67788544878176e-05, | |
| "loss": 0.5406, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.9423828125, | |
| "learning_rate": 8.57685161726715e-05, | |
| "loss": 0.5483, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 5.487012987012987, | |
| "grad_norm": 0.7421875, | |
| "learning_rate": 8.475965848134099e-05, | |
| "loss": 0.5399, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 5.51948051948052, | |
| "grad_norm": 0.8623046875, | |
| "learning_rate": 8.375238637397942e-05, | |
| "loss": 0.5427, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 5.551948051948052, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 8.274680464577816e-05, | |
| "loss": 0.4993, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 5.584415584415584, | |
| "grad_norm": 0.6953125, | |
| "learning_rate": 8.174301791606385e-05, | |
| "loss": 0.5258, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 5.616883116883117, | |
| "grad_norm": 0.81591796875, | |
| "learning_rate": 8.074113061741397e-05, | |
| "loss": 0.5342, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 5.64935064935065, | |
| "grad_norm": 0.8369140625, | |
| "learning_rate": 7.974124698479192e-05, | |
| "loss": 0.5295, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 5.681818181818182, | |
| "grad_norm": 0.77392578125, | |
| "learning_rate": 7.874347104470234e-05, | |
| "loss": 0.5291, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 1.908203125, | |
| "learning_rate": 7.774790660436858e-05, | |
| "loss": 0.5367, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 5.746753246753247, | |
| "grad_norm": 0.82861328125, | |
| "learning_rate": 7.675465724093254e-05, | |
| "loss": 0.5437, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 5.779220779220779, | |
| "grad_norm": 0.84228515625, | |
| "learning_rate": 7.576382629067877e-05, | |
| "loss": 0.5168, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 5.811688311688312, | |
| "grad_norm": 0.7587890625, | |
| "learning_rate": 7.47755168382835e-05, | |
| "loss": 0.5411, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 5.8441558441558445, | |
| "grad_norm": 0.77001953125, | |
| "learning_rate": 7.378983170608982e-05, | |
| "loss": 0.5443, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 5.876623376623376, | |
| "grad_norm": 0.75048828125, | |
| "learning_rate": 7.280687344341007e-05, | |
| "loss": 0.5365, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 5.909090909090909, | |
| "grad_norm": 0.7861328125, | |
| "learning_rate": 7.182674431585704e-05, | |
| "loss": 0.5274, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 5.941558441558442, | |
| "grad_norm": 0.83984375, | |
| "learning_rate": 7.084954629470417e-05, | |
| "loss": 0.5325, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 5.974025974025974, | |
| "grad_norm": 0.986328125, | |
| "learning_rate": 6.98753810462766e-05, | |
| "loss": 0.5503, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 6.0064935064935066, | |
| "grad_norm": 0.72314453125, | |
| "learning_rate": 6.890434992137407e-05, | |
| "loss": 0.5213, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 6.038961038961039, | |
| "grad_norm": 0.74560546875, | |
| "learning_rate": 6.793655394472644e-05, | |
| "loss": 0.4754, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 6.071428571428571, | |
| "grad_norm": 0.826171875, | |
| "learning_rate": 6.697209380448333e-05, | |
| "loss": 0.4703, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 6.103896103896104, | |
| "grad_norm": 0.90283203125, | |
| "learning_rate": 6.601106984173835e-05, | |
| "loss": 0.5074, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 6.136363636363637, | |
| "grad_norm": 0.91650390625, | |
| "learning_rate": 6.505358204009017e-05, | |
| "loss": 0.4941, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 6.1688311688311686, | |
| "grad_norm": 0.8486328125, | |
| "learning_rate": 6.409973001524012e-05, | |
| "loss": 0.4937, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 6.201298701298701, | |
| "grad_norm": 0.92822265625, | |
| "learning_rate": 6.314961300462841e-05, | |
| "loss": 0.5005, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 6.233766233766234, | |
| "grad_norm": 0.85546875, | |
| "learning_rate": 6.220332985710936e-05, | |
| "loss": 0.4936, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 6.266233766233766, | |
| "grad_norm": 0.91162109375, | |
| "learning_rate": 6.126097902266772e-05, | |
| "loss": 0.4682, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 6.298701298701299, | |
| "grad_norm": 0.88232421875, | |
| "learning_rate": 6.0322658542175736e-05, | |
| "loss": 0.5096, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 6.3311688311688314, | |
| "grad_norm": 0.90576171875, | |
| "learning_rate": 5.938846603719332e-05, | |
| "loss": 0.5064, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 6.363636363636363, | |
| "grad_norm": 0.931640625, | |
| "learning_rate": 5.845849869981137e-05, | |
| "loss": 0.5013, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 6.396103896103896, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 5.7532853282540387e-05, | |
| "loss": 0.4929, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 6.428571428571429, | |
| "grad_norm": 0.8662109375, | |
| "learning_rate": 5.6611626088244194e-05, | |
| "loss": 0.5128, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 6.461038961038961, | |
| "grad_norm": 0.9287109375, | |
| "learning_rate": 5.569491296012095e-05, | |
| "loss": 0.4912, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 6.4935064935064934, | |
| "grad_norm": 0.76904296875, | |
| "learning_rate": 5.478280927173145e-05, | |
| "loss": 0.4999, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.525974025974026, | |
| "grad_norm": 0.93701171875, | |
| "learning_rate": 5.387540991707697e-05, | |
| "loss": 0.4988, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 6.558441558441558, | |
| "grad_norm": 0.8828125, | |
| "learning_rate": 5.297280930072632e-05, | |
| "loss": 0.4825, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 6.590909090909091, | |
| "grad_norm": 0.83984375, | |
| "learning_rate": 5.207510132799436e-05, | |
| "loss": 0.5054, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 6.623376623376624, | |
| "grad_norm": 0.88671875, | |
| "learning_rate": 5.11823793951719e-05, | |
| "loss": 0.4975, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 6.6558441558441555, | |
| "grad_norm": 0.9208984375, | |
| "learning_rate": 5.029473637980926e-05, | |
| "loss": 0.497, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 6.688311688311688, | |
| "grad_norm": 0.86865234375, | |
| "learning_rate": 4.9412264631053216e-05, | |
| "loss": 0.4983, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 6.720779220779221, | |
| "grad_norm": 0.9228515625, | |
| "learning_rate": 4.853505596003905e-05, | |
| "loss": 0.522, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 6.753246753246753, | |
| "grad_norm": 0.86865234375, | |
| "learning_rate": 4.7663201630338816e-05, | |
| "loss": 0.4981, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 6.785714285714286, | |
| "grad_norm": 0.87060546875, | |
| "learning_rate": 4.6796792348466356e-05, | |
| "loss": 0.4795, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 6.818181818181818, | |
| "grad_norm": 1.0458984375, | |
| "learning_rate": 4.593591825444028e-05, | |
| "loss": 0.5133, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 6.85064935064935, | |
| "grad_norm": 0.828125, | |
| "learning_rate": 4.5080668912405855e-05, | |
| "loss": 0.5019, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 6.883116883116883, | |
| "grad_norm": 0.87158203125, | |
| "learning_rate": 4.423113330131707e-05, | |
| "loss": 0.4884, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 6.915584415584416, | |
| "grad_norm": 0.82568359375, | |
| "learning_rate": 4.3387399805679255e-05, | |
| "loss": 0.492, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 6.948051948051948, | |
| "grad_norm": 0.87255859375, | |
| "learning_rate": 4.25495562063537e-05, | |
| "loss": 0.489, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 6.98051948051948, | |
| "grad_norm": 0.90869140625, | |
| "learning_rate": 4.1717689671425e-05, | |
| "loss": 0.4903, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 7.012987012987013, | |
| "grad_norm": 0.87060546875, | |
| "learning_rate": 4.089188674713236e-05, | |
| "loss": 0.4801, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 7.045454545454546, | |
| "grad_norm": 0.94970703125, | |
| "learning_rate": 4.007223334886531e-05, | |
| "loss": 0.4534, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 7.077922077922078, | |
| "grad_norm": 0.92138671875, | |
| "learning_rate": 3.9258814752225284e-05, | |
| "loss": 0.4601, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 7.1103896103896105, | |
| "grad_norm": 0.87841796875, | |
| "learning_rate": 3.845171558415348e-05, | |
| "loss": 0.4498, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "grad_norm": 0.994140625, | |
| "learning_rate": 3.7651019814126654e-05, | |
| "loss": 0.4679, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 7.175324675324675, | |
| "grad_norm": 0.98193359375, | |
| "learning_rate": 3.6856810745420886e-05, | |
| "loss": 0.468, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 7.207792207792208, | |
| "grad_norm": 0.89111328125, | |
| "learning_rate": 3.606917100644488e-05, | |
| "loss": 0.4599, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 7.240259740259741, | |
| "grad_norm": 1.123046875, | |
| "learning_rate": 3.528818254214329e-05, | |
| "loss": 0.4696, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 0.85107421875, | |
| "learning_rate": 3.45139266054715e-05, | |
| "loss": 0.4594, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 7.305194805194805, | |
| "grad_norm": 0.86181640625, | |
| "learning_rate": 3.374648374894204e-05, | |
| "loss": 0.4488, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 7.337662337662338, | |
| "grad_norm": 0.97705078125, | |
| "learning_rate": 3.298593381624406e-05, | |
| "loss": 0.4619, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 7.37012987012987, | |
| "grad_norm": 0.90869140625, | |
| "learning_rate": 3.223235593393642e-05, | |
| "loss": 0.4717, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 7.402597402597403, | |
| "grad_norm": 0.974609375, | |
| "learning_rate": 3.1485828503215585e-05, | |
| "loss": 0.472, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 7.435064935064935, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 3.074642919175883e-05, | |
| "loss": 0.478, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 7.467532467532467, | |
| "grad_norm": 0.91015625, | |
| "learning_rate": 3.0014234925643837e-05, | |
| "loss": 0.4605, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 1.04296875, | |
| "learning_rate": 2.9289321881345254e-05, | |
| "loss": 0.4573, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 7.532467532467533, | |
| "grad_norm": 0.93994140625, | |
| "learning_rate": 2.8571765477809643e-05, | |
| "loss": 0.462, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 7.564935064935065, | |
| "grad_norm": 0.9765625, | |
| "learning_rate": 2.7861640368608844e-05, | |
| "loss": 0.4722, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 7.597402597402597, | |
| "grad_norm": 1.0263671875, | |
| "learning_rate": 2.71590204341731e-05, | |
| "loss": 0.4719, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 7.62987012987013, | |
| "grad_norm": 0.8818359375, | |
| "learning_rate": 2.6463978774104758e-05, | |
| "loss": 0.4476, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 7.662337662337662, | |
| "grad_norm": 1.0595703125, | |
| "learning_rate": 2.5776587699573006e-05, | |
| "loss": 0.4742, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 7.694805194805195, | |
| "grad_norm": 1.0439453125, | |
| "learning_rate": 2.509691872579075e-05, | |
| "loss": 0.4714, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 7.7272727272727275, | |
| "grad_norm": 0.89404296875, | |
| "learning_rate": 2.4425042564574184e-05, | |
| "loss": 0.4771, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 7.759740259740259, | |
| "grad_norm": 0.90673828125, | |
| "learning_rate": 2.3761029116986178e-05, | |
| "loss": 0.4537, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 7.792207792207792, | |
| "grad_norm": 0.958984375, | |
| "learning_rate": 2.3104947466063787e-05, | |
| "loss": 0.4678, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 7.824675324675325, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 2.2456865869630972e-05, | |
| "loss": 0.4819, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 7.857142857142857, | |
| "grad_norm": 0.95068359375, | |
| "learning_rate": 2.181685175319702e-05, | |
| "loss": 0.4759, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 7.8896103896103895, | |
| "grad_norm": 0.99755859375, | |
| "learning_rate": 2.118497170294195e-05, | |
| "loss": 0.4795, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 7.922077922077922, | |
| "grad_norm": 1.0068359375, | |
| "learning_rate": 2.0561291458788733e-05, | |
| "loss": 0.4593, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 7.954545454545455, | |
| "grad_norm": 1.1923828125, | |
| "learning_rate": 1.994587590756397e-05, | |
| "loss": 0.4653, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 7.987012987012987, | |
| "grad_norm": 0.9072265625, | |
| "learning_rate": 1.9338789076247e-05, | |
| "loss": 0.4776, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 8.019480519480519, | |
| "grad_norm": 0.8837890625, | |
| "learning_rate": 1.874009412530877e-05, | |
| "loss": 0.4631, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 8.051948051948052, | |
| "grad_norm": 0.9091796875, | |
| "learning_rate": 1.8149853342140645e-05, | |
| "loss": 0.4328, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 8.084415584415584, | |
| "grad_norm": 1.1220703125, | |
| "learning_rate": 1.7568128134574113e-05, | |
| "loss": 0.4424, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 8.116883116883116, | |
| "grad_norm": 0.95947265625, | |
| "learning_rate": 1.6994979024491942e-05, | |
| "loss": 0.4333, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 8.14935064935065, | |
| "grad_norm": 0.89794921875, | |
| "learning_rate": 1.64304656415317e-05, | |
| "loss": 0.4438, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 8.181818181818182, | |
| "grad_norm": 1.0498046875, | |
| "learning_rate": 1.587464671688187e-05, | |
| "loss": 0.4356, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 8.214285714285714, | |
| "grad_norm": 0.93212890625, | |
| "learning_rate": 1.5327580077171587e-05, | |
| "loss": 0.4312, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 8.246753246753247, | |
| "grad_norm": 0.90771484375, | |
| "learning_rate": 1.4789322638454351e-05, | |
| "loss": 0.4429, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 8.279220779220779, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 1.4259930400286669e-05, | |
| "loss": 0.4469, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 8.311688311688311, | |
| "grad_norm": 1.0224609375, | |
| "learning_rate": 1.373945843990192e-05, | |
| "loss": 0.4497, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 8.344155844155845, | |
| "grad_norm": 1.712890625, | |
| "learning_rate": 1.322796090648013e-05, | |
| "loss": 0.4467, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 8.376623376623376, | |
| "grad_norm": 0.923828125, | |
| "learning_rate": 1.272549101551438e-05, | |
| "loss": 0.4437, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 8.409090909090908, | |
| "grad_norm": 1.134765625, | |
| "learning_rate": 1.2232101043274436e-05, | |
| "loss": 0.4374, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 8.441558441558442, | |
| "grad_norm": 1.0517578125, | |
| "learning_rate": 1.1747842321367886e-05, | |
| "loss": 0.4516, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 8.474025974025974, | |
| "grad_norm": 0.9755859375, | |
| "learning_rate": 1.1272765231399685e-05, | |
| "loss": 0.4382, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 8.506493506493506, | |
| "grad_norm": 1.1240234375, | |
| "learning_rate": 1.0806919199730615e-05, | |
| "loss": 0.4438, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 8.53896103896104, | |
| "grad_norm": 0.9833984375, | |
| "learning_rate": 1.035035269233493e-05, | |
| "loss": 0.446, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 8.571428571428571, | |
| "grad_norm": 0.97412109375, | |
| "learning_rate": 9.903113209758096e-06, | |
| "loss": 0.4419, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 8.603896103896105, | |
| "grad_norm": 1.0947265625, | |
| "learning_rate": 9.465247282174805e-06, | |
| "loss": 0.4607, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 8.636363636363637, | |
| "grad_norm": 0.912109375, | |
| "learning_rate": 9.036800464548157e-06, | |
| "loss": 0.4575, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 8.668831168831169, | |
| "grad_norm": 0.91845703125, | |
| "learning_rate": 8.617817331890154e-06, | |
| "loss": 0.4486, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 8.7012987012987, | |
| "grad_norm": 1.0595703125, | |
| "learning_rate": 8.208341474624071e-06, | |
| "loss": 0.455, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 8.733766233766234, | |
| "grad_norm": 0.98095703125, | |
| "learning_rate": 7.808415494049514e-06, | |
| "loss": 0.4565, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 8.766233766233766, | |
| "grad_norm": 0.9755859375, | |
| "learning_rate": 7.4180809979102036e-06, | |
| "loss": 0.447, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 8.7987012987013, | |
| "grad_norm": 1.0478515625, | |
| "learning_rate": 7.0373785960650475e-06, | |
| "loss": 0.4431, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 8.831168831168831, | |
| "grad_norm": 0.97802734375, | |
| "learning_rate": 6.666347896263325e-06, | |
| "loss": 0.4479, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 8.863636363636363, | |
| "grad_norm": 0.8994140625, | |
| "learning_rate": 6.3050275000238414e-06, | |
| "loss": 0.4558, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 8.896103896103895, | |
| "grad_norm": 1.087890625, | |
| "learning_rate": 5.953454998618857e-06, | |
| "loss": 0.4357, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 8.928571428571429, | |
| "grad_norm": 1.0263671875, | |
| "learning_rate": 5.611666969163243e-06, | |
| "loss": 0.4464, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 8.96103896103896, | |
| "grad_norm": 0.93505859375, | |
| "learning_rate": 5.27969897080901e-06, | |
| "loss": 0.4459, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 8.993506493506494, | |
| "grad_norm": 0.96435546875, | |
| "learning_rate": 4.957585541045684e-06, | |
| "loss": 0.4749, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 9.025974025974026, | |
| "grad_norm": 1.185546875, | |
| "learning_rate": 4.6453601921072395e-06, | |
| "loss": 0.4351, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 9.058441558441558, | |
| "grad_norm": 0.95849609375, | |
| "learning_rate": 4.34305540748543e-06, | |
| "loss": 0.429, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 0.93701171875, | |
| "learning_rate": 4.050702638550275e-06, | |
| "loss": 0.4335, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 9.123376623376624, | |
| "grad_norm": 0.982421875, | |
| "learning_rate": 3.768332301277866e-06, | |
| "loss": 0.445, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 9.155844155844155, | |
| "grad_norm": 0.935546875, | |
| "learning_rate": 3.495973773086014e-06, | |
| "loss": 0.4316, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 9.188311688311689, | |
| "grad_norm": 1.0439453125, | |
| "learning_rate": 3.233655389777801e-06, | |
| "loss": 0.4395, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 9.220779220779221, | |
| "grad_norm": 0.9423828125, | |
| "learning_rate": 2.9814044425935606e-06, | |
| "loss": 0.4207, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 9.253246753246753, | |
| "grad_norm": 1.09765625, | |
| "learning_rate": 2.739247175371562e-06, | |
| "loss": 0.4503, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 9.285714285714286, | |
| "grad_norm": 1.0166015625, | |
| "learning_rate": 2.5072087818176382e-06, | |
| "loss": 0.454, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 9.318181818181818, | |
| "grad_norm": 1.01171875, | |
| "learning_rate": 2.2853134028840594e-06, | |
| "loss": 0.426, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 9.35064935064935, | |
| "grad_norm": 1.130859375, | |
| "learning_rate": 2.073584124257899e-06, | |
| "loss": 0.4464, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 9.383116883116884, | |
| "grad_norm": 0.888671875, | |
| "learning_rate": 1.8720429739592982e-06, | |
| "loss": 0.4535, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 9.415584415584416, | |
| "grad_norm": 1.0634765625, | |
| "learning_rate": 1.6807109200496995e-06, | |
| "loss": 0.4556, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 9.448051948051948, | |
| "grad_norm": 0.9482421875, | |
| "learning_rate": 1.4996078684503144e-06, | |
| "loss": 0.4236, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 9.480519480519481, | |
| "grad_norm": 0.85888671875, | |
| "learning_rate": 1.3287526608711131e-06, | |
| "loss": 0.4366, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 9.512987012987013, | |
| "grad_norm": 1.01953125, | |
| "learning_rate": 1.1681630728506699e-06, | |
| "loss": 0.4236, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 9.545454545454545, | |
| "grad_norm": 1.0693359375, | |
| "learning_rate": 1.0178558119067315e-06, | |
| "loss": 0.4398, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 9.577922077922079, | |
| "grad_norm": 0.96337890625, | |
| "learning_rate": 8.778465157979976e-07, | |
| "loss": 0.4412, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 9.61038961038961, | |
| "grad_norm": 0.982421875, | |
| "learning_rate": 7.481497508972312e-07, | |
| "loss": 0.4307, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 9.642857142857142, | |
| "grad_norm": 0.9599609375, | |
| "learning_rate": 6.287790106757396e-07, | |
| "loss": 0.4415, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 9.675324675324676, | |
| "grad_norm": 1.0146484375, | |
| "learning_rate": 5.19746714299596e-07, | |
| "loss": 0.4412, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 9.707792207792208, | |
| "grad_norm": 0.95361328125, | |
| "learning_rate": 4.210642053375069e-07, | |
| "loss": 0.4416, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 9.74025974025974, | |
| "grad_norm": 1.00390625, | |
| "learning_rate": 3.3274175058067846e-07, | |
| "loss": 0.4383, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 9.772727272727273, | |
| "grad_norm": 0.947265625, | |
| "learning_rate": 2.547885389746485e-07, | |
| "loss": 0.4401, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 9.805194805194805, | |
| "grad_norm": 1.0166015625, | |
| "learning_rate": 1.8721268066330676e-07, | |
| "loss": 0.4332, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 9.837662337662337, | |
| "grad_norm": 1.0302734375, | |
| "learning_rate": 1.300212061451367e-07, | |
| "loss": 0.4358, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 9.87012987012987, | |
| "grad_norm": 1.0673828125, | |
| "learning_rate": 8.322006554171146e-08, | |
| "loss": 0.4412, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 9.902597402597403, | |
| "grad_norm": 0.96826171875, | |
| "learning_rate": 4.6814127978722644e-08, | |
| "loss": 0.4329, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 9.935064935064934, | |
| "grad_norm": 1.03515625, | |
| "learning_rate": 2.080718107935198e-08, | |
| "loss": 0.4401, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 9.967532467532468, | |
| "grad_norm": 1.044921875, | |
| "learning_rate": 5.201930570242208e-09, | |
| "loss": 0.4407, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.93701171875, | |
| "learning_rate": 0.0, | |
| "loss": 0.4513, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 3080, | |
| "total_flos": 7.864751040233472e+16, | |
| "train_loss": 0.5693606966501706, | |
| "train_runtime": 1451.854, | |
| "train_samples_per_second": 8.486, | |
| "train_steps_per_second": 2.121 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3080, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 7.864751040233472e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |