| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 1.1846153846153846, |
| "eval_steps": 500, |
| "global_step": 180, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.006593406593406593, |
| "grad_norm": 1.9395878314971924, |
| "learning_rate": 0.0, |
| "loss": 3.0255, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.013186813186813187, |
| "grad_norm": 2.0909342765808105, |
| "learning_rate": 4e-05, |
| "loss": 3.2722, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.01978021978021978, |
| "grad_norm": 2.2211079597473145, |
| "learning_rate": 8e-05, |
| "loss": 3.0812, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.026373626373626374, |
| "grad_norm": 1.576534390449524, |
| "learning_rate": 0.00012, |
| "loss": 2.3772, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.03296703296703297, |
| "grad_norm": 2.288299560546875, |
| "learning_rate": 0.00016, |
| "loss": 2.7816, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.03956043956043956, |
| "grad_norm": 2.0976123809814453, |
| "learning_rate": 0.0002, |
| "loss": 1.8456, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.046153846153846156, |
| "grad_norm": 2.893791437149048, |
| "learning_rate": 0.00019955654101995565, |
| "loss": 1.9392, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.05274725274725275, |
| "grad_norm": 1.5747392177581787, |
| "learning_rate": 0.00019911308203991133, |
| "loss": 1.2524, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.05934065934065934, |
| "grad_norm": 0.9431101083755493, |
| "learning_rate": 0.00019866962305986697, |
| "loss": 1.2513, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.06593406593406594, |
| "grad_norm": 1.4205081462860107, |
| "learning_rate": 0.00019822616407982261, |
| "loss": 1.5582, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.07252747252747253, |
| "grad_norm": 1.0894768238067627, |
| "learning_rate": 0.00019778270509977829, |
| "loss": 1.2454, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.07912087912087912, |
| "grad_norm": 0.5371621251106262, |
| "learning_rate": 0.00019733924611973393, |
| "loss": 1.1853, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.08571428571428572, |
| "grad_norm": 0.5113345384597778, |
| "learning_rate": 0.0001968957871396896, |
| "loss": 1.0215, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.09230769230769231, |
| "grad_norm": 0.5522609353065491, |
| "learning_rate": 0.00019645232815964525, |
| "loss": 1.0844, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.0989010989010989, |
| "grad_norm": 0.5567106604576111, |
| "learning_rate": 0.00019600886917960092, |
| "loss": 1.28, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.1054945054945055, |
| "grad_norm": 0.5977707505226135, |
| "learning_rate": 0.00019556541019955653, |
| "loss": 1.3916, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.11208791208791209, |
| "grad_norm": 0.4428749680519104, |
| "learning_rate": 0.0001951219512195122, |
| "loss": 1.0281, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.11868131868131868, |
| "grad_norm": 0.39748522639274597, |
| "learning_rate": 0.00019467849223946785, |
| "loss": 0.9917, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.12527472527472527, |
| "grad_norm": 0.4602974057197571, |
| "learning_rate": 0.00019423503325942352, |
| "loss": 0.9289, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.13186813186813187, |
| "grad_norm": 0.46893274784088135, |
| "learning_rate": 0.00019379157427937917, |
| "loss": 1.0148, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.13846153846153847, |
| "grad_norm": 0.3629504144191742, |
| "learning_rate": 0.00019334811529933484, |
| "loss": 1.0688, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.14505494505494507, |
| "grad_norm": 0.3591447174549103, |
| "learning_rate": 0.00019290465631929045, |
| "loss": 1.1489, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.15164835164835164, |
| "grad_norm": 0.38264843821525574, |
| "learning_rate": 0.00019246119733924613, |
| "loss": 1.0042, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.15824175824175823, |
| "grad_norm": 0.30332762002944946, |
| "learning_rate": 0.00019201773835920177, |
| "loss": 1.0645, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.16483516483516483, |
| "grad_norm": 0.36956873536109924, |
| "learning_rate": 0.00019157427937915744, |
| "loss": 0.8483, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.17142857142857143, |
| "grad_norm": 0.35823047161102295, |
| "learning_rate": 0.00019113082039911309, |
| "loss": 0.9212, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.17802197802197803, |
| "grad_norm": 0.31688058376312256, |
| "learning_rate": 0.00019068736141906876, |
| "loss": 0.8109, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.18461538461538463, |
| "grad_norm": 0.34314897656440735, |
| "learning_rate": 0.0001902439024390244, |
| "loss": 0.9747, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.1912087912087912, |
| "grad_norm": 0.39364388585090637, |
| "learning_rate": 0.00018980044345898005, |
| "loss": 1.0013, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.1978021978021978, |
| "grad_norm": 0.30356618762016296, |
| "learning_rate": 0.00018935698447893572, |
| "loss": 0.9765, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.2043956043956044, |
| "grad_norm": 0.34586668014526367, |
| "learning_rate": 0.00018891352549889136, |
| "loss": 1.0919, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.210989010989011, |
| "grad_norm": 0.32317039370536804, |
| "learning_rate": 0.00018847006651884703, |
| "loss": 1.0067, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.2175824175824176, |
| "grad_norm": 0.44286298751831055, |
| "learning_rate": 0.00018802660753880268, |
| "loss": 0.9623, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.22417582417582418, |
| "grad_norm": 0.3818933367729187, |
| "learning_rate": 0.00018758314855875832, |
| "loss": 1.0142, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.23076923076923078, |
| "grad_norm": 0.33253738284111023, |
| "learning_rate": 0.00018713968957871397, |
| "loss": 0.8074, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.23736263736263735, |
| "grad_norm": 0.3408537805080414, |
| "learning_rate": 0.00018669623059866964, |
| "loss": 1.0326, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.24395604395604395, |
| "grad_norm": 0.40733590722084045, |
| "learning_rate": 0.00018625277161862528, |
| "loss": 1.0927, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.25054945054945055, |
| "grad_norm": 0.36787742376327515, |
| "learning_rate": 0.00018580931263858095, |
| "loss": 1.0677, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.2571428571428571, |
| "grad_norm": 0.3505491614341736, |
| "learning_rate": 0.0001853658536585366, |
| "loss": 0.9553, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.26373626373626374, |
| "grad_norm": 0.3779163062572479, |
| "learning_rate": 0.00018492239467849224, |
| "loss": 0.921, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.2703296703296703, |
| "grad_norm": 0.3095817565917969, |
| "learning_rate": 0.0001844789356984479, |
| "loss": 0.8463, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.27692307692307694, |
| "grad_norm": 0.3786693513393402, |
| "learning_rate": 0.00018403547671840356, |
| "loss": 1.0101, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.2835164835164835, |
| "grad_norm": 0.2848946750164032, |
| "learning_rate": 0.0001835920177383592, |
| "loss": 1.1081, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.29010989010989013, |
| "grad_norm": 0.2794632613658905, |
| "learning_rate": 0.00018314855875831487, |
| "loss": 0.8742, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.2967032967032967, |
| "grad_norm": 0.33480748534202576, |
| "learning_rate": 0.00018270509977827052, |
| "loss": 0.9947, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.3032967032967033, |
| "grad_norm": 0.37888801097869873, |
| "learning_rate": 0.00018226164079822616, |
| "loss": 0.9493, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.3098901098901099, |
| "grad_norm": 0.39233142137527466, |
| "learning_rate": 0.00018181818181818183, |
| "loss": 1.0789, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.31648351648351647, |
| "grad_norm": 0.2909736931324005, |
| "learning_rate": 0.00018137472283813748, |
| "loss": 1.0518, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.3230769230769231, |
| "grad_norm": 0.2732703983783722, |
| "learning_rate": 0.00018093126385809312, |
| "loss": 0.9017, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.32967032967032966, |
| "grad_norm": 0.30006876587867737, |
| "learning_rate": 0.0001804878048780488, |
| "loss": 0.8561, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.3362637362637363, |
| "grad_norm": 0.34367501735687256, |
| "learning_rate": 0.00018004434589800444, |
| "loss": 1.0616, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.34285714285714286, |
| "grad_norm": 0.2993384003639221, |
| "learning_rate": 0.00017960088691796008, |
| "loss": 0.7779, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.34945054945054943, |
| "grad_norm": 0.3281186819076538, |
| "learning_rate": 0.00017915742793791575, |
| "loss": 0.7872, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.35604395604395606, |
| "grad_norm": 0.2709483206272125, |
| "learning_rate": 0.0001787139689578714, |
| "loss": 0.9384, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.3626373626373626, |
| "grad_norm": 0.3327484428882599, |
| "learning_rate": 0.00017827050997782707, |
| "loss": 0.9123, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.36923076923076925, |
| "grad_norm": 0.32308903336524963, |
| "learning_rate": 0.00017782705099778271, |
| "loss": 1.0042, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.3758241758241758, |
| "grad_norm": 0.2704257369041443, |
| "learning_rate": 0.00017738359201773839, |
| "loss": 0.9699, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.3824175824175824, |
| "grad_norm": 0.28141987323760986, |
| "learning_rate": 0.000176940133037694, |
| "loss": 0.8826, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.389010989010989, |
| "grad_norm": 0.35217514634132385, |
| "learning_rate": 0.00017649667405764967, |
| "loss": 1.0742, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.3956043956043956, |
| "grad_norm": 0.3242128789424896, |
| "learning_rate": 0.00017605321507760532, |
| "loss": 0.8373, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.4021978021978022, |
| "grad_norm": 0.29047495126724243, |
| "learning_rate": 0.000175609756097561, |
| "loss": 0.885, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.4087912087912088, |
| "grad_norm": 0.30491581559181213, |
| "learning_rate": 0.00017516629711751663, |
| "loss": 1.0813, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.4153846153846154, |
| "grad_norm": 0.34017154574394226, |
| "learning_rate": 0.0001747228381374723, |
| "loss": 1.0067, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.421978021978022, |
| "grad_norm": 0.2905332148075104, |
| "learning_rate": 0.00017427937915742792, |
| "loss": 0.8561, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.42857142857142855, |
| "grad_norm": 0.3342151343822479, |
| "learning_rate": 0.0001738359201773836, |
| "loss": 0.9917, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.4351648351648352, |
| "grad_norm": 0.35391831398010254, |
| "learning_rate": 0.00017339246119733924, |
| "loss": 1.1116, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.44175824175824174, |
| "grad_norm": 0.34148970246315, |
| "learning_rate": 0.0001729490022172949, |
| "loss": 0.9165, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.44835164835164837, |
| "grad_norm": 0.35057196021080017, |
| "learning_rate": 0.00017250554323725056, |
| "loss": 0.9318, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.45494505494505494, |
| "grad_norm": 0.3118966221809387, |
| "learning_rate": 0.00017206208425720623, |
| "loss": 0.9556, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.46153846153846156, |
| "grad_norm": 0.29268085956573486, |
| "learning_rate": 0.00017161862527716187, |
| "loss": 0.8563, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.46813186813186813, |
| "grad_norm": 0.3315921127796173, |
| "learning_rate": 0.00017117516629711752, |
| "loss": 1.1628, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.4747252747252747, |
| "grad_norm": 0.35345011949539185, |
| "learning_rate": 0.0001707317073170732, |
| "loss": 0.944, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.48131868131868133, |
| "grad_norm": 0.2805996835231781, |
| "learning_rate": 0.00017028824833702883, |
| "loss": 0.8882, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.4879120879120879, |
| "grad_norm": 0.385016530752182, |
| "learning_rate": 0.0001698447893569845, |
| "loss": 0.799, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.4945054945054945, |
| "grad_norm": 0.35097435116767883, |
| "learning_rate": 0.00016940133037694015, |
| "loss": 0.9332, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.5010989010989011, |
| "grad_norm": 0.2864893674850464, |
| "learning_rate": 0.00016895787139689582, |
| "loss": 0.8632, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.5076923076923077, |
| "grad_norm": 0.2386821061372757, |
| "learning_rate": 0.00016851441241685144, |
| "loss": 0.9646, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.5142857142857142, |
| "grad_norm": 0.3139204978942871, |
| "learning_rate": 0.0001680709534368071, |
| "loss": 0.9282, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.5208791208791209, |
| "grad_norm": 0.37618762254714966, |
| "learning_rate": 0.00016762749445676275, |
| "loss": 1.1463, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.5274725274725275, |
| "grad_norm": 0.3567148447036743, |
| "learning_rate": 0.00016718403547671842, |
| "loss": 1.2093, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.5340659340659341, |
| "grad_norm": 0.3268345594406128, |
| "learning_rate": 0.00016674057649667407, |
| "loss": 0.8578, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.5406593406593406, |
| "grad_norm": 0.3362461030483246, |
| "learning_rate": 0.00016629711751662974, |
| "loss": 1.209, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.5472527472527473, |
| "grad_norm": 0.33477821946144104, |
| "learning_rate": 0.00016585365853658536, |
| "loss": 0.9915, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.5538461538461539, |
| "grad_norm": 0.38696935772895813, |
| "learning_rate": 0.00016541019955654103, |
| "loss": 0.8887, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.5604395604395604, |
| "grad_norm": 0.3261304199695587, |
| "learning_rate": 0.00016496674057649667, |
| "loss": 0.9255, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.567032967032967, |
| "grad_norm": 0.3052448034286499, |
| "learning_rate": 0.00016452328159645234, |
| "loss": 0.9299, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.5736263736263736, |
| "grad_norm": 0.3229495584964752, |
| "learning_rate": 0.000164079822616408, |
| "loss": 0.874, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.5802197802197803, |
| "grad_norm": 0.34713301062583923, |
| "learning_rate": 0.00016363636363636366, |
| "loss": 0.8682, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.5868131868131868, |
| "grad_norm": 0.32546767592430115, |
| "learning_rate": 0.0001631929046563193, |
| "loss": 0.9824, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.5934065934065934, |
| "grad_norm": 0.2888805866241455, |
| "learning_rate": 0.00016274944567627495, |
| "loss": 0.8598, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.6, |
| "grad_norm": 0.32607126235961914, |
| "learning_rate": 0.0001623059866962306, |
| "loss": 0.891, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.6065934065934065, |
| "grad_norm": 0.31503042578697205, |
| "learning_rate": 0.00016186252771618626, |
| "loss": 0.9513, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.6131868131868132, |
| "grad_norm": 0.398336797952652, |
| "learning_rate": 0.0001614190687361419, |
| "loss": 0.8821, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.6197802197802198, |
| "grad_norm": 0.32223382592201233, |
| "learning_rate": 0.00016097560975609758, |
| "loss": 0.9786, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.6263736263736264, |
| "grad_norm": 0.33958667516708374, |
| "learning_rate": 0.00016053215077605322, |
| "loss": 0.8952, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.6329670329670329, |
| "grad_norm": 0.2868139147758484, |
| "learning_rate": 0.00016008869179600887, |
| "loss": 0.8921, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.6395604395604395, |
| "grad_norm": 0.34224361181259155, |
| "learning_rate": 0.00015964523281596454, |
| "loss": 0.9362, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.6461538461538462, |
| "grad_norm": 0.30754712224006653, |
| "learning_rate": 0.00015920177383592018, |
| "loss": 0.9801, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.6527472527472528, |
| "grad_norm": 0.2931382656097412, |
| "learning_rate": 0.00015875831485587586, |
| "loss": 1.1286, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.6593406593406593, |
| "grad_norm": 0.30773454904556274, |
| "learning_rate": 0.0001583148558758315, |
| "loss": 1.0438, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.6659340659340659, |
| "grad_norm": 0.3150854706764221, |
| "learning_rate": 0.00015787139689578714, |
| "loss": 1.1029, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.6725274725274726, |
| "grad_norm": 0.34101125597953796, |
| "learning_rate": 0.0001574279379157428, |
| "loss": 1.3197, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.6791208791208792, |
| "grad_norm": 0.301902174949646, |
| "learning_rate": 0.00015698447893569846, |
| "loss": 0.9134, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.6857142857142857, |
| "grad_norm": 0.4397645592689514, |
| "learning_rate": 0.0001565410199556541, |
| "loss": 1.0168, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.6923076923076923, |
| "grad_norm": 0.30188363790512085, |
| "learning_rate": 0.00015609756097560978, |
| "loss": 0.9051, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.6989010989010989, |
| "grad_norm": 0.28673937916755676, |
| "learning_rate": 0.00015565410199556542, |
| "loss": 1.0421, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.7054945054945055, |
| "grad_norm": 0.34226083755493164, |
| "learning_rate": 0.00015521064301552106, |
| "loss": 0.8999, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.7120879120879121, |
| "grad_norm": 0.21212828159332275, |
| "learning_rate": 0.0001547671840354767, |
| "loss": 0.6621, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.7186813186813187, |
| "grad_norm": 0.33283287286758423, |
| "learning_rate": 0.00015432372505543238, |
| "loss": 1.0817, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.7252747252747253, |
| "grad_norm": 0.3267793357372284, |
| "learning_rate": 0.00015388026607538802, |
| "loss": 1.0826, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.7318681318681318, |
| "grad_norm": 0.33544689416885376, |
| "learning_rate": 0.0001534368070953437, |
| "loss": 1.043, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.7384615384615385, |
| "grad_norm": 0.2979942858219147, |
| "learning_rate": 0.00015299334811529934, |
| "loss": 0.9555, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.7450549450549451, |
| "grad_norm": 0.29721397161483765, |
| "learning_rate": 0.00015254988913525498, |
| "loss": 0.7911, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.7516483516483516, |
| "grad_norm": 0.23252011835575104, |
| "learning_rate": 0.00015210643015521066, |
| "loss": 0.7186, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.7582417582417582, |
| "grad_norm": 0.3264557719230652, |
| "learning_rate": 0.0001516629711751663, |
| "loss": 0.9136, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.7648351648351648, |
| "grad_norm": 0.3115456700325012, |
| "learning_rate": 0.00015121951219512197, |
| "loss": 1.0437, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.7714285714285715, |
| "grad_norm": 0.3569914996623993, |
| "learning_rate": 0.00015077605321507762, |
| "loss": 0.9559, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.778021978021978, |
| "grad_norm": 0.3666040301322937, |
| "learning_rate": 0.0001503325942350333, |
| "loss": 1.0863, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.7846153846153846, |
| "grad_norm": 0.44128793478012085, |
| "learning_rate": 0.0001498891352549889, |
| "loss": 1.0112, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.7912087912087912, |
| "grad_norm": 0.24702684581279755, |
| "learning_rate": 0.00014944567627494458, |
| "loss": 0.7882, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.7978021978021979, |
| "grad_norm": 0.32932835817337036, |
| "learning_rate": 0.00014900221729490022, |
| "loss": 0.8732, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.8043956043956044, |
| "grad_norm": 0.2903375029563904, |
| "learning_rate": 0.0001485587583148559, |
| "loss": 0.9835, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.810989010989011, |
| "grad_norm": 0.3277665376663208, |
| "learning_rate": 0.00014811529933481154, |
| "loss": 0.9524, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.8175824175824176, |
| "grad_norm": 0.3548404276371002, |
| "learning_rate": 0.0001476718403547672, |
| "loss": 1.0401, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.8241758241758241, |
| "grad_norm": 0.3156849145889282, |
| "learning_rate": 0.00014722838137472282, |
| "loss": 1.0754, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.8307692307692308, |
| "grad_norm": 0.37477758526802063, |
| "learning_rate": 0.0001467849223946785, |
| "loss": 0.9014, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.8373626373626374, |
| "grad_norm": 0.3087361752986908, |
| "learning_rate": 0.00014634146341463414, |
| "loss": 1.08, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.843956043956044, |
| "grad_norm": 0.2732657492160797, |
| "learning_rate": 0.0001458980044345898, |
| "loss": 0.8144, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.8505494505494505, |
| "grad_norm": 0.29096436500549316, |
| "learning_rate": 0.00014545454545454546, |
| "loss": 0.8827, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.8571428571428571, |
| "grad_norm": 0.29380616545677185, |
| "learning_rate": 0.00014501108647450113, |
| "loss": 1.0531, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.8637362637362638, |
| "grad_norm": 0.31562578678131104, |
| "learning_rate": 0.00014456762749445675, |
| "loss": 0.8875, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.8703296703296703, |
| "grad_norm": 0.35802748799324036, |
| "learning_rate": 0.00014412416851441242, |
| "loss": 1.0098, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.8769230769230769, |
| "grad_norm": 0.32378309965133667, |
| "learning_rate": 0.00014368070953436806, |
| "loss": 0.8362, |
| "step": 133 |
| }, |
| { |
| "epoch": 0.8835164835164835, |
| "grad_norm": 0.311889111995697, |
| "learning_rate": 0.00014323725055432373, |
| "loss": 0.9096, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.8901098901098901, |
| "grad_norm": 0.3975204825401306, |
| "learning_rate": 0.00014279379157427938, |
| "loss": 0.826, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.8967032967032967, |
| "grad_norm": 0.2858659625053406, |
| "learning_rate": 0.00014235033259423505, |
| "loss": 0.8492, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.9032967032967033, |
| "grad_norm": 0.4061797261238098, |
| "learning_rate": 0.0001419068736141907, |
| "loss": 0.9427, |
| "step": 137 |
| }, |
| { |
| "epoch": 0.9098901098901099, |
| "grad_norm": 0.3105581998825073, |
| "learning_rate": 0.00014146341463414634, |
| "loss": 0.9379, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.9164835164835164, |
| "grad_norm": 0.3545703589916229, |
| "learning_rate": 0.000141019955654102, |
| "loss": 0.9037, |
| "step": 139 |
| }, |
| { |
| "epoch": 0.9230769230769231, |
| "grad_norm": 0.290799617767334, |
| "learning_rate": 0.00014057649667405765, |
| "loss": 0.7515, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.9296703296703297, |
| "grad_norm": 0.3229195177555084, |
| "learning_rate": 0.00014013303769401332, |
| "loss": 0.9423, |
| "step": 141 |
| }, |
| { |
| "epoch": 0.9362637362637363, |
| "grad_norm": 0.38070327043533325, |
| "learning_rate": 0.00013968957871396897, |
| "loss": 1.0264, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.9428571428571428, |
| "grad_norm": 0.3201819062232971, |
| "learning_rate": 0.00013924611973392464, |
| "loss": 0.8616, |
| "step": 143 |
| }, |
| { |
| "epoch": 0.9494505494505494, |
| "grad_norm": 0.28324103355407715, |
| "learning_rate": 0.00013880266075388026, |
| "loss": 0.9933, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.9560439560439561, |
| "grad_norm": 0.3484077453613281, |
| "learning_rate": 0.00013835920177383593, |
| "loss": 0.9151, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.9626373626373627, |
| "grad_norm": 0.3093436062335968, |
| "learning_rate": 0.00013791574279379157, |
| "loss": 0.8514, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.9692307692307692, |
| "grad_norm": 0.2318611741065979, |
| "learning_rate": 0.00013747228381374724, |
| "loss": 0.7289, |
| "step": 147 |
| }, |
| { |
| "epoch": 0.9758241758241758, |
| "grad_norm": 0.4191138744354248, |
| "learning_rate": 0.0001370288248337029, |
| "loss": 1.0423, |
| "step": 148 |
| }, |
| { |
| "epoch": 0.9824175824175824, |
| "grad_norm": 0.29127630591392517, |
| "learning_rate": 0.00013658536585365856, |
| "loss": 1.0706, |
| "step": 149 |
| }, |
| { |
| "epoch": 0.989010989010989, |
| "grad_norm": 0.3293207585811615, |
| "learning_rate": 0.00013614190687361418, |
| "loss": 0.846, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.9956043956043956, |
| "grad_norm": 0.32789573073387146, |
| "learning_rate": 0.00013569844789356985, |
| "loss": 0.8972, |
| "step": 151 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 0.35556045174598694, |
| "learning_rate": 0.0001352549889135255, |
| "loss": 1.0373, |
| "step": 152 |
| }, |
| { |
| "epoch": 1.0065934065934066, |
| "grad_norm": 0.2797600030899048, |
| "learning_rate": 0.00013481152993348116, |
| "loss": 0.8272, |
| "step": 153 |
| }, |
| { |
| "epoch": 1.0131868131868131, |
| "grad_norm": 0.27939650416374207, |
| "learning_rate": 0.0001343680709534368, |
| "loss": 0.7378, |
| "step": 154 |
| }, |
| { |
| "epoch": 1.0197802197802197, |
| "grad_norm": 0.3140670955181122, |
| "learning_rate": 0.00013392461197339248, |
| "loss": 0.873, |
| "step": 155 |
| }, |
| { |
| "epoch": 1.0263736263736263, |
| "grad_norm": 0.3043338358402252, |
| "learning_rate": 0.00013348115299334812, |
| "loss": 0.728, |
| "step": 156 |
| }, |
| { |
| "epoch": 1.032967032967033, |
| "grad_norm": 0.23797346651554108, |
| "learning_rate": 0.00013303769401330377, |
| "loss": 0.8907, |
| "step": 157 |
| }, |
| { |
| "epoch": 1.0395604395604396, |
| "grad_norm": 0.3470970094203949, |
| "learning_rate": 0.00013259423503325944, |
| "loss": 0.9333, |
| "step": 158 |
| }, |
| { |
| "epoch": 1.0461538461538462, |
| "grad_norm": 0.36303016543388367, |
| "learning_rate": 0.00013215077605321509, |
| "loss": 0.7435, |
| "step": 159 |
| }, |
| { |
| "epoch": 1.0527472527472528, |
| "grad_norm": 0.3077246844768524, |
| "learning_rate": 0.00013170731707317076, |
| "loss": 0.7731, |
| "step": 160 |
| }, |
| { |
| "epoch": 1.0593406593406594, |
| "grad_norm": 0.31969889998435974, |
| "learning_rate": 0.0001312638580931264, |
| "loss": 0.8777, |
| "step": 161 |
| }, |
| { |
| "epoch": 1.065934065934066, |
| "grad_norm": 0.3754861056804657, |
| "learning_rate": 0.00013082039911308205, |
| "loss": 1.098, |
| "step": 162 |
| }, |
| { |
| "epoch": 1.0725274725274725, |
| "grad_norm": 0.29991087317466736, |
| "learning_rate": 0.0001303769401330377, |
| "loss": 0.7567, |
| "step": 163 |
| }, |
| { |
| "epoch": 1.079120879120879, |
| "grad_norm": 0.3924933075904846, |
| "learning_rate": 0.00012993348115299336, |
| "loss": 0.8253, |
| "step": 164 |
| }, |
| { |
| "epoch": 1.0857142857142856, |
| "grad_norm": 0.3110440969467163, |
| "learning_rate": 0.000129490022172949, |
| "loss": 0.9294, |
| "step": 165 |
| }, |
| { |
| "epoch": 1.0923076923076924, |
| "grad_norm": 0.477671355009079, |
| "learning_rate": 0.00012904656319290468, |
| "loss": 0.9962, |
| "step": 166 |
| }, |
| { |
| "epoch": 1.098901098901099, |
| "grad_norm": 0.36420226097106934, |
| "learning_rate": 0.00012860310421286032, |
| "loss": 0.8716, |
| "step": 167 |
| }, |
| { |
| "epoch": 1.1054945054945056, |
| "grad_norm": 0.3950567841529846, |
| "learning_rate": 0.00012815964523281597, |
| "loss": 0.6489, |
| "step": 168 |
| }, |
| { |
| "epoch": 1.1120879120879121, |
| "grad_norm": 0.49215519428253174, |
| "learning_rate": 0.0001277161862527716, |
| "loss": 0.8981, |
| "step": 169 |
| }, |
| { |
| "epoch": 1.1186813186813187, |
| "grad_norm": 0.3717999756336212, |
| "learning_rate": 0.00012727272727272728, |
| "loss": 1.0126, |
| "step": 170 |
| }, |
| { |
| "epoch": 1.1252747252747253, |
| "grad_norm": 0.32597070932388306, |
| "learning_rate": 0.00012682926829268293, |
| "loss": 0.8131, |
| "step": 171 |
| }, |
| { |
| "epoch": 1.1318681318681318, |
| "grad_norm": 0.3318406939506531, |
| "learning_rate": 0.0001263858093126386, |
| "loss": 0.9198, |
| "step": 172 |
| }, |
| { |
| "epoch": 1.1384615384615384, |
| "grad_norm": 0.4110514223575592, |
| "learning_rate": 0.00012594235033259424, |
| "loss": 0.9242, |
| "step": 173 |
| }, |
| { |
| "epoch": 1.145054945054945, |
| "grad_norm": 0.3442305326461792, |
| "learning_rate": 0.00012549889135254989, |
| "loss": 0.9842, |
| "step": 174 |
| }, |
| { |
| "epoch": 1.1516483516483516, |
| "grad_norm": 0.40250134468078613, |
| "learning_rate": 0.00012505543237250553, |
| "loss": 0.8919, |
| "step": 175 |
| }, |
| { |
| "epoch": 1.1582417582417581, |
| "grad_norm": 0.3540174067020416, |
| "learning_rate": 0.0001246119733924612, |
| "loss": 0.8814, |
| "step": 176 |
| }, |
| { |
| "epoch": 1.164835164835165, |
| "grad_norm": 0.4426930844783783, |
| "learning_rate": 0.00012416851441241685, |
| "loss": 0.6934, |
| "step": 177 |
| }, |
| { |
| "epoch": 1.1714285714285715, |
| "grad_norm": 0.3581725060939789, |
| "learning_rate": 0.00012372505543237252, |
| "loss": 0.688, |
| "step": 178 |
| }, |
| { |
| "epoch": 1.178021978021978, |
| "grad_norm": 0.36629343032836914, |
| "learning_rate": 0.00012328159645232816, |
| "loss": 1.0895, |
| "step": 179 |
| }, |
| { |
| "epoch": 1.1846153846153846, |
| "grad_norm": 0.31905603408813477, |
| "learning_rate": 0.0001228381374722838, |
| "loss": 0.7644, |
| "step": 180 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 456, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 15, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 2.97165952487424e+16, |
| "train_batch_size": 22, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
