| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 13.0, | |
| "eval_steps": 500, | |
| "global_step": 130, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 6.427576939603807, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 1.0366, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 6.949204825256894, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 1.0581, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 6.602063062056644, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 1.0347, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 6.170546559184251, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.0547, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 4.592112691051119, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 0.9809, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 3.660644378455656, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 1.005, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 2.7822380588657314, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 0.9507, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.4593261742846764, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.9218, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 4.065863581987457, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 0.9427, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.050375723195787, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.9662, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 3.341270383897445, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 0.8789, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 3.2579847654666634, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.8938, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 2.52718940415014, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8731, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 1.8935708864959517, | |
| "learning_rate": 9.998197638354428e-06, | |
| "loss": 0.8488, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.604825554990027, | |
| "learning_rate": 9.992791852820709e-06, | |
| "loss": 0.807, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1.61804116528192, | |
| "learning_rate": 9.983786540671052e-06, | |
| "loss": 0.7946, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 1.3650017675416546, | |
| "learning_rate": 9.971188194237141e-06, | |
| "loss": 0.8041, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 1.1051598088407006, | |
| "learning_rate": 9.955005896229543e-06, | |
| "loss": 0.7773, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 1.177478351631234, | |
| "learning_rate": 9.935251313189564e-06, | |
| "loss": 0.7369, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.287022188169648, | |
| "learning_rate": 9.911938687078324e-06, | |
| "loss": 0.764, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 1.1513904182806198, | |
| "learning_rate": 9.885084825009085e-06, | |
| "loss": 0.739, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 0.8378184128545715, | |
| "learning_rate": 9.854709087130261e-06, | |
| "loss": 0.7259, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 1.0949668166461517, | |
| "learning_rate": 9.820833372667813e-06, | |
| "loss": 0.6721, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.0217091291906188, | |
| "learning_rate": 9.783482104137127e-06, | |
| "loss": 0.6935, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.9873556246882189, | |
| "learning_rate": 9.742682209735727e-06, | |
| "loss": 0.6811, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 1.0032737740907482, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.7031, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 0.8951917372556527, | |
| "learning_rate": 9.650856666246693e-06, | |
| "loss": 0.6736, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.8978445347997116, | |
| "learning_rate": 9.599897218294122e-06, | |
| "loss": 0.6993, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 1.0334419489190376, | |
| "learning_rate": 9.54562149901362e-06, | |
| "loss": 0.631, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.848796104093161, | |
| "learning_rate": 9.488068638195072e-06, | |
| "loss": 0.6274, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.7412909357921983, | |
| "learning_rate": 9.427280128266049e-06, | |
| "loss": 0.5814, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.7987332233836216, | |
| "learning_rate": 9.363299794378072e-06, | |
| "loss": 0.601, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 0.7431232648462397, | |
| "learning_rate": 9.296173762811084e-06, | |
| "loss": 0.5563, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 0.7192385108701194, | |
| "learning_rate": 9.225950427718974e-06, | |
| "loss": 0.584, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 0.7663617730407009, | |
| "learning_rate": 9.152680416240059e-06, | |
| "loss": 0.6227, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.7025115415032245, | |
| "learning_rate": 9.076416551997721e-06, | |
| "loss": 0.5874, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 0.7969651245220044, | |
| "learning_rate": 8.997213817017508e-06, | |
| "loss": 0.6063, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.7089588397644075, | |
| "learning_rate": 8.915129312088112e-06, | |
| "loss": 0.5779, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.794628317575661, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.5842, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.7453080274347142, | |
| "learning_rate": 8.742553740855507e-06, | |
| "loss": 0.5703, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 0.7001670135651602, | |
| "learning_rate": 8.652187091988516e-06, | |
| "loss": 0.4658, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 0.751229569949347, | |
| "learning_rate": 8.559187418346703e-06, | |
| "loss": 0.5244, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 0.726611828371733, | |
| "learning_rate": 8.463621767547998e-06, | |
| "loss": 0.5293, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 0.7176265166856064, | |
| "learning_rate": 8.36555903713785e-06, | |
| "loss": 0.5018, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 0.6887105692961704, | |
| "learning_rate": 8.265069924917925e-06, | |
| "loss": 0.5149, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 0.691663822333585, | |
| "learning_rate": 8.162226877976886e-06, | |
| "loss": 0.4888, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 0.6782822066608379, | |
| "learning_rate": 8.057104040460062e-06, | |
| "loss": 0.515, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.7535378251785477, | |
| "learning_rate": 7.949777200115617e-06, | |
| "loss": 0.4995, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 0.73942606231889, | |
| "learning_rate": 7.84032373365578e-06, | |
| "loss": 0.4939, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.5953765664012706, | |
| "learning_rate": 7.728822550972523e-06, | |
| "loss": 0.4581, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 0.7338494712254903, | |
| "learning_rate": 7.615354038247889e-06, | |
| "loss": 0.4306, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 0.6864950044935947, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.4396, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "grad_norm": 0.7507014982287982, | |
| "learning_rate": 7.382843600106539e-06, | |
| "loss": 0.4318, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "grad_norm": 0.6697991705708543, | |
| "learning_rate": 7.263969301848188e-06, | |
| "loss": 0.4157, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 0.6831422856253547, | |
| "learning_rate": 7.143462807015271e-06, | |
| "loss": 0.3936, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 0.6650630706492774, | |
| "learning_rate": 7.021410994121525e-06, | |
| "loss": 0.4256, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "grad_norm": 0.644108167018625, | |
| "learning_rate": 6.897901855769483e-06, | |
| "loss": 0.4078, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "grad_norm": 0.6324541017780231, | |
| "learning_rate": 6.773024435212678e-06, | |
| "loss": 0.3917, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.9, | |
| "grad_norm": 0.7498556087646469, | |
| "learning_rate": 6.646868762160399e-06, | |
| "loss": 0.4081, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.5845553692872227, | |
| "learning_rate": 6.519525787871235e-06, | |
| "loss": 0.4124, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "grad_norm": 0.8186143631248243, | |
| "learning_rate": 6.391087319582264e-06, | |
| "loss": 0.3681, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "grad_norm": 0.7007622328787126, | |
| "learning_rate": 6.261645954321109e-06, | |
| "loss": 0.3807, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.3, | |
| "grad_norm": 0.7369673161549185, | |
| "learning_rate": 6.131295012148613e-06, | |
| "loss": 0.3679, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 0.9695072418581587, | |
| "learning_rate": 6.000128468880223e-06, | |
| "loss": 0.3161, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "grad_norm": 0.6682687764823072, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.3532, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "grad_norm": 0.8132268408392879, | |
| "learning_rate": 5.735727354158581e-06, | |
| "loss": 0.3086, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.7, | |
| "grad_norm": 0.7513667140250347, | |
| "learning_rate": 5.6026834012766155e-06, | |
| "loss": 0.3446, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "grad_norm": 0.706377907830683, | |
| "learning_rate": 5.469204947015897e-06, | |
| "loss": 0.3356, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.9, | |
| "grad_norm": 0.7394601784180472, | |
| "learning_rate": 5.335388221955012e-06, | |
| "loss": 0.3013, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.7084866266508238, | |
| "learning_rate": 5.201329700547077e-06, | |
| "loss": 0.3354, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "grad_norm": 0.7388019162961375, | |
| "learning_rate": 5.067126031566988e-06, | |
| "loss": 0.2859, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 0.7594555147392841, | |
| "learning_rate": 4.932873968433014e-06, | |
| "loss": 0.2632, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.3, | |
| "grad_norm": 0.7279673465106814, | |
| "learning_rate": 4.798670299452926e-06, | |
| "loss": 0.3063, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "grad_norm": 0.9319655660378228, | |
| "learning_rate": 4.664611778044988e-06, | |
| "loss": 0.2861, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.6704570547991645, | |
| "learning_rate": 4.530795052984104e-06, | |
| "loss": 0.2962, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 0.7427352789176518, | |
| "learning_rate": 4.397316598723385e-06, | |
| "loss": 0.2878, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.7, | |
| "grad_norm": 0.7005963412829718, | |
| "learning_rate": 4.264272645841419e-06, | |
| "loss": 0.2556, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "grad_norm": 0.7579150401497047, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.2579, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.9, | |
| "grad_norm": 0.6563495169422253, | |
| "learning_rate": 3.999871531119779e-06, | |
| "loss": 0.2901, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.6379617571256964, | |
| "learning_rate": 3.86870498785139e-06, | |
| "loss": 0.272, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.1, | |
| "grad_norm": 0.6256577497024163, | |
| "learning_rate": 3.7383540456788915e-06, | |
| "loss": 0.2587, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "grad_norm": 0.6195096843246825, | |
| "learning_rate": 3.6089126804177373e-06, | |
| "loss": 0.2281, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.3, | |
| "grad_norm": 0.7270579247394137, | |
| "learning_rate": 3.480474212128766e-06, | |
| "loss": 0.2034, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "grad_norm": 0.6517246776245681, | |
| "learning_rate": 3.3531312378396026e-06, | |
| "loss": 0.2482, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 0.6810587187844922, | |
| "learning_rate": 3.226975564787322e-06, | |
| "loss": 0.2338, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "grad_norm": 0.7137913157718943, | |
| "learning_rate": 3.1020981442305187e-06, | |
| "loss": 0.2064, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "grad_norm": 0.6602685740768546, | |
| "learning_rate": 2.978589005878476e-06, | |
| "loss": 0.2252, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "grad_norm": 0.5924819653887085, | |
| "learning_rate": 2.8565371929847286e-06, | |
| "loss": 0.2204, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.9, | |
| "grad_norm": 0.6840887784565872, | |
| "learning_rate": 2.736030698151815e-06, | |
| "loss": 0.1918, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.6013103328375718, | |
| "learning_rate": 2.6171563998934605e-06, | |
| "loss": 0.2441, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.1, | |
| "grad_norm": 0.7225682019428351, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.1675, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "grad_norm": 0.5758913271702294, | |
| "learning_rate": 2.384645961752113e-06, | |
| "loss": 0.2107, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.3, | |
| "grad_norm": 0.681785905225741, | |
| "learning_rate": 2.2711774490274767e-06, | |
| "loss": 0.2226, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.4, | |
| "grad_norm": 0.7255894257677337, | |
| "learning_rate": 2.159676266344222e-06, | |
| "loss": 0.1899, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "grad_norm": 0.6465026241868481, | |
| "learning_rate": 2.050222799884387e-06, | |
| "loss": 0.186, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 0.6475959452445863, | |
| "learning_rate": 1.942895959539939e-06, | |
| "loss": 0.1872, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 9.7, | |
| "grad_norm": 0.6652079497145243, | |
| "learning_rate": 1.8377731220231144e-06, | |
| "loss": 0.1791, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "grad_norm": 0.6815536749526544, | |
| "learning_rate": 1.7349300750820758e-06, | |
| "loss": 0.1687, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.9, | |
| "grad_norm": 0.5668479155062388, | |
| "learning_rate": 1.6344409628621482e-06, | |
| "loss": 0.2108, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.5813922276526535, | |
| "learning_rate": 1.5363782324520033e-06, | |
| "loss": 0.1573, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.1, | |
| "grad_norm": 0.5806763330407928, | |
| "learning_rate": 1.4408125816532981e-06, | |
| "loss": 0.1677, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "grad_norm": 0.553482442243413, | |
| "learning_rate": 1.347812908011485e-06, | |
| "loss": 0.1523, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.3, | |
| "grad_norm": 0.5895044868296303, | |
| "learning_rate": 1.257446259144494e-06, | |
| "loss": 0.1652, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "grad_norm": 0.6728282034435286, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.1876, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 10.5, | |
| "grad_norm": 0.6168559494592303, | |
| "learning_rate": 1.0848706879118893e-06, | |
| "loss": 0.1716, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 10.6, | |
| "grad_norm": 0.5766037147981325, | |
| "learning_rate": 1.0027861829824953e-06, | |
| "loss": 0.1777, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 10.7, | |
| "grad_norm": 0.5606241075736775, | |
| "learning_rate": 9.235834480022788e-07, | |
| "loss": 0.1431, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "grad_norm": 0.5796701979793849, | |
| "learning_rate": 8.473195837599419e-07, | |
| "loss": 0.1456, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.9, | |
| "grad_norm": 0.5441073434898442, | |
| "learning_rate": 7.740495722810271e-07, | |
| "loss": 0.1633, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.551884623659146, | |
| "learning_rate": 7.03826237188916e-07, | |
| "loss": 0.1602, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.1, | |
| "grad_norm": 0.6712424096431827, | |
| "learning_rate": 6.367002056219285e-07, | |
| "loss": 0.1361, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "grad_norm": 0.574081085459768, | |
| "learning_rate": 5.727198717339511e-07, | |
| "loss": 0.1401, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.3, | |
| "grad_norm": 0.5762711416493509, | |
| "learning_rate": 5.119313618049309e-07, | |
| "loss": 0.1718, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.4, | |
| "grad_norm": 0.5174473303370842, | |
| "learning_rate": 4.54378500986381e-07, | |
| "loss": 0.1434, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 11.5, | |
| "grad_norm": 0.5029979064526134, | |
| "learning_rate": 4.001027817058789e-07, | |
| "loss": 0.1416, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "grad_norm": 0.6166403006699419, | |
| "learning_rate": 3.49143333753309e-07, | |
| "loss": 0.1689, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 11.7, | |
| "grad_norm": 0.6209212510291084, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.1435, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 11.8, | |
| "grad_norm": 0.6044047129226829, | |
| "learning_rate": 2.573177902642726e-07, | |
| "loss": 0.1555, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 11.9, | |
| "grad_norm": 0.5486653051881393, | |
| "learning_rate": 2.1651789586287442e-07, | |
| "loss": 0.1464, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.5062585328069695, | |
| "learning_rate": 1.7916662733218848e-07, | |
| "loss": 0.1498, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 12.1, | |
| "grad_norm": 0.5443974723830177, | |
| "learning_rate": 1.4529091286973994e-07, | |
| "loss": 0.1509, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 12.2, | |
| "grad_norm": 0.4678442086479432, | |
| "learning_rate": 1.1491517499091498e-07, | |
| "loss": 0.1291, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 12.3, | |
| "grad_norm": 0.4639323476583933, | |
| "learning_rate": 8.80613129216762e-08, | |
| "loss": 0.1218, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "grad_norm": 0.5247097751710195, | |
| "learning_rate": 6.474868681043578e-08, | |
| "loss": 0.1551, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 0.45747930798066005, | |
| "learning_rate": 4.499410377045765e-08, | |
| "loss": 0.1503, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 12.6, | |
| "grad_norm": 0.48362972580828334, | |
| "learning_rate": 2.8811805762860578e-08, | |
| "loss": 0.1581, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 12.7, | |
| "grad_norm": 0.46925396240140826, | |
| "learning_rate": 1.6213459328950355e-08, | |
| "loss": 0.1525, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "grad_norm": 0.47130993477619143, | |
| "learning_rate": 7.2081471792911914e-09, | |
| "loss": 0.1541, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 12.9, | |
| "grad_norm": 0.4558593399302876, | |
| "learning_rate": 1.8023616455731253e-09, | |
| "loss": 0.1356, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.46775589777469023, | |
| "learning_rate": 0.0, | |
| "loss": 0.1601, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "step": 130, | |
| "total_flos": 7.074897800147763e+16, | |
| "train_loss": 0.4226666835638193, | |
| "train_runtime": 2279.8038, | |
| "train_samples_per_second": 1.802, | |
| "train_steps_per_second": 0.057 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 130, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 13, | |
| "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": 7.074897800147763e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |