| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 16.93658536585366, | |
| "eval_steps": 500, | |
| "global_step": 136, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11707317073170732, | |
| "grad_norm": 1.713741660118103, | |
| "learning_rate": 7.142857142857143e-07, | |
| "loss": 0.391, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.23414634146341465, | |
| "grad_norm": 1.7321815490722656, | |
| "learning_rate": 1.4285714285714286e-06, | |
| "loss": 0.3876, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.35121951219512193, | |
| "grad_norm": 1.869097352027893, | |
| "learning_rate": 2.1428571428571427e-06, | |
| "loss": 0.4163, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4682926829268293, | |
| "grad_norm": 1.741655707359314, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 0.4025, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5853658536585366, | |
| "grad_norm": 1.5887295007705688, | |
| "learning_rate": 3.5714285714285718e-06, | |
| "loss": 0.383, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.7024390243902439, | |
| "grad_norm": 1.5869324207305908, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 0.3968, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.8195121951219512, | |
| "grad_norm": 1.3928492069244385, | |
| "learning_rate": 5e-06, | |
| "loss": 0.3868, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.9365853658536586, | |
| "grad_norm": 1.317595362663269, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 0.3871, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.1170731707317074, | |
| "grad_norm": 1.3958970308303833, | |
| "learning_rate": 6.4285714285714295e-06, | |
| "loss": 0.7529, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.2341463414634146, | |
| "grad_norm": 0.588860273361206, | |
| "learning_rate": 7.1428571428571436e-06, | |
| "loss": 0.361, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.3512195121951218, | |
| "grad_norm": 0.5455438494682312, | |
| "learning_rate": 7.857142857142858e-06, | |
| "loss": 0.3647, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.4682926829268292, | |
| "grad_norm": 1.2507004737854004, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 0.3795, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.5853658536585367, | |
| "grad_norm": 1.3338210582733154, | |
| "learning_rate": 9.285714285714288e-06, | |
| "loss": 0.3783, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.7024390243902439, | |
| "grad_norm": 1.2802133560180664, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3637, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.819512195121951, | |
| "grad_norm": 1.1926186084747314, | |
| "learning_rate": 9.998342337571566e-06, | |
| "loss": 0.3663, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.9365853658536585, | |
| "grad_norm": 0.8796733021736145, | |
| "learning_rate": 9.993370449424153e-06, | |
| "loss": 0.3446, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.1170731707317074, | |
| "grad_norm": 1.2581405639648438, | |
| "learning_rate": 9.985087632242634e-06, | |
| "loss": 0.6666, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 2.234146341463415, | |
| "grad_norm": 0.6097764372825623, | |
| "learning_rate": 9.973499378072947e-06, | |
| "loss": 0.3404, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 2.351219512195122, | |
| "grad_norm": 0.5183754563331604, | |
| "learning_rate": 9.958613370680507e-06, | |
| "loss": 0.331, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.4682926829268292, | |
| "grad_norm": 0.5353206992149353, | |
| "learning_rate": 9.940439480455386e-06, | |
| "loss": 0.3364, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.5853658536585367, | |
| "grad_norm": 0.5248769521713257, | |
| "learning_rate": 9.918989757867584e-06, | |
| "loss": 0.3289, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.7024390243902436, | |
| "grad_norm": 0.5104326009750366, | |
| "learning_rate": 9.89427842547679e-06, | |
| "loss": 0.3207, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.819512195121951, | |
| "grad_norm": 0.429746150970459, | |
| "learning_rate": 9.866321868501914e-06, | |
| "loss": 0.3354, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.9365853658536585, | |
| "grad_norm": 0.4242963492870331, | |
| "learning_rate": 9.835138623956603e-06, | |
| "loss": 0.3149, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 3.1170731707317074, | |
| "grad_norm": 0.8024505972862244, | |
| "learning_rate": 9.80074936835801e-06, | |
| "loss": 0.6097, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.234146341463415, | |
| "grad_norm": 0.3315274715423584, | |
| "learning_rate": 9.763176904016914e-06, | |
| "loss": 0.3016, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 3.351219512195122, | |
| "grad_norm": 0.3314255475997925, | |
| "learning_rate": 9.722446143918307e-06, | |
| "loss": 0.2876, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 3.4682926829268292, | |
| "grad_norm": 0.30298393964767456, | |
| "learning_rate": 9.678584095202468e-06, | |
| "loss": 0.2912, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.5853658536585367, | |
| "grad_norm": 0.3056536614894867, | |
| "learning_rate": 9.631619841257477e-06, | |
| "loss": 0.2855, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.7024390243902436, | |
| "grad_norm": 0.3028172552585602, | |
| "learning_rate": 9.581584522435025e-06, | |
| "loss": 0.2848, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.819512195121951, | |
| "grad_norm": 0.3157103657722473, | |
| "learning_rate": 9.528511315402358e-06, | |
| "loss": 0.2891, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.9365853658536585, | |
| "grad_norm": 0.28788891434669495, | |
| "learning_rate": 9.472435411143979e-06, | |
| "loss": 0.274, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 4.117073170731707, | |
| "grad_norm": 0.562315821647644, | |
| "learning_rate": 9.413393991627737e-06, | |
| "loss": 0.5517, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 4.234146341463415, | |
| "grad_norm": 0.24586421251296997, | |
| "learning_rate": 9.351426205150778e-06, | |
| "loss": 0.2557, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 4.351219512195122, | |
| "grad_norm": 0.24129828810691833, | |
| "learning_rate": 9.286573140381663e-06, | |
| "loss": 0.2506, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 4.46829268292683, | |
| "grad_norm": 0.2212359756231308, | |
| "learning_rate": 9.218877799115929e-06, | |
| "loss": 0.2583, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 4.585365853658536, | |
| "grad_norm": 0.20721423625946045, | |
| "learning_rate": 9.148385067763094e-06, | |
| "loss": 0.2509, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 4.702439024390244, | |
| "grad_norm": 0.22884435951709747, | |
| "learning_rate": 9.075141687584056e-06, | |
| "loss": 0.248, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.819512195121951, | |
| "grad_norm": 0.24287782609462738, | |
| "learning_rate": 8.999196223698599e-06, | |
| "loss": 0.2443, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.9365853658536585, | |
| "grad_norm": 0.2220289707183838, | |
| "learning_rate": 8.920599032883553e-06, | |
| "loss": 0.235, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 5.117073170731707, | |
| "grad_norm": 0.4429185092449188, | |
| "learning_rate": 8.839402230183e-06, | |
| "loss": 0.4562, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 5.234146341463415, | |
| "grad_norm": 0.2300422042608261, | |
| "learning_rate": 8.755659654352599e-06, | |
| "loss": 0.2245, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 5.351219512195122, | |
| "grad_norm": 0.21638786792755127, | |
| "learning_rate": 8.669426832160997e-06, | |
| "loss": 0.2268, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 5.46829268292683, | |
| "grad_norm": 0.2475624680519104, | |
| "learning_rate": 8.580760941571968e-06, | |
| "loss": 0.2026, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 5.585365853658536, | |
| "grad_norm": 0.2283526360988617, | |
| "learning_rate": 8.489720773831717e-06, | |
| "loss": 0.203, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 5.702439024390244, | |
| "grad_norm": 0.19043877720832825, | |
| "learning_rate": 8.396366694486466e-06, | |
| "loss": 0.2039, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 5.819512195121951, | |
| "grad_norm": 0.2230452299118042, | |
| "learning_rate": 8.30076060335616e-06, | |
| "loss": 0.1957, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 5.9365853658536585, | |
| "grad_norm": 0.2054775357246399, | |
| "learning_rate": 8.202965893490877e-06, | |
| "loss": 0.202, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 6.117073170731707, | |
| "grad_norm": 0.19972442090511322, | |
| "learning_rate": 8.103047409137114e-06, | |
| "loss": 0.3737, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 6.234146341463415, | |
| "grad_norm": 0.20269611477851868, | |
| "learning_rate": 8.001071402741843e-06, | |
| "loss": 0.1804, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.351219512195122, | |
| "grad_norm": 0.2113686352968216, | |
| "learning_rate": 7.897105491022819e-06, | |
| "loss": 0.1755, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 6.46829268292683, | |
| "grad_norm": 0.19944241642951965, | |
| "learning_rate": 7.791218610134324e-06, | |
| "loss": 0.1576, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 6.585365853658536, | |
| "grad_norm": 0.1904984563589096, | |
| "learning_rate": 7.683480969958005e-06, | |
| "loss": 0.1586, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 6.702439024390244, | |
| "grad_norm": 0.19858157634735107, | |
| "learning_rate": 7.5739640075491546e-06, | |
| "loss": 0.1628, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 6.819512195121951, | |
| "grad_norm": 0.1883428692817688, | |
| "learning_rate": 7.462740339769323e-06, | |
| "loss": 0.1632, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 6.9365853658536585, | |
| "grad_norm": 0.17427878081798553, | |
| "learning_rate": 7.349883715136601e-06, | |
| "loss": 0.1653, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 7.117073170731707, | |
| "grad_norm": 0.37863096594810486, | |
| "learning_rate": 7.235468964925571e-06, | |
| "loss": 0.3005, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 7.234146341463415, | |
| "grad_norm": 0.1933353841304779, | |
| "learning_rate": 7.119571953549305e-06, | |
| "loss": 0.1439, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 7.351219512195122, | |
| "grad_norm": 0.1970248520374298, | |
| "learning_rate": 7.002269528256334e-06, | |
| "loss": 0.132, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 7.46829268292683, | |
| "grad_norm": 0.19500941038131714, | |
| "learning_rate": 6.883639468175926e-06, | |
| "loss": 0.1264, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.585365853658536, | |
| "grad_norm": 0.20704524219036102, | |
| "learning_rate": 6.763760432745475e-06, | |
| "loss": 0.135, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 7.702439024390244, | |
| "grad_norm": 0.21069225668907166, | |
| "learning_rate": 6.6427119095541745e-06, | |
| "loss": 0.1168, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 7.819512195121951, | |
| "grad_norm": 0.185212641954422, | |
| "learning_rate": 6.520574161637591e-06, | |
| "loss": 0.1252, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 7.9365853658536585, | |
| "grad_norm": 0.18015943467617035, | |
| "learning_rate": 6.397428174258048e-06, | |
| "loss": 0.1249, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 8.117073170731707, | |
| "grad_norm": 0.5003377199172974, | |
| "learning_rate": 6.273355601206143e-06, | |
| "loss": 0.2434, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 8.234146341463415, | |
| "grad_norm": 0.2011335790157318, | |
| "learning_rate": 6.148438710658979e-06, | |
| "loss": 0.0987, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 8.351219512195122, | |
| "grad_norm": 0.1703123301267624, | |
| "learning_rate": 6.022760330631006e-06, | |
| "loss": 0.1057, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 8.46829268292683, | |
| "grad_norm": 0.17905612289905548, | |
| "learning_rate": 5.896403794053679e-06, | |
| "loss": 0.1019, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 8.585365853658537, | |
| "grad_norm": 0.22483955323696136, | |
| "learning_rate": 5.76945288352031e-06, | |
| "loss": 0.1018, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 8.702439024390245, | |
| "grad_norm": 0.18365544080734253, | |
| "learning_rate": 5.641991775732756e-06, | |
| "loss": 0.0906, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 8.819512195121952, | |
| "grad_norm": 0.16530556976795197, | |
| "learning_rate": 5.514104985686802e-06, | |
| "loss": 0.0961, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 8.93658536585366, | |
| "grad_norm": 0.18940307199954987, | |
| "learning_rate": 5.385877310633233e-06, | |
| "loss": 0.0921, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 9.117073170731707, | |
| "grad_norm": 0.3870483934879303, | |
| "learning_rate": 5.257393773851733e-06, | |
| "loss": 0.1844, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 9.234146341463415, | |
| "grad_norm": 0.19212201237678528, | |
| "learning_rate": 5.1287395682749444e-06, | |
| "loss": 0.0799, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 9.351219512195122, | |
| "grad_norm": 0.19690005481243134, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0757, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 9.46829268292683, | |
| "grad_norm": 0.18531833589076996, | |
| "learning_rate": 4.871260431725058e-06, | |
| "loss": 0.0746, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 9.585365853658537, | |
| "grad_norm": 0.2481042891740799, | |
| "learning_rate": 4.742606226148268e-06, | |
| "loss": 0.064, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 9.702439024390245, | |
| "grad_norm": 0.1442754566669464, | |
| "learning_rate": 4.614122689366769e-06, | |
| "loss": 0.0741, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 9.819512195121952, | |
| "grad_norm": 0.1746453046798706, | |
| "learning_rate": 4.485895014313198e-06, | |
| "loss": 0.0712, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 9.93658536585366, | |
| "grad_norm": 0.19312545657157898, | |
| "learning_rate": 4.358008224267245e-06, | |
| "loss": 0.0747, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 10.117073170731707, | |
| "grad_norm": 0.4135829508304596, | |
| "learning_rate": 4.230547116479691e-06, | |
| "loss": 0.1344, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 10.234146341463415, | |
| "grad_norm": 0.14636236429214478, | |
| "learning_rate": 4.103596205946323e-06, | |
| "loss": 0.0564, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 10.351219512195122, | |
| "grad_norm": 0.18952858448028564, | |
| "learning_rate": 3.977239669368998e-06, | |
| "loss": 0.0567, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 10.46829268292683, | |
| "grad_norm": 0.23925091326236725, | |
| "learning_rate": 3.851561289341023e-06, | |
| "loss": 0.0575, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 10.585365853658537, | |
| "grad_norm": 0.20485427975654602, | |
| "learning_rate": 3.726644398793857e-06, | |
| "loss": 0.0571, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 10.702439024390245, | |
| "grad_norm": 0.15516437590122223, | |
| "learning_rate": 3.6025718257419532e-06, | |
| "loss": 0.0515, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 10.819512195121952, | |
| "grad_norm": 0.20174065232276917, | |
| "learning_rate": 3.4794258383624115e-06, | |
| "loss": 0.0538, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 10.93658536585366, | |
| "grad_norm": 0.23884522914886475, | |
| "learning_rate": 3.3572880904458267e-06, | |
| "loss": 0.0557, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 11.117073170731707, | |
| "grad_norm": 0.3778754770755768, | |
| "learning_rate": 3.236239567254526e-06, | |
| "loss": 0.1121, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 11.234146341463415, | |
| "grad_norm": 0.15436743199825287, | |
| "learning_rate": 3.116360531824074e-06, | |
| "loss": 0.0422, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 11.351219512195122, | |
| "grad_norm": 0.14751587808132172, | |
| "learning_rate": 2.997730471743667e-06, | |
| "loss": 0.0417, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 11.46829268292683, | |
| "grad_norm": 0.23487572371959686, | |
| "learning_rate": 2.880428046450697e-06, | |
| "loss": 0.0396, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 11.585365853658537, | |
| "grad_norm": 0.2069467157125473, | |
| "learning_rate": 2.7645310350744296e-06, | |
| "loss": 0.0426, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 11.702439024390245, | |
| "grad_norm": 0.15420763194561005, | |
| "learning_rate": 2.6501162848634023e-06, | |
| "loss": 0.0413, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 11.819512195121952, | |
| "grad_norm": 0.1196109727025032, | |
| "learning_rate": 2.537259660230679e-06, | |
| "loss": 0.0404, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 11.93658536585366, | |
| "grad_norm": 0.1467064768075943, | |
| "learning_rate": 2.426035992450848e-06, | |
| "loss": 0.0407, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 12.117073170731707, | |
| "grad_norm": 0.19361571967601776, | |
| "learning_rate": 2.316519030041998e-06, | |
| "loss": 0.0791, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 12.234146341463415, | |
| "grad_norm": 0.13011707365512848, | |
| "learning_rate": 2.2087813898656775e-06, | |
| "loss": 0.0346, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 12.351219512195122, | |
| "grad_norm": 0.10667014867067337, | |
| "learning_rate": 2.102894508977182e-06, | |
| "loss": 0.0327, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 12.46829268292683, | |
| "grad_norm": 0.14411939680576324, | |
| "learning_rate": 1.9989285972581595e-06, | |
| "loss": 0.0326, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 12.585365853658537, | |
| "grad_norm": 0.19949135184288025, | |
| "learning_rate": 1.896952590862886e-06, | |
| "loss": 0.0343, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 12.702439024390245, | |
| "grad_norm": 0.18770211935043335, | |
| "learning_rate": 1.7970341065091246e-06, | |
| "loss": 0.031, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 12.819512195121952, | |
| "grad_norm": 0.1456199288368225, | |
| "learning_rate": 1.699239396643841e-06, | |
| "loss": 0.0283, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 12.93658536585366, | |
| "grad_norm": 0.140044167637825, | |
| "learning_rate": 1.6036333055135345e-06, | |
| "loss": 0.0331, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 13.117073170731707, | |
| "grad_norm": 0.3027644157409668, | |
| "learning_rate": 1.5102792261682813e-06, | |
| "loss": 0.0651, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 13.234146341463415, | |
| "grad_norm": 0.16215696930885315, | |
| "learning_rate": 1.4192390584280347e-06, | |
| "loss": 0.0296, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 13.351219512195122, | |
| "grad_norm": 0.1622275710105896, | |
| "learning_rate": 1.330573167839005e-06, | |
| "loss": 0.0265, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 13.46829268292683, | |
| "grad_norm": 0.13922300934791565, | |
| "learning_rate": 1.2443403456474017e-06, | |
| "loss": 0.0277, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 13.585365853658537, | |
| "grad_norm": 0.08746293187141418, | |
| "learning_rate": 1.1605977698170001e-06, | |
| "loss": 0.0237, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 13.702439024390245, | |
| "grad_norm": 0.09428137540817261, | |
| "learning_rate": 1.0794009671164484e-06, | |
| "loss": 0.0255, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 13.819512195121952, | |
| "grad_norm": 0.10664989054203033, | |
| "learning_rate": 1.0008037763014033e-06, | |
| "loss": 0.0265, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 13.93658536585366, | |
| "grad_norm": 0.12778587639331818, | |
| "learning_rate": 9.248583124159438e-07, | |
| "loss": 0.0265, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 14.117073170731707, | |
| "grad_norm": 0.25158995389938354, | |
| "learning_rate": 8.516149322369055e-07, | |
| "loss": 0.0504, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 14.234146341463415, | |
| "grad_norm": 0.08846470713615417, | |
| "learning_rate": 7.811222008840719e-07, | |
| "loss": 0.0218, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 14.351219512195122, | |
| "grad_norm": 0.08048608899116516, | |
| "learning_rate": 7.13426859618338e-07, | |
| "loss": 0.0234, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 14.46829268292683, | |
| "grad_norm": 0.08340264111757278, | |
| "learning_rate": 6.485737948492237e-07, | |
| "loss": 0.0237, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 14.585365853658537, | |
| "grad_norm": 0.08076529949903488, | |
| "learning_rate": 5.866060083722624e-07, | |
| "loss": 0.0212, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 14.702439024390245, | |
| "grad_norm": 0.08138881623744965, | |
| "learning_rate": 5.275645888560233e-07, | |
| "loss": 0.0221, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 14.819512195121952, | |
| "grad_norm": 0.08238395303487778, | |
| "learning_rate": 4.71488684597643e-07, | |
| "loss": 0.0222, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 14.93658536585366, | |
| "grad_norm": 0.08457454293966293, | |
| "learning_rate": 4.184154775649768e-07, | |
| "loss": 0.0231, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 15.117073170731707, | |
| "grad_norm": 0.21646344661712646, | |
| "learning_rate": 3.683801587425251e-07, | |
| "loss": 0.0459, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 15.234146341463415, | |
| "grad_norm": 0.08210847526788712, | |
| "learning_rate": 3.214159047975324e-07, | |
| "loss": 0.0243, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 15.351219512195122, | |
| "grad_norm": 0.0840378999710083, | |
| "learning_rate": 2.7755385608169374e-07, | |
| "loss": 0.0205, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 15.46829268292683, | |
| "grad_norm": 0.07520610839128494, | |
| "learning_rate": 2.368230959830875e-07, | |
| "loss": 0.0196, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 15.585365853658537, | |
| "grad_norm": 0.07876352220773697, | |
| "learning_rate": 1.992506316419912e-07, | |
| "loss": 0.0204, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 15.702439024390245, | |
| "grad_norm": 0.07992194592952728, | |
| "learning_rate": 1.6486137604339813e-07, | |
| "loss": 0.0211, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 15.819512195121952, | |
| "grad_norm": 0.08172820508480072, | |
| "learning_rate": 1.3367813149808728e-07, | |
| "loss": 0.0229, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 15.93658536585366, | |
| "grad_norm": 0.07874112576246262, | |
| "learning_rate": 1.0572157452321097e-07, | |
| "loss": 0.0221, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 16.117073170731707, | |
| "grad_norm": 0.19976511597633362, | |
| "learning_rate": 8.101024213241826e-08, | |
| "loss": 0.0391, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 16.234146341463415, | |
| "grad_norm": 0.0725630447268486, | |
| "learning_rate": 5.9560519544614725e-08, | |
| "loss": 0.0225, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 16.351219512195122, | |
| "grad_norm": 0.07620209455490112, | |
| "learning_rate": 4.138662931949255e-08, | |
| "loss": 0.0197, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 16.46829268292683, | |
| "grad_norm": 0.06960424035787582, | |
| "learning_rate": 2.6500621927054716e-08, | |
| "loss": 0.0205, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 16.585365853658537, | |
| "grad_norm": 0.0744452178478241, | |
| "learning_rate": 1.4912367757366485e-08, | |
| "loss": 0.0204, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 16.702439024390245, | |
| "grad_norm": 0.07866977900266647, | |
| "learning_rate": 6.629550575847355e-09, | |
| "loss": 0.0212, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 16.819512195121952, | |
| "grad_norm": 0.07277688384056091, | |
| "learning_rate": 1.657662428434792e-09, | |
| "loss": 0.0186, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 16.93658536585366, | |
| "grad_norm": 0.06631285697221756, | |
| "learning_rate": 0.0, | |
| "loss": 0.0242, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 16.93658536585366, | |
| "step": 136, | |
| "total_flos": 483897984679936.0, | |
| "train_loss": 0.16699845409568617, | |
| "train_runtime": 26245.2704, | |
| "train_samples_per_second": 0.531, | |
| "train_steps_per_second": 0.005 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 136, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 17, | |
| "save_steps": 10000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 483897984679936.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |