| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9878213802435725, | |
| "eval_steps": 500, | |
| "global_step": 276, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010825439783491205, | |
| "grad_norm": 5.8087664609607845, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 0.8566, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02165087956698241, | |
| "grad_norm": 5.8131844736866745, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 0.8603, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03247631935047361, | |
| "grad_norm": 5.440621255632965, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 0.8525, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.04330175913396482, | |
| "grad_norm": 2.3054665992357184, | |
| "learning_rate": 1.1428571428571429e-05, | |
| "loss": 0.7759, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.05412719891745602, | |
| "grad_norm": 3.924860005723019, | |
| "learning_rate": 1.4285714285714287e-05, | |
| "loss": 0.7543, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06495263870094722, | |
| "grad_norm": 4.208977456644038, | |
| "learning_rate": 1.7142857142857142e-05, | |
| "loss": 0.7509, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07577807848443843, | |
| "grad_norm": 4.359799562993377, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7343, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.08660351826792964, | |
| "grad_norm": 2.9096972595545263, | |
| "learning_rate": 2.2857142857142858e-05, | |
| "loss": 0.729, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.09742895805142084, | |
| "grad_norm": 2.753429344328329, | |
| "learning_rate": 2.5714285714285718e-05, | |
| "loss": 0.672, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.10825439783491204, | |
| "grad_norm": 2.1289654887761356, | |
| "learning_rate": 2.8571428571428574e-05, | |
| "loss": 0.6659, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.11907983761840325, | |
| "grad_norm": 1.5620330251776127, | |
| "learning_rate": 3.142857142857143e-05, | |
| "loss": 0.6379, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.12990527740189445, | |
| "grad_norm": 1.3572209762178864, | |
| "learning_rate": 3.4285714285714284e-05, | |
| "loss": 0.6303, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.14073071718538566, | |
| "grad_norm": 1.5509649753365637, | |
| "learning_rate": 3.714285714285715e-05, | |
| "loss": 0.614, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.15155615696887687, | |
| "grad_norm": 1.2221200154613043, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6245, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.16238159675236807, | |
| "grad_norm": 1.364235921605161, | |
| "learning_rate": 4.2857142857142856e-05, | |
| "loss": 0.61, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.17320703653585928, | |
| "grad_norm": 1.189733323627339, | |
| "learning_rate": 4.5714285714285716e-05, | |
| "loss": 0.5977, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.18403247631935046, | |
| "grad_norm": 1.7575676747700886, | |
| "learning_rate": 4.857142857142857e-05, | |
| "loss": 0.5936, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.19485791610284167, | |
| "grad_norm": 1.0997504244414555, | |
| "learning_rate": 5.1428571428571436e-05, | |
| "loss": 0.5755, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.20568335588633288, | |
| "grad_norm": 1.6966476141679678, | |
| "learning_rate": 5.4285714285714295e-05, | |
| "loss": 0.5698, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2165087956698241, | |
| "grad_norm": 1.4996825457561533, | |
| "learning_rate": 5.714285714285715e-05, | |
| "loss": 0.5742, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2273342354533153, | |
| "grad_norm": 1.7195794249062655, | |
| "learning_rate": 6.000000000000001e-05, | |
| "loss": 0.5694, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2381596752368065, | |
| "grad_norm": 1.3451466364919609, | |
| "learning_rate": 6.285714285714286e-05, | |
| "loss": 0.5683, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2489851150202977, | |
| "grad_norm": 1.894775974408309, | |
| "learning_rate": 6.571428571428571e-05, | |
| "loss": 0.5623, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2598105548037889, | |
| "grad_norm": 1.3966190302733898, | |
| "learning_rate": 6.857142857142857e-05, | |
| "loss": 0.5673, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2706359945872801, | |
| "grad_norm": 1.9745789162966707, | |
| "learning_rate": 7.142857142857143e-05, | |
| "loss": 0.5599, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2814614343707713, | |
| "grad_norm": 1.4927835449790416, | |
| "learning_rate": 7.42857142857143e-05, | |
| "loss": 0.5393, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2922868741542625, | |
| "grad_norm": 1.9041578436403848, | |
| "learning_rate": 7.714285714285715e-05, | |
| "loss": 0.545, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.30311231393775373, | |
| "grad_norm": 1.4926590028287288, | |
| "learning_rate": 8e-05, | |
| "loss": 0.5515, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.31393775372124494, | |
| "grad_norm": 1.6334956142809471, | |
| "learning_rate": 7.999679062421315e-05, | |
| "loss": 0.5486, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.32476319350473615, | |
| "grad_norm": 3.2641403044755752, | |
| "learning_rate": 7.998716301185722e-05, | |
| "loss": 0.5478, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.33558863328822736, | |
| "grad_norm": 3929.33892996564, | |
| "learning_rate": 7.997111870786354e-05, | |
| "loss": 1.6128, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.34641407307171856, | |
| "grad_norm": 10.911696810971137, | |
| "learning_rate": 7.994866028684212e-05, | |
| "loss": 0.5695, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3572395128552097, | |
| "grad_norm": 2.0445482049515733, | |
| "learning_rate": 7.991979135266861e-05, | |
| "loss": 0.5748, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3680649526387009, | |
| "grad_norm": 1.340044633834453, | |
| "learning_rate": 7.98845165379059e-05, | |
| "loss": 0.5537, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.37889039242219213, | |
| "grad_norm": 2.1153484180739963, | |
| "learning_rate": 7.984284150306085e-05, | |
| "loss": 0.543, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.38971583220568334, | |
| "grad_norm": 1.2802706235141443, | |
| "learning_rate": 7.97947729356758e-05, | |
| "loss": 0.5335, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.40054127198917455, | |
| "grad_norm": 2.1978971171709003, | |
| "learning_rate": 7.974031854925562e-05, | |
| "loss": 0.5411, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.41136671177266576, | |
| "grad_norm": 1.5255430027571375, | |
| "learning_rate": 7.967948708202972e-05, | |
| "loss": 0.5335, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.42219215155615697, | |
| "grad_norm": 1.8471678385687695, | |
| "learning_rate": 7.961228829555003e-05, | |
| "loss": 0.5393, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.4330175913396482, | |
| "grad_norm": 1.3402125559936133, | |
| "learning_rate": 7.953873297312447e-05, | |
| "loss": 0.5243, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4438430311231394, | |
| "grad_norm": 1.8106418023497959, | |
| "learning_rate": 7.945883291808655e-05, | |
| "loss": 0.5261, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.4546684709066306, | |
| "grad_norm": 1.4152890082976677, | |
| "learning_rate": 7.937260095190137e-05, | |
| "loss": 0.5202, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4654939106901218, | |
| "grad_norm": 1.3073632081364428, | |
| "learning_rate": 7.928005091210817e-05, | |
| "loss": 0.5163, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.476319350473613, | |
| "grad_norm": 1.2185742379112405, | |
| "learning_rate": 7.918119765009979e-05, | |
| "loss": 0.5145, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.4871447902571042, | |
| "grad_norm": 1.1307175104637475, | |
| "learning_rate": 7.907605702873948e-05, | |
| "loss": 0.5173, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4979702300405954, | |
| "grad_norm": 1.5615281762012732, | |
| "learning_rate": 7.896464591981549e-05, | |
| "loss": 0.5146, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.5087956698240866, | |
| "grad_norm": 1.661842657431127, | |
| "learning_rate": 7.884698220133357e-05, | |
| "loss": 0.5144, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.5196211096075778, | |
| "grad_norm": 0.9645999991081547, | |
| "learning_rate": 7.872308475464818e-05, | |
| "loss": 0.5093, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.530446549391069, | |
| "grad_norm": 1.2076664817970877, | |
| "learning_rate": 7.859297346143258e-05, | |
| "loss": 0.5082, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.5412719891745602, | |
| "grad_norm": 1.5720938935578466, | |
| "learning_rate": 7.84566692004885e-05, | |
| "loss": 0.5075, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5520974289580515, | |
| "grad_norm": 1.7727617552441766, | |
| "learning_rate": 7.831419384439565e-05, | |
| "loss": 0.5061, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.5629228687415426, | |
| "grad_norm": 0.7992807977962048, | |
| "learning_rate": 7.816557025600196e-05, | |
| "loss": 0.4959, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5737483085250338, | |
| "grad_norm": 2.1466918628904588, | |
| "learning_rate": 7.80108222847547e-05, | |
| "loss": 0.4999, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.584573748308525, | |
| "grad_norm": 1.2411877400240239, | |
| "learning_rate": 7.784997476287349e-05, | |
| "loss": 0.5127, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5953991880920162, | |
| "grad_norm": 1.8015582678737536, | |
| "learning_rate": 7.76830535013654e-05, | |
| "loss": 0.4999, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.6062246278755075, | |
| "grad_norm": 1.3586321087574318, | |
| "learning_rate": 7.751008528588322e-05, | |
| "loss": 0.5055, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.6170500676589986, | |
| "grad_norm": 1.7146896357463135, | |
| "learning_rate": 7.733109787242708e-05, | |
| "loss": 0.5011, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.6278755074424899, | |
| "grad_norm": 1.14091441377822, | |
| "learning_rate": 7.71461199828905e-05, | |
| "loss": 0.4994, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.638700947225981, | |
| "grad_norm": 1.2740719297995, | |
| "learning_rate": 7.695518130045147e-05, | |
| "loss": 0.4865, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6495263870094723, | |
| "grad_norm": 1.2663598979505173, | |
| "learning_rate": 7.675831246480923e-05, | |
| "loss": 0.4934, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6603518267929634, | |
| "grad_norm": 1.2797944310392075, | |
| "learning_rate": 7.655554506726747e-05, | |
| "loss": 0.5039, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.6711772665764547, | |
| "grad_norm": 1.3838168196069196, | |
| "learning_rate": 7.6346911645665e-05, | |
| "loss": 0.4933, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6820027063599459, | |
| "grad_norm": 1.7189329763004875, | |
| "learning_rate": 7.61324456791544e-05, | |
| "loss": 0.5039, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6928281461434371, | |
| "grad_norm": 0.8056952169520525, | |
| "learning_rate": 7.591218158282968e-05, | |
| "loss": 0.496, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.7036535859269283, | |
| "grad_norm": 1.4161621521145464, | |
| "learning_rate": 7.568615470220369e-05, | |
| "loss": 0.497, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.7144790257104194, | |
| "grad_norm": 1.1742977969559418, | |
| "learning_rate": 7.545440130753634e-05, | |
| "loss": 0.5032, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.7253044654939107, | |
| "grad_norm": 1.554226462240922, | |
| "learning_rate": 7.52169585880143e-05, | |
| "loss": 0.4975, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.7361299052774019, | |
| "grad_norm": 1.0676967952547851, | |
| "learning_rate": 7.497386464578329e-05, | |
| "loss": 0.5055, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.7469553450608931, | |
| "grad_norm": 1.6373366615764569, | |
| "learning_rate": 7.472515848983394e-05, | |
| "loss": 0.5024, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7577807848443843, | |
| "grad_norm": 1.0798983229476786, | |
| "learning_rate": 7.447088002974199e-05, | |
| "loss": 0.48, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7686062246278755, | |
| "grad_norm": 1.71937852314963, | |
| "learning_rate": 7.421107006926408e-05, | |
| "loss": 0.4854, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.7794316644113667, | |
| "grad_norm": 1.3186745236022845, | |
| "learning_rate": 7.394577029979004e-05, | |
| "loss": 0.4977, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.790257104194858, | |
| "grad_norm": 1.7501120115784738, | |
| "learning_rate": 7.367502329365268e-05, | |
| "loss": 0.5004, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.8010825439783491, | |
| "grad_norm": 1.3486345921440117, | |
| "learning_rate": 7.33988724972963e-05, | |
| "loss": 0.4968, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.8119079837618404, | |
| "grad_norm": 1.784526687443049, | |
| "learning_rate": 7.311736222430487e-05, | |
| "loss": 0.491, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.8227334235453315, | |
| "grad_norm": 1.2096257660245169, | |
| "learning_rate": 7.283053764829106e-05, | |
| "loss": 0.484, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.8335588633288228, | |
| "grad_norm": 1.5112482452095675, | |
| "learning_rate": 7.253844479564737e-05, | |
| "loss": 0.4817, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.8443843031123139, | |
| "grad_norm": 1.243473583940459, | |
| "learning_rate": 7.224113053816021e-05, | |
| "loss": 0.4852, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.8552097428958051, | |
| "grad_norm": 1.5663481840321567, | |
| "learning_rate": 7.193864258548855e-05, | |
| "loss": 0.489, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.8660351826792964, | |
| "grad_norm": 1.0782094949200172, | |
| "learning_rate": 7.163102947750794e-05, | |
| "loss": 0.4838, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8768606224627875, | |
| "grad_norm": 1.5986154352316757, | |
| "learning_rate": 7.131834057652142e-05, | |
| "loss": 0.4925, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.8876860622462788, | |
| "grad_norm": 1.371379633845542, | |
| "learning_rate": 7.100062605933835e-05, | |
| "loss": 0.4814, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.8985115020297699, | |
| "grad_norm": 1.302557126672863, | |
| "learning_rate": 7.067793690922268e-05, | |
| "loss": 0.4835, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.9093369418132612, | |
| "grad_norm": 1.090603750825841, | |
| "learning_rate": 7.035032490771165e-05, | |
| "loss": 0.4826, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.9201623815967523, | |
| "grad_norm": 1.0864878831546754, | |
| "learning_rate": 7.001784262630652e-05, | |
| "loss": 0.4794, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.9309878213802436, | |
| "grad_norm": 0.8709298151810692, | |
| "learning_rate": 6.968054341803644e-05, | |
| "loss": 0.4828, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.9418132611637348, | |
| "grad_norm": 1.6941280905464255, | |
| "learning_rate": 6.933848140889705e-05, | |
| "loss": 0.4807, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.952638700947226, | |
| "grad_norm": 0.930295842416618, | |
| "learning_rate": 6.89917114891648e-05, | |
| "loss": 0.467, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.9634641407307172, | |
| "grad_norm": 1.3410428094816311, | |
| "learning_rate": 6.864028930458892e-05, | |
| "loss": 0.4813, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.9742895805142084, | |
| "grad_norm": 1.4185244104880752, | |
| "learning_rate": 6.828427124746191e-05, | |
| "loss": 0.4823, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.9851150202976996, | |
| "grad_norm": 0.8396448430164011, | |
| "learning_rate": 6.792371444757037e-05, | |
| "loss": 0.4787, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.9959404600811907, | |
| "grad_norm": 1.1676153586045548, | |
| "learning_rate": 6.755867676302747e-05, | |
| "loss": 0.4825, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.006765899864682, | |
| "grad_norm": 2.146342320876203, | |
| "learning_rate": 6.718921677098853e-05, | |
| "loss": 0.7519, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.0175913396481733, | |
| "grad_norm": 1.2204156668696224, | |
| "learning_rate": 6.681539375825115e-05, | |
| "loss": 0.4591, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.0284167794316643, | |
| "grad_norm": 0.9418529660384006, | |
| "learning_rate": 6.643726771174164e-05, | |
| "loss": 0.4564, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.0392422192151556, | |
| "grad_norm": 1.1839114213551731, | |
| "learning_rate": 6.60548993088889e-05, | |
| "loss": 0.4574, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0500676589986468, | |
| "grad_norm": 1.4575234413047131, | |
| "learning_rate": 6.56683499078876e-05, | |
| "loss": 0.4503, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.060893098782138, | |
| "grad_norm": 0.8758736510165871, | |
| "learning_rate": 6.527768153785216e-05, | |
| "loss": 0.4557, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.0717185385656292, | |
| "grad_norm": 1.1203958511631036, | |
| "learning_rate": 6.488295688886295e-05, | |
| "loss": 0.4538, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.0825439783491204, | |
| "grad_norm": 1.4383565732547114, | |
| "learning_rate": 6.448423930190653e-05, | |
| "loss": 0.4517, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0933694181326117, | |
| "grad_norm": 0.7642643301268309, | |
| "learning_rate": 6.408159275871132e-05, | |
| "loss": 0.45, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.104194857916103, | |
| "grad_norm": 0.9189424705853121, | |
| "learning_rate": 6.36750818714807e-05, | |
| "loss": 0.4526, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.115020297699594, | |
| "grad_norm": 0.9382870992965567, | |
| "learning_rate": 6.326477187252455e-05, | |
| "loss": 0.455, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.1258457374830853, | |
| "grad_norm": 1.2949430415873158, | |
| "learning_rate": 6.28507286037917e-05, | |
| "loss": 0.4495, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.1366711772665765, | |
| "grad_norm": 0.7351344885973718, | |
| "learning_rate": 6.243301850630419e-05, | |
| "loss": 0.4516, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.1474966170500678, | |
| "grad_norm": 0.980693126870663, | |
| "learning_rate": 6.201170860949565e-05, | |
| "loss": 0.4544, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.1583220568335588, | |
| "grad_norm": 1.516092843633026, | |
| "learning_rate": 6.15868665204552e-05, | |
| "loss": 0.4518, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.16914749661705, | |
| "grad_norm": 0.5987661096912231, | |
| "learning_rate": 6.11585604130785e-05, | |
| "loss": 0.4437, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.1799729364005414, | |
| "grad_norm": 1.2043721169403026, | |
| "learning_rate": 6.072685901712808e-05, | |
| "loss": 0.4528, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.1907983761840324, | |
| "grad_norm": 1.0400040833177389, | |
| "learning_rate": 6.02918316072043e-05, | |
| "loss": 0.4565, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.2016238159675237, | |
| "grad_norm": 0.9070577060618072, | |
| "learning_rate": 5.9853547991628967e-05, | |
| "loss": 0.4546, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.212449255751015, | |
| "grad_norm": 0.9325395739201259, | |
| "learning_rate": 5.941207850124325e-05, | |
| "loss": 0.4489, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.2232746955345062, | |
| "grad_norm": 0.921059943088959, | |
| "learning_rate": 5.896749397812181e-05, | |
| "loss": 0.4448, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.2341001353179972, | |
| "grad_norm": 0.8172704617982391, | |
| "learning_rate": 5.8519865764204834e-05, | |
| "loss": 0.4504, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.2449255751014885, | |
| "grad_norm": 0.8839107108189724, | |
| "learning_rate": 5.8069265689849884e-05, | |
| "loss": 0.4568, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.2557510148849798, | |
| "grad_norm": 1.2575966730831944, | |
| "learning_rate": 5.761576606230538e-05, | |
| "loss": 0.4489, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.266576454668471, | |
| "grad_norm": 0.5333063557898303, | |
| "learning_rate": 5.7159439654107506e-05, | |
| "loss": 0.4441, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.277401894451962, | |
| "grad_norm": 1.0494605977905378, | |
| "learning_rate": 5.6700359691402533e-05, | |
| "loss": 0.4449, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.2882273342354533, | |
| "grad_norm": 0.9725379237388848, | |
| "learning_rate": 5.6238599842196285e-05, | |
| "loss": 0.4497, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.2990527740189446, | |
| "grad_norm": 0.7230962046892564, | |
| "learning_rate": 5.5774234204532746e-05, | |
| "loss": 0.4417, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.3098782138024356, | |
| "grad_norm": 0.7902778339531633, | |
| "learning_rate": 5.5307337294603595e-05, | |
| "loss": 0.4493, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.320703653585927, | |
| "grad_norm": 0.7504923802044895, | |
| "learning_rate": 5.483798403479072e-05, | |
| "loss": 0.4438, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.3315290933694182, | |
| "grad_norm": 0.5320675700936504, | |
| "learning_rate": 5.436624974164349e-05, | |
| "loss": 0.4456, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.3423545331529092, | |
| "grad_norm": 0.5518673717581679, | |
| "learning_rate": 5.389221011379281e-05, | |
| "loss": 0.4439, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.3531799729364005, | |
| "grad_norm": 0.6874367401876698, | |
| "learning_rate": 5.3415941219803895e-05, | |
| "loss": 0.4439, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.3640054127198917, | |
| "grad_norm": 0.617879380906548, | |
| "learning_rate": 5.2937519485969525e-05, | |
| "loss": 0.4428, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.374830852503383, | |
| "grad_norm": 0.6012050158242978, | |
| "learning_rate": 5.245702168404616e-05, | |
| "loss": 0.4362, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.3856562922868743, | |
| "grad_norm": 0.4340883446785468, | |
| "learning_rate": 5.1974524918934336e-05, | |
| "loss": 0.4383, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.3964817320703653, | |
| "grad_norm": 0.5607932307214518, | |
| "learning_rate": 5.14901066163058e-05, | |
| "loss": 0.4376, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.4073071718538566, | |
| "grad_norm": 0.6865271257452217, | |
| "learning_rate": 5.1003844510179126e-05, | |
| "loss": 0.442, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.4181326116373478, | |
| "grad_norm": 0.3997948939760999, | |
| "learning_rate": 5.0515816630445795e-05, | |
| "loss": 0.4439, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.4289580514208389, | |
| "grad_norm": 0.4937361614442335, | |
| "learning_rate": 5.002610129034883e-05, | |
| "loss": 0.4448, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.4397834912043301, | |
| "grad_norm": 0.5370860413198207, | |
| "learning_rate": 4.953477707391597e-05, | |
| "loss": 0.4389, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.4506089309878214, | |
| "grad_norm": 0.4729853469838054, | |
| "learning_rate": 4.90419228233494e-05, | |
| "loss": 0.4403, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.4614343707713127, | |
| "grad_norm": 0.36807074398732437, | |
| "learning_rate": 4.854761762637403e-05, | |
| "loss": 0.4405, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.472259810554804, | |
| "grad_norm": 0.4057735308453249, | |
| "learning_rate": 4.805194080354641e-05, | |
| "loss": 0.4394, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.483085250338295, | |
| "grad_norm": 0.3910899543447501, | |
| "learning_rate": 4.7554971895526175e-05, | |
| "loss": 0.4367, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.4939106901217862, | |
| "grad_norm": 0.3564039577666558, | |
| "learning_rate": 4.705679065031235e-05, | |
| "loss": 0.443, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.5047361299052775, | |
| "grad_norm": 0.31533444364126456, | |
| "learning_rate": 4.6557477010446206e-05, | |
| "loss": 0.4413, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.5155615696887685, | |
| "grad_norm": 0.4443341910745228, | |
| "learning_rate": 4.605711110018307e-05, | |
| "loss": 0.441, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.5263870094722598, | |
| "grad_norm": 0.34010263018035086, | |
| "learning_rate": 4.555577321263477e-05, | |
| "loss": 0.4374, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.537212449255751, | |
| "grad_norm": 0.33919807935275176, | |
| "learning_rate": 4.505354379688518e-05, | |
| "loss": 0.4379, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.548037889039242, | |
| "grad_norm": 0.3528543669249499, | |
| "learning_rate": 4.4550503445080606e-05, | |
| "loss": 0.4441, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.5588633288227334, | |
| "grad_norm": 0.27772886705003613, | |
| "learning_rate": 4.4046732879497295e-05, | |
| "loss": 0.4364, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.5696887686062246, | |
| "grad_norm": 0.34956265254322005, | |
| "learning_rate": 4.354231293958801e-05, | |
| "loss": 0.4342, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.5805142083897157, | |
| "grad_norm": 0.3811128302087095, | |
| "learning_rate": 4.3037324569009854e-05, | |
| "loss": 0.4346, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.5913396481732072, | |
| "grad_norm": 0.3273603589996229, | |
| "learning_rate": 4.2531848802635264e-05, | |
| "loss": 0.4303, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.6021650879566982, | |
| "grad_norm": 0.28942663396412893, | |
| "learning_rate": 4.202596675354851e-05, | |
| "loss": 0.4501, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.6129905277401895, | |
| "grad_norm": 0.29314299543457817, | |
| "learning_rate": 4.151975960002958e-05, | |
| "loss": 0.4322, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.6238159675236807, | |
| "grad_norm": 0.3009437148190999, | |
| "learning_rate": 4.101330857252752e-05, | |
| "loss": 0.4393, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.6346414073071718, | |
| "grad_norm": 0.26040476529158696, | |
| "learning_rate": 4.050669494062561e-05, | |
| "loss": 0.4398, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.645466847090663, | |
| "grad_norm": 0.2615430242126928, | |
| "learning_rate": 4e-05, | |
| "loss": 0.4333, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.6562922868741543, | |
| "grad_norm": 0.29882957344640304, | |
| "learning_rate": 3.9493305059374405e-05, | |
| "loss": 0.4464, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.6671177266576453, | |
| "grad_norm": 0.25317331378708147, | |
| "learning_rate": 3.8986691427472496e-05, | |
| "loss": 0.4386, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.6779431664411368, | |
| "grad_norm": 0.2431387634446952, | |
| "learning_rate": 3.8480240399970436e-05, | |
| "loss": 0.4315, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.6887686062246279, | |
| "grad_norm": 0.2622280156077215, | |
| "learning_rate": 3.7974033246451496e-05, | |
| "loss": 0.4395, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.699594046008119, | |
| "grad_norm": 0.3288283111265646, | |
| "learning_rate": 3.746815119736475e-05, | |
| "loss": 0.4258, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.7104194857916104, | |
| "grad_norm": 0.21372843881028097, | |
| "learning_rate": 3.696267543099016e-05, | |
| "loss": 0.4364, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.7212449255751014, | |
| "grad_norm": 0.24709372213160263, | |
| "learning_rate": 3.6457687060412e-05, | |
| "loss": 0.4435, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.7320703653585927, | |
| "grad_norm": 0.21109175516844855, | |
| "learning_rate": 3.595326712050272e-05, | |
| "loss": 0.4372, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.742895805142084, | |
| "grad_norm": 0.25642603072599424, | |
| "learning_rate": 3.5449496554919414e-05, | |
| "loss": 0.4341, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.753721244925575, | |
| "grad_norm": 0.18539949931204602, | |
| "learning_rate": 3.494645620311484e-05, | |
| "loss": 0.44, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.7645466847090663, | |
| "grad_norm": 0.21475008841196183, | |
| "learning_rate": 3.444422678736525e-05, | |
| "loss": 0.4329, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.7753721244925575, | |
| "grad_norm": 0.2632871534174217, | |
| "learning_rate": 3.394288889981695e-05, | |
| "loss": 0.4308, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.7861975642760486, | |
| "grad_norm": 0.21696836902341513, | |
| "learning_rate": 3.34425229895538e-05, | |
| "loss": 0.4376, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.79702300405954, | |
| "grad_norm": 0.26623981980485834, | |
| "learning_rate": 3.294320934968768e-05, | |
| "loss": 0.4304, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.8078484438430311, | |
| "grad_norm": 0.2164206058578031, | |
| "learning_rate": 3.2445028104473845e-05, | |
| "loss": 0.4366, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.8186738836265224, | |
| "grad_norm": 0.25869927615505595, | |
| "learning_rate": 3.194805919645359e-05, | |
| "loss": 0.4333, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.8294993234100136, | |
| "grad_norm": 0.19684106352344385, | |
| "learning_rate": 3.145238237362596e-05, | |
| "loss": 0.4414, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.8403247631935047, | |
| "grad_norm": 0.20359484794712124, | |
| "learning_rate": 3.0958077176650606e-05, | |
| "loss": 0.4323, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.851150202976996, | |
| "grad_norm": 0.18191803193358724, | |
| "learning_rate": 3.0465222926084036e-05, | |
| "loss": 0.4283, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.8619756427604872, | |
| "grad_norm": 0.20053975230176904, | |
| "learning_rate": 2.997389870965118e-05, | |
| "loss": 0.4292, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.8728010825439783, | |
| "grad_norm": 0.19458549890449223, | |
| "learning_rate": 2.948418336955421e-05, | |
| "loss": 0.4237, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.8836265223274695, | |
| "grad_norm": 0.1810208465637645, | |
| "learning_rate": 2.899615548982088e-05, | |
| "loss": 0.4359, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.8944519621109608, | |
| "grad_norm": 0.19951878692554267, | |
| "learning_rate": 2.8509893383694213e-05, | |
| "loss": 0.4298, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.9052774018944518, | |
| "grad_norm": 0.1873602918054878, | |
| "learning_rate": 2.8025475081065684e-05, | |
| "loss": 0.4252, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.9161028416779433, | |
| "grad_norm": 0.21424662828484559, | |
| "learning_rate": 2.754297831595385e-05, | |
| "loss": 0.426, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.9269282814614344, | |
| "grad_norm": 0.22062924549343704, | |
| "learning_rate": 2.7062480514030478e-05, | |
| "loss": 0.4362, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.9377537212449256, | |
| "grad_norm": 0.2185890922228221, | |
| "learning_rate": 2.658405878019612e-05, | |
| "loss": 0.4301, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.9485791610284169, | |
| "grad_norm": 0.2074067861622232, | |
| "learning_rate": 2.6107789886207195e-05, | |
| "loss": 0.4329, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.959404600811908, | |
| "grad_norm": 0.17987563674452678, | |
| "learning_rate": 2.563375025835652e-05, | |
| "loss": 0.4315, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.9702300405953992, | |
| "grad_norm": 0.1966429768916275, | |
| "learning_rate": 2.5162015965209295e-05, | |
| "loss": 0.4285, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.9810554803788905, | |
| "grad_norm": 0.16025413751219228, | |
| "learning_rate": 2.4692662705396412e-05, | |
| "loss": 0.4352, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.9918809201623815, | |
| "grad_norm": 0.1824113480199386, | |
| "learning_rate": 2.4225765795467267e-05, | |
| "loss": 0.4366, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.002706359945873, | |
| "grad_norm": 0.3072484857319511, | |
| "learning_rate": 2.376140015780372e-05, | |
| "loss": 0.6866, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.013531799729364, | |
| "grad_norm": 0.24070769469641556, | |
| "learning_rate": 2.3299640308597487e-05, | |
| "loss": 0.4053, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.024357239512855, | |
| "grad_norm": 0.18689164117902526, | |
| "learning_rate": 2.2840560345892518e-05, | |
| "loss": 0.4046, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.0351826792963466, | |
| "grad_norm": 0.23479456361657802, | |
| "learning_rate": 2.2384233937694626e-05, | |
| "loss": 0.4052, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.0460081190798376, | |
| "grad_norm": 0.19984450718107008, | |
| "learning_rate": 2.1930734310150116e-05, | |
| "loss": 0.3993, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.0568335588633286, | |
| "grad_norm": 0.21673886132694872, | |
| "learning_rate": 2.1480134235795173e-05, | |
| "loss": 0.404, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.06765899864682, | |
| "grad_norm": 0.1853122367099065, | |
| "learning_rate": 2.10325060218782e-05, | |
| "loss": 0.4092, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.078484438430311, | |
| "grad_norm": 0.23045910920678292, | |
| "learning_rate": 2.0587921498756768e-05, | |
| "loss": 0.4127, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.089309878213802, | |
| "grad_norm": 0.19591355548021902, | |
| "learning_rate": 2.014645200837105e-05, | |
| "loss": 0.4054, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.1001353179972937, | |
| "grad_norm": 0.19977890878163287, | |
| "learning_rate": 1.9708168392795718e-05, | |
| "loss": 0.3971, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.1109607577807847, | |
| "grad_norm": 0.2004713363308186, | |
| "learning_rate": 1.9273140982871936e-05, | |
| "loss": 0.4065, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.121786197564276, | |
| "grad_norm": 0.21075286637362836, | |
| "learning_rate": 1.8841439586921515e-05, | |
| "loss": 0.4056, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.1326116373477673, | |
| "grad_norm": 0.22168049552774613, | |
| "learning_rate": 1.841313347954482e-05, | |
| "loss": 0.4002, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.1434370771312583, | |
| "grad_norm": 0.17429194042979973, | |
| "learning_rate": 1.7988291390504348e-05, | |
| "loss": 0.3996, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.15426251691475, | |
| "grad_norm": 0.19385637822362273, | |
| "learning_rate": 1.7566981493695828e-05, | |
| "loss": 0.406, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.165087956698241, | |
| "grad_norm": 0.20865997182785667, | |
| "learning_rate": 1.71492713962083e-05, | |
| "loss": 0.4085, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.175913396481732, | |
| "grad_norm": 0.15240836641103603, | |
| "learning_rate": 1.673522812747544e-05, | |
| "loss": 0.4042, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.1867388362652234, | |
| "grad_norm": 0.17399041045507257, | |
| "learning_rate": 1.6324918128519306e-05, | |
| "loss": 0.4011, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.1975642760487144, | |
| "grad_norm": 0.16385878458723385, | |
| "learning_rate": 1.5918407241288678e-05, | |
| "loss": 0.4072, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.208389715832206, | |
| "grad_norm": 0.13571253259057953, | |
| "learning_rate": 1.5515760698093485e-05, | |
| "loss": 0.4064, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.219215155615697, | |
| "grad_norm": 0.14062791240457395, | |
| "learning_rate": 1.511704311113705e-05, | |
| "loss": 0.404, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.230040595399188, | |
| "grad_norm": 0.1466710812309978, | |
| "learning_rate": 1.4722318462147844e-05, | |
| "loss": 0.4047, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.2408660351826795, | |
| "grad_norm": 0.14236159352983788, | |
| "learning_rate": 1.4331650092112406e-05, | |
| "loss": 0.4085, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.2516914749661705, | |
| "grad_norm": 0.1492512421183965, | |
| "learning_rate": 1.394510069111112e-05, | |
| "loss": 0.3992, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.2625169147496615, | |
| "grad_norm": 0.13214088611895553, | |
| "learning_rate": 1.3562732288258377e-05, | |
| "loss": 0.4045, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.273342354533153, | |
| "grad_norm": 0.13004710791062976, | |
| "learning_rate": 1.3184606241748857e-05, | |
| "loss": 0.3996, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.284167794316644, | |
| "grad_norm": 0.1379992554189758, | |
| "learning_rate": 1.2810783229011486e-05, | |
| "loss": 0.4011, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.2949932341001356, | |
| "grad_norm": 0.12792260792325943, | |
| "learning_rate": 1.2441323236972536e-05, | |
| "loss": 0.4014, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.3058186738836266, | |
| "grad_norm": 0.12551832471668423, | |
| "learning_rate": 1.2076285552429642e-05, | |
| "loss": 0.4091, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.3166441136671176, | |
| "grad_norm": 0.12157726859049582, | |
| "learning_rate": 1.1715728752538103e-05, | |
| "loss": 0.4034, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.3274695534506087, | |
| "grad_norm": 0.12069426576395426, | |
| "learning_rate": 1.1359710695411086e-05, | |
| "loss": 0.4044, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.3382949932341, | |
| "grad_norm": 0.13997082917503958, | |
| "learning_rate": 1.100828851083521e-05, | |
| "loss": 0.4099, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.349120433017591, | |
| "grad_norm": 0.12256862298627384, | |
| "learning_rate": 1.0661518591102973e-05, | |
| "loss": 0.4038, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.3599458728010827, | |
| "grad_norm": 0.12729477088608893, | |
| "learning_rate": 1.0319456581963578e-05, | |
| "loss": 0.4114, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.3707713125845737, | |
| "grad_norm": 0.13294964285537444, | |
| "learning_rate": 9.982157373693502e-06, | |
| "loss": 0.4089, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.381596752368065, | |
| "grad_norm": 0.12133786615241259, | |
| "learning_rate": 9.649675092288366e-06, | |
| "loss": 0.396, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.3924221921515563, | |
| "grad_norm": 0.1270304631668414, | |
| "learning_rate": 9.322063090777331e-06, | |
| "loss": 0.4038, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.4032476319350473, | |
| "grad_norm": 0.1320928328419506, | |
| "learning_rate": 8.99937394066165e-06, | |
| "loss": 0.41, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.414073071718539, | |
| "grad_norm": 0.12628554847051537, | |
| "learning_rate": 8.681659423478587e-06, | |
| "loss": 0.4089, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.42489851150203, | |
| "grad_norm": 0.13268411657935994, | |
| "learning_rate": 8.368970522492064e-06, | |
| "loss": 0.4055, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.435723951285521, | |
| "grad_norm": 0.12710838810128255, | |
| "learning_rate": 8.06135741451146e-06, | |
| "loss": 0.4042, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.4465493910690124, | |
| "grad_norm": 0.12486799356135045, | |
| "learning_rate": 7.758869461839808e-06, | |
| "loss": 0.4069, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.4573748308525034, | |
| "grad_norm": 0.11589641701781822, | |
| "learning_rate": 7.461555204352655e-06, | |
| "loss": 0.4007, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.4682002706359945, | |
| "grad_norm": 0.1279624723064858, | |
| "learning_rate": 7.169462351708958e-06, | |
| "loss": 0.4027, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.479025710419486, | |
| "grad_norm": 0.11280014309470966, | |
| "learning_rate": 6.882637775695147e-06, | |
| "loss": 0.399, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.489851150202977, | |
| "grad_norm": 0.11503215242439603, | |
| "learning_rate": 6.60112750270371e-06, | |
| "loss": 0.4003, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.500676589986468, | |
| "grad_norm": 0.11639784194163648, | |
| "learning_rate": 6.324976706347317e-06, | |
| "loss": 0.4093, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.5115020297699595, | |
| "grad_norm": 0.10526301515587191, | |
| "learning_rate": 6.054229700209959e-06, | |
| "loss": 0.4072, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.5223274695534506, | |
| "grad_norm": 0.10905344992752257, | |
| "learning_rate": 5.788929930735916e-06, | |
| "loss": 0.4016, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.533152909336942, | |
| "grad_norm": 0.1076850596517059, | |
| "learning_rate": 5.529119970258014e-06, | |
| "loss": 0.4046, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.543978349120433, | |
| "grad_norm": 0.10088542435761866, | |
| "learning_rate": 5.274841510166062e-06, | |
| "loss": 0.3977, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.554803788903924, | |
| "grad_norm": 0.09759924893913105, | |
| "learning_rate": 5.026135354216717e-06, | |
| "loss": 0.3967, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.565629228687415, | |
| "grad_norm": 0.09849111853235039, | |
| "learning_rate": 4.783041411985716e-06, | |
| "loss": 0.4044, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.5764546684709067, | |
| "grad_norm": 0.09952216439787918, | |
| "learning_rate": 4.545598692463675e-06, | |
| "loss": 0.4028, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.5872801082543977, | |
| "grad_norm": 0.10024732745834458, | |
| "learning_rate": 4.3138452977963266e-06, | |
| "loss": 0.4005, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.598105548037889, | |
| "grad_norm": 0.10025937597236387, | |
| "learning_rate": 4.087818417170337e-06, | |
| "loss": 0.4075, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.60893098782138, | |
| "grad_norm": 0.1004809905553361, | |
| "learning_rate": 3.867554320845601e-06, | |
| "loss": 0.41, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.6197564276048713, | |
| "grad_norm": 0.09709409878445463, | |
| "learning_rate": 3.6530883543350038e-06, | |
| "loss": 0.4096, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.6305818673883627, | |
| "grad_norm": 0.09883426050159856, | |
| "learning_rate": 3.4444549327325325e-06, | |
| "loss": 0.3998, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.641407307171854, | |
| "grad_norm": 0.10815993060300129, | |
| "learning_rate": 3.241687535190776e-06, | |
| "loss": 0.3985, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.6522327469553453, | |
| "grad_norm": 0.10563998579889254, | |
| "learning_rate": 3.0448186995485307e-06, | |
| "loss": 0.3999, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.6630581867388363, | |
| "grad_norm": 0.09566056794573148, | |
| "learning_rate": 2.853880017109516e-06, | |
| "loss": 0.4003, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.6738836265223274, | |
| "grad_norm": 0.09005281867113217, | |
| "learning_rate": 2.6689021275729366e-06, | |
| "loss": 0.4064, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.6847090663058184, | |
| "grad_norm": 0.09770175833329853, | |
| "learning_rate": 2.489914714116788e-06, | |
| "loss": 0.4012, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.69553450608931, | |
| "grad_norm": 0.10115637396980438, | |
| "learning_rate": 2.316946498634605e-06, | |
| "loss": 0.3975, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.706359945872801, | |
| "grad_norm": 0.10005262850620415, | |
| "learning_rate": 2.1500252371265253e-06, | |
| "loss": 0.4084, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.7171853856562924, | |
| "grad_norm": 0.09928418099460232, | |
| "learning_rate": 1.989177715245307e-06, | |
| "loss": 0.4044, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.7280108254397835, | |
| "grad_norm": 0.09347435747222115, | |
| "learning_rate": 1.8344297439980475e-06, | |
| "loss": 0.4082, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.7388362652232745, | |
| "grad_norm": 0.09044247763169987, | |
| "learning_rate": 1.685806155604346e-06, | |
| "loss": 0.4069, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.749661705006766, | |
| "grad_norm": 0.09067861636765942, | |
| "learning_rate": 1.5433307995115043e-06, | |
| "loss": 0.401, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.760487144790257, | |
| "grad_norm": 0.08656832055717045, | |
| "learning_rate": 1.4070265385674176e-06, | |
| "loss": 0.3965, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.7713125845737485, | |
| "grad_norm": 0.08765335369318135, | |
| "learning_rate": 1.276915245351833e-06, | |
| "loss": 0.3956, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.7821380243572396, | |
| "grad_norm": 0.08586498239574822, | |
| "learning_rate": 1.1530177986664425e-06, | |
| "loss": 0.4097, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.7929634641407306, | |
| "grad_norm": 0.0852742067113504, | |
| "learning_rate": 1.0353540801845229e-06, | |
| "loss": 0.402, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.803788903924222, | |
| "grad_norm": 0.09264848868643123, | |
| "learning_rate": 9.239429712605274e-07, | |
| "loss": 0.4008, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.814614343707713, | |
| "grad_norm": 0.08260763330441036, | |
| "learning_rate": 8.188023499002206e-07, | |
| "loss": 0.4064, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.825439783491204, | |
| "grad_norm": 0.08245520268533377, | |
| "learning_rate": 7.199490878918314e-07, | |
| "loss": 0.3956, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.8362652232746957, | |
| "grad_norm": 0.08234361429294809, | |
| "learning_rate": 6.273990480986314e-07, | |
| "loss": 0.4009, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.8470906630581867, | |
| "grad_norm": 0.08368092436176568, | |
| "learning_rate": 5.411670819134651e-07, | |
| "loss": 0.4053, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.8579161028416777, | |
| "grad_norm": 0.08727445900118284, | |
| "learning_rate": 4.6126702687554483e-07, | |
| "loss": 0.4011, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.8687415426251692, | |
| "grad_norm": 0.08490599228995778, | |
| "learning_rate": 3.8771170444996895e-07, | |
| "loss": 0.4053, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.8795669824086603, | |
| "grad_norm": 0.08339809754319051, | |
| "learning_rate": 3.2051291797027925e-07, | |
| "loss": 0.4038, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.8903924221921518, | |
| "grad_norm": 0.08357288940308696, | |
| "learning_rate": 2.59681450744389e-07, | |
| "loss": 0.3995, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.901217861975643, | |
| "grad_norm": 0.0809308823366729, | |
| "learning_rate": 2.0522706432419382e-07, | |
| "loss": 0.4009, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.912043301759134, | |
| "grad_norm": 0.08231369213258534, | |
| "learning_rate": 1.5715849693916264e-07, | |
| "loss": 0.4027, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.9228687415426253, | |
| "grad_norm": 0.08198954303443719, | |
| "learning_rate": 1.1548346209410366e-07, | |
| "loss": 0.4036, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.9336941813261164, | |
| "grad_norm": 0.08331345127712436, | |
| "learning_rate": 8.020864733140343e-08, | |
| "loss": 0.4008, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.944519621109608, | |
| "grad_norm": 0.08275059935563994, | |
| "learning_rate": 5.133971315788966e-08, | |
| "loss": 0.4068, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.955345060893099, | |
| "grad_norm": 0.08385499559681316, | |
| "learning_rate": 2.8881292136468952e-08, | |
| "loss": 0.4042, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.96617050067659, | |
| "grad_norm": 0.08577557063889475, | |
| "learning_rate": 1.2836988142779228e-08, | |
| "loss": 0.4029, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.976995940460081, | |
| "grad_norm": 0.08158836488623979, | |
| "learning_rate": 3.209375786856761e-09, | |
| "loss": 0.4109, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.9878213802435725, | |
| "grad_norm": 0.07897929665596683, | |
| "learning_rate": 0.0, | |
| "loss": 0.3988, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.9878213802435725, | |
| "step": 276, | |
| "total_flos": 7.10230108051551e+18, | |
| "train_loss": 0.4706156641460847, | |
| "train_runtime": 60487.3598, | |
| "train_samples_per_second": 2.345, | |
| "train_steps_per_second": 0.005 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 276, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.10230108051551e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |