| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.0639897616381379, | |
| "eval_steps": 500, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00015997440409534473, | |
| "grad_norm": 131.17303466796875, | |
| "learning_rate": 0.0, | |
| "loss": 2.02, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00031994880819068947, | |
| "grad_norm": 158.60177612304688, | |
| "learning_rate": 1.3236987738623674e-07, | |
| "loss": 1.8427, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00047992321228603423, | |
| "grad_norm": 159.34176635742188, | |
| "learning_rate": 2.0980129188224264e-07, | |
| "loss": 1.8302, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0006398976163813789, | |
| "grad_norm": 160.90457153320312, | |
| "learning_rate": 2.647397547724735e-07, | |
| "loss": 1.8208, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0007998720204767237, | |
| "grad_norm": 163.9342803955078, | |
| "learning_rate": 3.073533372198984e-07, | |
| "loss": 1.8017, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0009598464245720685, | |
| "grad_norm": 165.28424072265625, | |
| "learning_rate": 3.421711692684794e-07, | |
| "loss": 1.7632, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0011198208286674132, | |
| "grad_norm": 154.68109130859375, | |
| "learning_rate": 3.7160922681241326e-07, | |
| "loss": 1.6268, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0012797952327627579, | |
| "grad_norm": 91.99791717529297, | |
| "learning_rate": 3.9710963215871023e-07, | |
| "loss": 1.3067, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0014397696368581027, | |
| "grad_norm": 91.15142059326172, | |
| "learning_rate": 4.1960258376448527e-07, | |
| "loss": 1.2644, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0015997440409534474, | |
| "grad_norm": 95.00140380859375, | |
| "learning_rate": 4.397232146061353e-07, | |
| "loss": 1.2029, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0017597184450487923, | |
| "grad_norm": 100.01094055175781, | |
| "learning_rate": 4.5792453918508963e-07, | |
| "loss": 1.1511, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.001919692849144137, | |
| "grad_norm": 75.83676147460938, | |
| "learning_rate": 4.745410466547161e-07, | |
| "loss": 0.7515, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.002079667253239482, | |
| "grad_norm": 38.47466278076172, | |
| "learning_rate": 4.898267517654968e-07, | |
| "loss": 0.5764, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0022396416573348264, | |
| "grad_norm": 35.91501998901367, | |
| "learning_rate": 5.0397910419865e-07, | |
| "loss": 0.5374, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.002399616061430171, | |
| "grad_norm": 35.598350524902344, | |
| "learning_rate": 5.17154629102141e-07, | |
| "loss": 0.5144, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0025595904655255157, | |
| "grad_norm": 38.198551177978516, | |
| "learning_rate": 5.29479509544947e-07, | |
| "loss": 0.4709, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.002719564869620861, | |
| "grad_norm": 48.83864974975586, | |
| "learning_rate": 5.410569551171063e-07, | |
| "loss": 0.4102, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0028795392737162055, | |
| "grad_norm": 50.42375946044922, | |
| "learning_rate": 5.519724611507219e-07, | |
| "loss": 0.3637, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.00303951367781155, | |
| "grad_norm": 26.880878448486328, | |
| "learning_rate": 5.622976441001489e-07, | |
| "loss": 0.325, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.003199488081906895, | |
| "grad_norm": 50.65513229370117, | |
| "learning_rate": 5.720930919923719e-07, | |
| "loss": 0.2883, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0033594624860022394, | |
| "grad_norm": 12.885412216186523, | |
| "learning_rate": 5.814105186946559e-07, | |
| "loss": 0.237, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0035194368900975845, | |
| "grad_norm": 10.100977897644043, | |
| "learning_rate": 5.902944165713264e-07, | |
| "loss": 0.2293, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.003679411294192929, | |
| "grad_norm": 7.237063407897949, | |
| "learning_rate": 5.98783341472312e-07, | |
| "loss": 0.2131, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.003839385698288274, | |
| "grad_norm": 5.179779529571533, | |
| "learning_rate": 6.069109240409529e-07, | |
| "loss": 0.2051, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0039993601023836185, | |
| "grad_norm": 3.619863271713257, | |
| "learning_rate": 6.147066744397969e-07, | |
| "loss": 0.1905, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.004159334506478964, | |
| "grad_norm": 3.3004238605499268, | |
| "learning_rate": 6.221966291517337e-07, | |
| "loss": 0.1855, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.004319308910574308, | |
| "grad_norm": 2.556411027908325, | |
| "learning_rate": 6.294038756467279e-07, | |
| "loss": 0.179, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.004479283314669653, | |
| "grad_norm": 3.1660890579223633, | |
| "learning_rate": 6.363489815848868e-07, | |
| "loss": 0.1762, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.004639257718764998, | |
| "grad_norm": 2.2603609561920166, | |
| "learning_rate": 6.430503486697051e-07, | |
| "loss": 0.1743, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.004799232122860342, | |
| "grad_norm": 3.1162281036376953, | |
| "learning_rate": 6.495245064883779e-07, | |
| "loss": 0.1727, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.004959206526955687, | |
| "grad_norm": 3.651181697845459, | |
| "learning_rate": 6.557863581532573e-07, | |
| "loss": 0.1671, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.0051191809310510315, | |
| "grad_norm": 2.72200345993042, | |
| "learning_rate": 6.618493869311837e-07, | |
| "loss": 0.1652, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.005279155335146377, | |
| "grad_norm": 2.1462044715881348, | |
| "learning_rate": 6.677258310673321e-07, | |
| "loss": 0.1698, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.005439129739241722, | |
| "grad_norm": 2.3578836917877197, | |
| "learning_rate": 6.734268325033431e-07, | |
| "loss": 0.1678, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.005599104143337066, | |
| "grad_norm": 3.18251633644104, | |
| "learning_rate": 6.789625640323117e-07, | |
| "loss": 0.1666, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.005759078547432411, | |
| "grad_norm": 2.3386144638061523, | |
| "learning_rate": 6.843423385369588e-07, | |
| "loss": 0.1606, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.005919052951527755, | |
| "grad_norm": 3.2586636543273926, | |
| "learning_rate": 6.895747032576225e-07, | |
| "loss": 0.1585, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.0060790273556231, | |
| "grad_norm": 3.786763906478882, | |
| "learning_rate": 6.946675214863857e-07, | |
| "loss": 0.1624, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.006239001759718445, | |
| "grad_norm": 2.901278495788574, | |
| "learning_rate": 6.996280436477395e-07, | |
| "loss": 0.16, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.00639897616381379, | |
| "grad_norm": 1.7482857704162598, | |
| "learning_rate": 7.044629693786087e-07, | |
| "loss": 0.1596, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.006558950567909135, | |
| "grad_norm": 2.393767833709717, | |
| "learning_rate": 7.091785019416827e-07, | |
| "loss": 0.1558, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.006718924972004479, | |
| "grad_norm": 3.0040743350982666, | |
| "learning_rate": 7.137803960808927e-07, | |
| "loss": 0.154, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.006878899376099824, | |
| "grad_norm": 2.8855950832366943, | |
| "learning_rate": 7.182740002451747e-07, | |
| "loss": 0.1568, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.007038873780195169, | |
| "grad_norm": 2.0385258197784424, | |
| "learning_rate": 7.22664293957563e-07, | |
| "loss": 0.1579, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.007198848184290513, | |
| "grad_norm": 1.4387277364730835, | |
| "learning_rate": 7.269559209843836e-07, | |
| "loss": 0.1487, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.007358822588385858, | |
| "grad_norm": 2.1492249965667725, | |
| "learning_rate": 7.311532188585489e-07, | |
| "loss": 0.1538, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.007518796992481203, | |
| "grad_norm": 1.820937991142273, | |
| "learning_rate": 7.352602452275264e-07, | |
| "loss": 0.1495, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.007678771396576548, | |
| "grad_norm": 1.9386242628097534, | |
| "learning_rate": 7.392808014271896e-07, | |
| "loss": 0.1515, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.007838745800671893, | |
| "grad_norm": 2.2448389530181885, | |
| "learning_rate": 7.432184536248265e-07, | |
| "loss": 0.1516, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.007998720204767237, | |
| "grad_norm": 1.6324615478515625, | |
| "learning_rate": 7.470765518260336e-07, | |
| "loss": 0.1494, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.008158694608862581, | |
| "grad_norm": 1.9566981792449951, | |
| "learning_rate": 7.50858246999349e-07, | |
| "loss": 0.1499, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.008318669012957927, | |
| "grad_norm": 1.8566670417785645, | |
| "learning_rate": 7.545665065379704e-07, | |
| "loss": 0.147, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.008478643417053271, | |
| "grad_norm": 1.4080177545547485, | |
| "learning_rate": 7.582041282486483e-07, | |
| "loss": 0.1484, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.008638617821148616, | |
| "grad_norm": 2.2273147106170654, | |
| "learning_rate": 7.617737530329646e-07, | |
| "loss": 0.1505, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.008798592225243962, | |
| "grad_norm": 1.559259057044983, | |
| "learning_rate": 7.652778764049881e-07, | |
| "loss": 0.1469, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.008958566629339306, | |
| "grad_norm": 1.5281285047531128, | |
| "learning_rate": 7.687188589711236e-07, | |
| "loss": 0.1479, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.00911854103343465, | |
| "grad_norm": 1.222518801689148, | |
| "learning_rate": 7.720989359823916e-07, | |
| "loss": 0.1431, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.009278515437529996, | |
| "grad_norm": 1.6629137992858887, | |
| "learning_rate": 7.754202260559418e-07, | |
| "loss": 0.1463, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.00943848984162534, | |
| "grad_norm": 1.7817844152450562, | |
| "learning_rate": 7.786847391510249e-07, | |
| "loss": 0.1461, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.009598464245720684, | |
| "grad_norm": 1.7894457578659058, | |
| "learning_rate": 7.818943838746145e-07, | |
| "loss": 0.1475, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.009758438649816029, | |
| "grad_norm": 1.6237293481826782, | |
| "learning_rate": 7.850509741831755e-07, | |
| "loss": 0.1441, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.009918413053911375, | |
| "grad_norm": 1.5423122644424438, | |
| "learning_rate": 7.881562355394939e-07, | |
| "loss": 0.1448, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.010078387458006719, | |
| "grad_norm": 1.381689429283142, | |
| "learning_rate": 7.912118105768986e-07, | |
| "loss": 0.1415, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.010238361862102063, | |
| "grad_norm": 1.8088704347610474, | |
| "learning_rate": 7.942192643174205e-07, | |
| "loss": 0.1426, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.010398336266197409, | |
| "grad_norm": 2.4634368419647217, | |
| "learning_rate": 7.971800889853952e-07, | |
| "loss": 0.1459, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.010558310670292753, | |
| "grad_norm": 1.7363865375518799, | |
| "learning_rate": 8.000957084535689e-07, | |
| "loss": 0.1438, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.010718285074388097, | |
| "grad_norm": 1.5139596462249756, | |
| "learning_rate": 8.029674823548714e-07, | |
| "loss": 0.1411, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.010878259478483443, | |
| "grad_norm": 1.4979151487350464, | |
| "learning_rate": 8.057967098895799e-07, | |
| "loss": 0.142, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.011038233882578788, | |
| "grad_norm": 1.7654547691345215, | |
| "learning_rate": 8.085846333545548e-07, | |
| "loss": 0.1424, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.011198208286674132, | |
| "grad_norm": 1.9629215002059937, | |
| "learning_rate": 8.113324414185485e-07, | |
| "loss": 0.1458, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.011358182690769478, | |
| "grad_norm": 1.9196921586990356, | |
| "learning_rate": 8.140412721651974e-07, | |
| "loss": 0.1465, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.011518157094864822, | |
| "grad_norm": 2.098177909851074, | |
| "learning_rate": 8.167122159231955e-07, | |
| "loss": 0.1439, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.011678131498960166, | |
| "grad_norm": 1.7505509853363037, | |
| "learning_rate": 8.193463179012649e-07, | |
| "loss": 0.1421, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.01183810590305551, | |
| "grad_norm": 1.7455815076828003, | |
| "learning_rate": 8.219445806438593e-07, | |
| "loss": 0.1382, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.011998080307150856, | |
| "grad_norm": 1.5361164808273315, | |
| "learning_rate": 8.245079663220395e-07, | |
| "loss": 0.142, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.0121580547112462, | |
| "grad_norm": 2.7622509002685547, | |
| "learning_rate": 8.270373988726225e-07, | |
| "loss": 0.1474, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.012318029115341545, | |
| "grad_norm": 2.590261936187744, | |
| "learning_rate": 8.295337659975029e-07, | |
| "loss": 0.1442, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.01247800351943689, | |
| "grad_norm": 1.9166169166564941, | |
| "learning_rate": 8.319979210339763e-07, | |
| "loss": 0.1401, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.012637977923532235, | |
| "grad_norm": 1.3086128234863281, | |
| "learning_rate": 8.344306847059229e-07, | |
| "loss": 0.1392, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.01279795232762758, | |
| "grad_norm": 1.7761530876159668, | |
| "learning_rate": 8.368328467648454e-07, | |
| "loss": 0.1389, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.012957926731722925, | |
| "grad_norm": 1.8375579118728638, | |
| "learning_rate": 8.392051675289705e-07, | |
| "loss": 0.1399, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.01311790113581827, | |
| "grad_norm": 1.7359107732772827, | |
| "learning_rate": 8.415483793279194e-07, | |
| "loss": 0.1411, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.013277875539913614, | |
| "grad_norm": 1.9930511713027954, | |
| "learning_rate": 8.43863187859817e-07, | |
| "loss": 0.135, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.013437849944008958, | |
| "grad_norm": 1.8547563552856445, | |
| "learning_rate": 8.461502734671294e-07, | |
| "loss": 0.1365, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.013597824348104304, | |
| "grad_norm": 1.7807646989822388, | |
| "learning_rate": 8.484102923370048e-07, | |
| "loss": 0.1377, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.013757798752199648, | |
| "grad_norm": 2.3836963176727295, | |
| "learning_rate": 8.506438776314115e-07, | |
| "loss": 0.1392, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.013917773156294992, | |
| "grad_norm": 1.9490735530853271, | |
| "learning_rate": 8.528516405519477e-07, | |
| "loss": 0.1352, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.014077747560390338, | |
| "grad_norm": 1.6703108549118042, | |
| "learning_rate": 8.550341713437998e-07, | |
| "loss": 0.1404, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.014237721964485682, | |
| "grad_norm": 1.2594908475875854, | |
| "learning_rate": 8.571920402429763e-07, | |
| "loss": 0.1343, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.014397696368581027, | |
| "grad_norm": 1.4328629970550537, | |
| "learning_rate": 8.593257983706205e-07, | |
| "loss": 0.1399, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.014557670772676373, | |
| "grad_norm": 1.3369605541229248, | |
| "learning_rate": 8.6143597857791e-07, | |
| "loss": 0.1411, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.014717645176771717, | |
| "grad_norm": 1.4118566513061523, | |
| "learning_rate": 8.635230962447856e-07, | |
| "loss": 0.1378, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.014877619580867061, | |
| "grad_norm": 1.9336967468261719, | |
| "learning_rate": 8.655876500354998e-07, | |
| "loss": 0.1367, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.015037593984962405, | |
| "grad_norm": 1.9851653575897217, | |
| "learning_rate": 8.676301226137632e-07, | |
| "loss": 0.1398, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.015197568389057751, | |
| "grad_norm": 1.2664474248886108, | |
| "learning_rate": 8.696509813200474e-07, | |
| "loss": 0.1358, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.015357542793153095, | |
| "grad_norm": 2.4485650062561035, | |
| "learning_rate": 8.716506788134264e-07, | |
| "loss": 0.1409, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.01551751719724844, | |
| "grad_norm": 1.413927435874939, | |
| "learning_rate": 8.736296536801594e-07, | |
| "loss": 0.1342, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.015677491601343786, | |
| "grad_norm": 2.10976505279541, | |
| "learning_rate": 8.755883310110635e-07, | |
| "loss": 0.1428, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.01583746600543913, | |
| "grad_norm": 1.454032063484192, | |
| "learning_rate": 8.775271229495747e-07, | |
| "loss": 0.1377, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.015997440409534474, | |
| "grad_norm": 1.451311707496643, | |
| "learning_rate": 8.794464292122706e-07, | |
| "loss": 0.1366, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.016157414813629818, | |
| "grad_norm": 1.2541290521621704, | |
| "learning_rate": 8.813466375834887e-07, | |
| "loss": 0.1371, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.016317389217725162, | |
| "grad_norm": 1.826319694519043, | |
| "learning_rate": 8.832281243855856e-07, | |
| "loss": 0.1363, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.01647736362182051, | |
| "grad_norm": 1.3739150762557983, | |
| "learning_rate": 8.8509125492625e-07, | |
| "loss": 0.1389, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.016637338025915854, | |
| "grad_norm": 1.5730900764465332, | |
| "learning_rate": 8.869363839242072e-07, | |
| "loss": 0.1363, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.0167973124300112, | |
| "grad_norm": 1.4488335847854614, | |
| "learning_rate": 8.887638559145543e-07, | |
| "loss": 0.1331, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.016957286834106543, | |
| "grad_norm": 1.3076649904251099, | |
| "learning_rate": 8.905740056348848e-07, | |
| "loss": 0.1328, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.017117261238201887, | |
| "grad_norm": 1.6239938735961914, | |
| "learning_rate": 8.923671583932827e-07, | |
| "loss": 0.1339, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.01727723564229723, | |
| "grad_norm": 1.7959277629852295, | |
| "learning_rate": 8.941436304192014e-07, | |
| "loss": 0.1353, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.01743721004639258, | |
| "grad_norm": 1.9266117811203003, | |
| "learning_rate": 8.959037291981712e-07, | |
| "loss": 0.1377, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.017597184450487923, | |
| "grad_norm": 2.10133695602417, | |
| "learning_rate": 8.976477537912248e-07, | |
| "loss": 0.1359, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.017757158854583267, | |
| "grad_norm": 1.285016417503357, | |
| "learning_rate": 8.99375995139865e-07, | |
| "loss": 0.134, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.01791713325867861, | |
| "grad_norm": 1.2501912117004395, | |
| "learning_rate": 9.010887363573603e-07, | |
| "loss": 0.1353, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.018077107662773956, | |
| "grad_norm": 1.2144287824630737, | |
| "learning_rate": 9.027862530070898e-07, | |
| "loss": 0.132, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.0182370820668693, | |
| "grad_norm": 1.5820159912109375, | |
| "learning_rate": 9.044688133686283e-07, | |
| "loss": 0.1327, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.018397056470964644, | |
| "grad_norm": 1.3742053508758545, | |
| "learning_rate": 9.061366786922105e-07, | |
| "loss": 0.1358, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.018557030875059992, | |
| "grad_norm": 1.3718464374542236, | |
| "learning_rate": 9.077901034421785e-07, | |
| "loss": 0.136, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.018717005279155336, | |
| "grad_norm": 1.246771216392517, | |
| "learning_rate": 9.094293355299821e-07, | |
| "loss": 0.1336, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.01887697968325068, | |
| "grad_norm": 1.442410945892334, | |
| "learning_rate": 9.110546165372615e-07, | |
| "loss": 0.1335, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.019036954087346025, | |
| "grad_norm": 1.5307565927505493, | |
| "learning_rate": 9.126661819295197e-07, | |
| "loss": 0.1338, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.01919692849144137, | |
| "grad_norm": 1.809870958328247, | |
| "learning_rate": 9.142642612608513e-07, | |
| "loss": 0.1316, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.019356902895536713, | |
| "grad_norm": 1.5888453722000122, | |
| "learning_rate": 9.158490783701793e-07, | |
| "loss": 0.1339, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.019516877299632057, | |
| "grad_norm": 1.5183619260787964, | |
| "learning_rate": 9.174208515694123e-07, | |
| "loss": 0.1324, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.019676851703727405, | |
| "grad_norm": 1.18596351146698, | |
| "learning_rate": 9.189797938239252e-07, | |
| "loss": 0.1335, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.01983682610782275, | |
| "grad_norm": 0.9591464996337891, | |
| "learning_rate": 9.205261129257306e-07, | |
| "loss": 0.1284, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.019996800511918093, | |
| "grad_norm": 1.1490577459335327, | |
| "learning_rate": 9.220600116596953e-07, | |
| "loss": 0.1337, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.020156774916013438, | |
| "grad_norm": 1.3413796424865723, | |
| "learning_rate": 9.235816879631353e-07, | |
| "loss": 0.1318, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.020316749320108782, | |
| "grad_norm": 1.8173480033874512, | |
| "learning_rate": 9.250913350791019e-07, | |
| "loss": 0.1348, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.020476723724204126, | |
| "grad_norm": 1.4662258625030518, | |
| "learning_rate": 9.265891417036572e-07, | |
| "loss": 0.1333, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.020636698128299474, | |
| "grad_norm": 1.722621202468872, | |
| "learning_rate": 9.280752921274173e-07, | |
| "loss": 0.1337, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.020796672532394818, | |
| "grad_norm": 2.0989277362823486, | |
| "learning_rate": 9.295499663716319e-07, | |
| "loss": 0.1347, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.020956646936490162, | |
| "grad_norm": 1.9588260650634766, | |
| "learning_rate": 9.310133403190494e-07, | |
| "loss": 0.1346, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.021116621340585506, | |
| "grad_norm": 1.5125082731246948, | |
| "learning_rate": 9.324655858398056e-07, | |
| "loss": 0.1376, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.02127659574468085, | |
| "grad_norm": 2.068347454071045, | |
| "learning_rate": 9.339068709125623e-07, | |
| "loss": 0.1358, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.021436570148776195, | |
| "grad_norm": 1.3576561212539673, | |
| "learning_rate": 9.353373597411082e-07, | |
| "loss": 0.1341, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.02159654455287154, | |
| "grad_norm": 1.4498059749603271, | |
| "learning_rate": 9.367572128666264e-07, | |
| "loss": 0.1307, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.021756518956966887, | |
| "grad_norm": 1.832219123840332, | |
| "learning_rate": 9.381665872758167e-07, | |
| "loss": 0.1322, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.02191649336106223, | |
| "grad_norm": 1.6702324151992798, | |
| "learning_rate": 9.395656365050596e-07, | |
| "loss": 0.1341, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.022076467765157575, | |
| "grad_norm": 2.3712918758392334, | |
| "learning_rate": 9.409545107407915e-07, | |
| "loss": 0.132, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.02223644216925292, | |
| "grad_norm": 2.47230863571167, | |
| "learning_rate": 9.423333569162554e-07, | |
| "loss": 0.1384, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.022396416573348264, | |
| "grad_norm": 1.5399249792099, | |
| "learning_rate": 9.437023188047852e-07, | |
| "loss": 0.1293, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.022556390977443608, | |
| "grad_norm": 2.937544822692871, | |
| "learning_rate": 9.450615371097691e-07, | |
| "loss": 0.1364, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.022716365381538955, | |
| "grad_norm": 2.1480307579040527, | |
| "learning_rate": 9.464111495514342e-07, | |
| "loss": 0.1317, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.0228763397856343, | |
| "grad_norm": 1.650770902633667, | |
| "learning_rate": 9.477512909505864e-07, | |
| "loss": 0.1328, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.023036314189729644, | |
| "grad_norm": 1.477109432220459, | |
| "learning_rate": 9.490820933094322e-07, | |
| "loss": 0.1319, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.023196288593824988, | |
| "grad_norm": 2.036853551864624, | |
| "learning_rate": 9.504036858896035e-07, | |
| "loss": 0.1326, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.023356262997920332, | |
| "grad_norm": 1.4756255149841309, | |
| "learning_rate": 9.517161952875015e-07, | |
| "loss": 0.1324, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.023516237402015677, | |
| "grad_norm": 1.1842366456985474, | |
| "learning_rate": 9.53019745507069e-07, | |
| "loss": 0.1293, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.02367621180611102, | |
| "grad_norm": 1.311472773551941, | |
| "learning_rate": 9.54314458030096e-07, | |
| "loss": 0.1276, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.02383618621020637, | |
| "grad_norm": 1.8847229480743408, | |
| "learning_rate": 9.556004518841566e-07, | |
| "loss": 0.1314, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.023996160614301713, | |
| "grad_norm": 1.67835533618927, | |
| "learning_rate": 9.568778437082761e-07, | |
| "loss": 0.1321, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.024156135018397057, | |
| "grad_norm": 1.356938123703003, | |
| "learning_rate": 9.581467478164157e-07, | |
| "loss": 0.1309, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.0243161094224924, | |
| "grad_norm": 1.6580119132995605, | |
| "learning_rate": 9.594072762588592e-07, | |
| "loss": 0.1316, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.024476083826587745, | |
| "grad_norm": 1.3679087162017822, | |
| "learning_rate": 9.606595388815916e-07, | |
| "loss": 0.131, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.02463605823068309, | |
| "grad_norm": 1.2950249910354614, | |
| "learning_rate": 9.619036433837396e-07, | |
| "loss": 0.1297, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.024796032634778434, | |
| "grad_norm": 1.3981151580810547, | |
| "learning_rate": 9.631396953731557e-07, | |
| "loss": 0.1288, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.02495600703887378, | |
| "grad_norm": 1.2822041511535645, | |
| "learning_rate": 9.64367798420213e-07, | |
| "loss": 0.1307, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.025115981442969126, | |
| "grad_norm": 1.3140027523040771, | |
| "learning_rate": 9.65588054109883e-07, | |
| "loss": 0.1331, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.02527595584706447, | |
| "grad_norm": 1.517618179321289, | |
| "learning_rate": 9.668005620921596e-07, | |
| "loss": 0.1325, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.025435930251159814, | |
| "grad_norm": 1.2211989164352417, | |
| "learning_rate": 9.680054201308908e-07, | |
| "loss": 0.1314, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.02559590465525516, | |
| "grad_norm": 1.1967936754226685, | |
| "learning_rate": 9.692027241510822e-07, | |
| "loss": 0.129, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.025755879059350503, | |
| "grad_norm": 1.2219634056091309, | |
| "learning_rate": 9.703925682847255e-07, | |
| "loss": 0.131, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.02591585346344585, | |
| "grad_norm": 1.4189143180847168, | |
| "learning_rate": 9.71575044915207e-07, | |
| "loss": 0.1287, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.026075827867541194, | |
| "grad_norm": 1.3790985345840454, | |
| "learning_rate": 9.727502447203536e-07, | |
| "loss": 0.1301, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.02623580227163654, | |
| "grad_norm": 1.5076289176940918, | |
| "learning_rate": 9.739182567141562e-07, | |
| "loss": 0.1314, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.026395776675731883, | |
| "grad_norm": 1.3452013731002808, | |
| "learning_rate": 9.750791682872306e-07, | |
| "loss": 0.131, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.026555751079827227, | |
| "grad_norm": 1.466188669204712, | |
| "learning_rate": 9.762330652460538e-07, | |
| "loss": 0.1303, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.02671572548392257, | |
| "grad_norm": 1.9016494750976562, | |
| "learning_rate": 9.773800318510204e-07, | |
| "loss": 0.1311, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.026875699888017916, | |
| "grad_norm": 1.223925232887268, | |
| "learning_rate": 9.78520150853366e-07, | |
| "loss": 0.1295, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.027035674292113263, | |
| "grad_norm": 1.5380191802978516, | |
| "learning_rate": 9.796535035309936e-07, | |
| "loss": 0.1278, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.027195648696208607, | |
| "grad_norm": 1.4442285299301147, | |
| "learning_rate": 9.807801697232416e-07, | |
| "loss": 0.1324, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.02735562310030395, | |
| "grad_norm": 1.7630947828292847, | |
| "learning_rate": 9.819002278646342e-07, | |
| "loss": 0.1322, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.027515597504399296, | |
| "grad_norm": 1.2535282373428345, | |
| "learning_rate": 9.830137550176481e-07, | |
| "loss": 0.1315, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.02767557190849464, | |
| "grad_norm": 1.4018826484680176, | |
| "learning_rate": 9.841208269045278e-07, | |
| "loss": 0.1332, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.027835546312589984, | |
| "grad_norm": 1.1194539070129395, | |
| "learning_rate": 9.852215179381846e-07, | |
| "loss": 0.1304, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.027995520716685332, | |
| "grad_norm": 1.3209128379821777, | |
| "learning_rate": 9.863159012522102e-07, | |
| "loss": 0.1318, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.028155495120780676, | |
| "grad_norm": 1.637170672416687, | |
| "learning_rate": 9.874040487300364e-07, | |
| "loss": 0.128, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.02831546952487602, | |
| "grad_norm": 2.5420658588409424, | |
| "learning_rate": 9.884860310332674e-07, | |
| "loss": 0.1307, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.028475443928971365, | |
| "grad_norm": 1.7741079330444336, | |
| "learning_rate": 9.89561917629213e-07, | |
| "loss": 0.1309, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.02863541833306671, | |
| "grad_norm": 1.5465468168258667, | |
| "learning_rate": 9.906317768176538e-07, | |
| "loss": 0.1295, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.028795392737162053, | |
| "grad_norm": 1.5184584856033325, | |
| "learning_rate": 9.916956757568572e-07, | |
| "loss": 0.1302, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.028955367141257397, | |
| "grad_norm": 1.4520975351333618, | |
| "learning_rate": 9.92753680488876e-07, | |
| "loss": 0.1278, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.029115341545352745, | |
| "grad_norm": 1.275576114654541, | |
| "learning_rate": 9.93805855964147e-07, | |
| "loss": 0.1286, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.02927531594944809, | |
| "grad_norm": 1.4705110788345337, | |
| "learning_rate": 9.948522660654181e-07, | |
| "loss": 0.1278, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.029435290353543433, | |
| "grad_norm": 1.7193312644958496, | |
| "learning_rate": 9.958929736310225e-07, | |
| "loss": 0.1297, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.029595264757638778, | |
| "grad_norm": 1.483162522315979, | |
| "learning_rate": 9.969280404775208e-07, | |
| "loss": 0.1297, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.029755239161734122, | |
| "grad_norm": 1.191413402557373, | |
| "learning_rate": 9.979575274217366e-07, | |
| "loss": 0.1279, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.029915213565829466, | |
| "grad_norm": 1.3899563550949097, | |
| "learning_rate": 9.989814943021958e-07, | |
| "loss": 0.1308, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.03007518796992481, | |
| "grad_norm": 1.1342469453811646, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1244, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.030235162374020158, | |
| "grad_norm": 1.0823456048965454, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1296, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.030395136778115502, | |
| "grad_norm": 1.1805393695831299, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1267, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.030555111182210846, | |
| "grad_norm": 1.355592131614685, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1293, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.03071508558630619, | |
| "grad_norm": 1.7999428510665894, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1313, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.030875059990401535, | |
| "grad_norm": 1.5406898260116577, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1303, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.03103503439449688, | |
| "grad_norm": 1.6359665393829346, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1297, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.031195008798592227, | |
| "grad_norm": 1.3248281478881836, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1283, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.03135498320268757, | |
| "grad_norm": 1.4763351678848267, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1251, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.03151495760678291, | |
| "grad_norm": 1.383988618850708, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1327, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.03167493201087826, | |
| "grad_norm": 1.0943655967712402, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1305, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.03183490641497361, | |
| "grad_norm": 1.2787405252456665, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1273, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.03199488081906895, | |
| "grad_norm": 1.597632646560669, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1261, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.032154855223164296, | |
| "grad_norm": 1.0324817895889282, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1249, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.032314829627259636, | |
| "grad_norm": 1.453323245048523, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1266, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.032474804031354984, | |
| "grad_norm": 1.1299526691436768, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1237, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.032634778435450325, | |
| "grad_norm": 1.1185365915298462, | |
| "learning_rate": 1e-06, | |
| "loss": 0.129, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.03279475283954567, | |
| "grad_norm": 1.024652123451233, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1256, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.03295472724364102, | |
| "grad_norm": 1.4425640106201172, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1224, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.03311470164773636, | |
| "grad_norm": 1.5149112939834595, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1275, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.03327467605183171, | |
| "grad_norm": 1.4966243505477905, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1273, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.03343465045592705, | |
| "grad_norm": 1.7751713991165161, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1313, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.0335946248600224, | |
| "grad_norm": 1.3089805841445923, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1223, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.03375459926411774, | |
| "grad_norm": 1.116400957107544, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1288, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.033914573668213085, | |
| "grad_norm": 1.0689202547073364, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1266, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.03407454807230843, | |
| "grad_norm": 1.4078688621520996, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1247, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.034234522476403774, | |
| "grad_norm": 1.030319333076477, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1292, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.03439449688049912, | |
| "grad_norm": 1.2724783420562744, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1261, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.03455447128459446, | |
| "grad_norm": 1.1030813455581665, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1304, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.03471444568868981, | |
| "grad_norm": 1.212585210800171, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1279, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.03487442009278516, | |
| "grad_norm": 1.146895408630371, | |
| "learning_rate": 1e-06, | |
| "loss": 0.124, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.0350343944968805, | |
| "grad_norm": 1.102738857269287, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1278, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.035194368900975846, | |
| "grad_norm": 1.302136778831482, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1279, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.03535434330507119, | |
| "grad_norm": 1.228232502937317, | |
| "learning_rate": 1e-06, | |
| "loss": 0.128, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.035514317709166535, | |
| "grad_norm": 0.8978828191757202, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1232, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.035674292113261875, | |
| "grad_norm": 1.0016555786132812, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1261, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.03583426651735722, | |
| "grad_norm": 1.0512871742248535, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1239, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.03599424092145257, | |
| "grad_norm": 1.1789789199829102, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1224, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.03615421532554791, | |
| "grad_norm": 1.0191493034362793, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1218, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.03631418972964326, | |
| "grad_norm": 1.345475196838379, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1243, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.0364741641337386, | |
| "grad_norm": 2.024815082550049, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1266, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.03663413853783395, | |
| "grad_norm": 1.0629597902297974, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1244, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.03679411294192929, | |
| "grad_norm": 0.9979191422462463, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1242, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.036954087346024636, | |
| "grad_norm": 1.2679247856140137, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1224, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.037114061750119984, | |
| "grad_norm": 1.0288176536560059, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1256, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.037274036154215325, | |
| "grad_norm": 1.3102436065673828, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1262, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.03743401055831067, | |
| "grad_norm": 1.1388421058654785, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1236, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.03759398496240601, | |
| "grad_norm": 1.0879204273223877, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1268, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.03775395936650136, | |
| "grad_norm": 1.7273410558700562, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1279, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.0379139337705967, | |
| "grad_norm": 1.3870850801467896, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1262, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.03807390817469205, | |
| "grad_norm": 1.259337067604065, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1267, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.0382338825787874, | |
| "grad_norm": 1.4889863729476929, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1282, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.03839385698288274, | |
| "grad_norm": 1.2688918113708496, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1263, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.038553831386978085, | |
| "grad_norm": 1.057685375213623, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1221, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.038713805791073426, | |
| "grad_norm": 1.0359437465667725, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1285, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.038873780195168774, | |
| "grad_norm": 1.0317096710205078, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1236, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.039033754599264114, | |
| "grad_norm": 1.4080708026885986, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1256, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.03919372900335946, | |
| "grad_norm": 1.2293702363967896, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1225, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.03935370340745481, | |
| "grad_norm": 1.2161550521850586, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1273, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.03951367781155015, | |
| "grad_norm": 0.9967581629753113, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1232, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.0396736522156455, | |
| "grad_norm": 1.0089436769485474, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1252, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.03983362661974084, | |
| "grad_norm": 1.1048007011413574, | |
| "learning_rate": 1e-06, | |
| "loss": 0.123, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.03999360102383619, | |
| "grad_norm": 1.3069252967834473, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1275, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.040153575427931534, | |
| "grad_norm": 1.3315153121948242, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1262, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.040313549832026875, | |
| "grad_norm": 1.1331770420074463, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1243, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.04047352423612222, | |
| "grad_norm": 1.024278998374939, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1241, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.040633498640217564, | |
| "grad_norm": 1.151720404624939, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1229, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.04079347304431291, | |
| "grad_norm": 1.107699990272522, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1213, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.04095344744840825, | |
| "grad_norm": 1.2880581617355347, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1275, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.0411134218525036, | |
| "grad_norm": 1.247376561164856, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1266, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.04127339625659895, | |
| "grad_norm": 1.3562711477279663, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1218, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.04143337066069429, | |
| "grad_norm": 1.2490332126617432, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1242, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.041593345064789636, | |
| "grad_norm": 0.9172055125236511, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1253, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.041753319468884977, | |
| "grad_norm": 1.6371047496795654, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1223, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.041913293872980324, | |
| "grad_norm": 1.1314542293548584, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1226, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.042073268277075665, | |
| "grad_norm": 1.001624584197998, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1221, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.04223324268117101, | |
| "grad_norm": 1.015631079673767, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1234, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.04239321708526636, | |
| "grad_norm": 1.6776975393295288, | |
| "learning_rate": 1e-06, | |
| "loss": 0.127, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.0425531914893617, | |
| "grad_norm": 1.4266294240951538, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1254, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.04271316589345705, | |
| "grad_norm": 1.0626451969146729, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1199, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.04287314029755239, | |
| "grad_norm": 1.2195415496826172, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1216, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.04303311470164774, | |
| "grad_norm": 1.431025505065918, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1231, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.04319308910574308, | |
| "grad_norm": 0.9598281383514404, | |
| "learning_rate": 1e-06, | |
| "loss": 0.122, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.043353063509838426, | |
| "grad_norm": 1.2865647077560425, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1234, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.04351303791393377, | |
| "grad_norm": 1.2584363222122192, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1222, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.043673012318029114, | |
| "grad_norm": 1.6006242036819458, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1216, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.04383298672212446, | |
| "grad_norm": 1.14664626121521, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1238, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.0439929611262198, | |
| "grad_norm": 1.0439503192901611, | |
| "learning_rate": 1e-06, | |
| "loss": 0.123, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.04415293553031515, | |
| "grad_norm": 1.2464580535888672, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1228, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.04431290993441049, | |
| "grad_norm": 1.0424617528915405, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1242, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.04447288433850584, | |
| "grad_norm": 1.3555285930633545, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1192, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.044632858742601186, | |
| "grad_norm": 1.4239215850830078, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1258, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.04479283314669653, | |
| "grad_norm": 1.1337487697601318, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1235, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.044952807550791875, | |
| "grad_norm": 1.1882452964782715, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1192, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.045112781954887216, | |
| "grad_norm": 1.1663519144058228, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1241, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.04527275635898256, | |
| "grad_norm": 1.189931869506836, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1227, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.04543273076307791, | |
| "grad_norm": 1.3207061290740967, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1196, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.04559270516717325, | |
| "grad_norm": 1.5193383693695068, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1255, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.0457526795712686, | |
| "grad_norm": 0.9933753609657288, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1196, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.04591265397536394, | |
| "grad_norm": 0.9570215344429016, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1185, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.04607262837945929, | |
| "grad_norm": 1.136502742767334, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1201, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.04623260278355463, | |
| "grad_norm": 1.320960283279419, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1228, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.046392577187649976, | |
| "grad_norm": 1.0228922367095947, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1198, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.046552551591745324, | |
| "grad_norm": 0.9063114523887634, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1192, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.046712525995840665, | |
| "grad_norm": 1.1633445024490356, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1236, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.04687250039993601, | |
| "grad_norm": 1.2757052183151245, | |
| "learning_rate": 1e-06, | |
| "loss": 0.12, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.04703247480403135, | |
| "grad_norm": 1.551196575164795, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1241, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.0471924492081267, | |
| "grad_norm": 1.631995439529419, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1238, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.04735242361222204, | |
| "grad_norm": 1.4675942659378052, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1182, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.04751239801631739, | |
| "grad_norm": 1.1384583711624146, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1195, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.04767237242041274, | |
| "grad_norm": 0.9831569194793701, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1189, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.04783234682450808, | |
| "grad_norm": 1.4235044717788696, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1202, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.047992321228603425, | |
| "grad_norm": 1.4275295734405518, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1226, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.048152295632698766, | |
| "grad_norm": 1.2116954326629639, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1205, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.048312270036794114, | |
| "grad_norm": 1.479552149772644, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1215, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.048472244440889455, | |
| "grad_norm": 1.3289475440979004, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1215, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.0486322188449848, | |
| "grad_norm": 1.4380484819412231, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1214, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.04879219324908015, | |
| "grad_norm": 1.1001449823379517, | |
| "learning_rate": 1e-06, | |
| "loss": 0.12, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.04895216765317549, | |
| "grad_norm": 1.2235729694366455, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1171, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.04911214205727084, | |
| "grad_norm": 1.2234548330307007, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1227, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.04927211646136618, | |
| "grad_norm": 1.6779841184616089, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1225, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.04943209086546153, | |
| "grad_norm": 1.339382529258728, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1219, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.04959206526955687, | |
| "grad_norm": 0.8880839943885803, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1208, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.049752039673652215, | |
| "grad_norm": 1.1524003744125366, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1212, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.04991201407774756, | |
| "grad_norm": 1.2433319091796875, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1196, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.050071988481842904, | |
| "grad_norm": 1.299512505531311, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1226, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.05023196288593825, | |
| "grad_norm": 1.0843769311904907, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1209, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.05039193729003359, | |
| "grad_norm": 1.617997169494629, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1212, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.05055191169412894, | |
| "grad_norm": 1.2156031131744385, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1185, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.05071188609822429, | |
| "grad_norm": 1.658170461654663, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1264, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.05087186050231963, | |
| "grad_norm": 1.341801404953003, | |
| "learning_rate": 1e-06, | |
| "loss": 0.124, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.051031834906414976, | |
| "grad_norm": 1.3077876567840576, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1217, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.05119180931051032, | |
| "grad_norm": 1.2550292015075684, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1218, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.051351783714605664, | |
| "grad_norm": 0.9955998063087463, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1205, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.051511758118701005, | |
| "grad_norm": 1.0665310621261597, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1204, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.05167173252279635, | |
| "grad_norm": 0.8823798894882202, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1184, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.0518317069268917, | |
| "grad_norm": 1.2012176513671875, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1192, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.05199168133098704, | |
| "grad_norm": 1.2626415491104126, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1187, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.05215165573508239, | |
| "grad_norm": 1.4473867416381836, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1203, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.05231163013917773, | |
| "grad_norm": 1.1338282823562622, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1202, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.05247160454327308, | |
| "grad_norm": 1.3163774013519287, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1215, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.05263157894736842, | |
| "grad_norm": 1.1282827854156494, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1181, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.052791553351463766, | |
| "grad_norm": 0.8695802092552185, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1199, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.052951527755559114, | |
| "grad_norm": 0.9970802664756775, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1184, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.053111502159654454, | |
| "grad_norm": 0.9834699630737305, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1165, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.0532714765637498, | |
| "grad_norm": 1.0248655080795288, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1232, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.05343145096784514, | |
| "grad_norm": 1.1430097818374634, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1224, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.05359142537194049, | |
| "grad_norm": 1.295326828956604, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1228, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.05375139977603583, | |
| "grad_norm": 1.067046880722046, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1211, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.05391137418013118, | |
| "grad_norm": 1.057966947555542, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1165, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.054071348584226527, | |
| "grad_norm": 1.1167185306549072, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1221, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.05423132298832187, | |
| "grad_norm": 0.8829674124717712, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1185, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.054391297392417215, | |
| "grad_norm": 1.3520281314849854, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1214, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.054551271796512556, | |
| "grad_norm": 1.4164739847183228, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1184, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.0547112462006079, | |
| "grad_norm": 1.1109625101089478, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1199, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.054871220604703244, | |
| "grad_norm": 1.1353943347930908, | |
| "learning_rate": 1e-06, | |
| "loss": 0.118, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.05503119500879859, | |
| "grad_norm": 1.3248083591461182, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1207, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.05519116941289394, | |
| "grad_norm": 1.3598345518112183, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1178, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.05535114381698928, | |
| "grad_norm": 1.3764357566833496, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1185, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.05551111822108463, | |
| "grad_norm": 1.07554292678833, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1193, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.05567109262517997, | |
| "grad_norm": 1.3993921279907227, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1209, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.055831067029275316, | |
| "grad_norm": 1.208682894706726, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1188, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.055991041433370664, | |
| "grad_norm": 0.9713140726089478, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1187, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.056151015837466005, | |
| "grad_norm": 0.9984699487686157, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1146, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.05631099024156135, | |
| "grad_norm": 1.1007826328277588, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1201, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.05647096464565669, | |
| "grad_norm": 1.3435072898864746, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1177, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.05663093904975204, | |
| "grad_norm": 1.3530515432357788, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1193, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.05679091345384738, | |
| "grad_norm": 1.0946826934814453, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1154, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.05695088785794273, | |
| "grad_norm": 1.277107834815979, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1196, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.05711086226203808, | |
| "grad_norm": 1.668004035949707, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1201, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.05727083666613342, | |
| "grad_norm": 1.1211800575256348, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1162, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.057430811070228766, | |
| "grad_norm": 1.0412259101867676, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1174, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.057590785474324106, | |
| "grad_norm": 1.1593021154403687, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1197, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.057750759878419454, | |
| "grad_norm": 1.2287545204162598, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1205, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.057910734282514795, | |
| "grad_norm": 1.0975946187973022, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1147, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.05807070868661014, | |
| "grad_norm": 1.073879361152649, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1166, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.05823068309070549, | |
| "grad_norm": 1.0244841575622559, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1181, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.05839065749480083, | |
| "grad_norm": 1.0585442781448364, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1197, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.05855063189889618, | |
| "grad_norm": 0.9430259466171265, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1155, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.05871060630299152, | |
| "grad_norm": 0.9988534450531006, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1179, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.05887058070708687, | |
| "grad_norm": 1.4290752410888672, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1166, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.05903055511118221, | |
| "grad_norm": 1.2006633281707764, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1183, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.059190529515277555, | |
| "grad_norm": 1.0609732866287231, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1164, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.0593505039193729, | |
| "grad_norm": 1.2493401765823364, | |
| "learning_rate": 1e-06, | |
| "loss": 0.117, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.059510478323468244, | |
| "grad_norm": 1.0093525648117065, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1171, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.05967045272756359, | |
| "grad_norm": 1.1668469905853271, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1177, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.05983042713165893, | |
| "grad_norm": 1.0411137342453003, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1154, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.05999040153575428, | |
| "grad_norm": 1.7043784856796265, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1163, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.06015037593984962, | |
| "grad_norm": 1.1405059099197388, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1179, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.06031035034394497, | |
| "grad_norm": 1.1912827491760254, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1151, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.060470324748040316, | |
| "grad_norm": 1.0265625715255737, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1169, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.06063029915213566, | |
| "grad_norm": 1.2455023527145386, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1156, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.060790273556231005, | |
| "grad_norm": 1.0697797536849976, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1112, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.060950247960326345, | |
| "grad_norm": 1.2517694234848022, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1139, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.06111022236442169, | |
| "grad_norm": 1.1439390182495117, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1157, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.06127019676851704, | |
| "grad_norm": 1.0659351348876953, | |
| "learning_rate": 1e-06, | |
| "loss": 0.115, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.06143017117261238, | |
| "grad_norm": 1.1062361001968384, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1164, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.06159014557670773, | |
| "grad_norm": 1.227656364440918, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1177, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.06175011998080307, | |
| "grad_norm": 1.4164737462997437, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1148, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.06191009438489842, | |
| "grad_norm": 1.2910661697387695, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1167, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.06207006878899376, | |
| "grad_norm": 0.9569298624992371, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1148, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.062230043193089106, | |
| "grad_norm": 1.004881739616394, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1158, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.062390017597184454, | |
| "grad_norm": 1.2112892866134644, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1147, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.0625499920012798, | |
| "grad_norm": 1.5320332050323486, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1168, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.06270996640537514, | |
| "grad_norm": 1.3150376081466675, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1164, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.06286994080947049, | |
| "grad_norm": 1.4928754568099976, | |
| "learning_rate": 1e-06, | |
| "loss": 0.119, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.06302991521356582, | |
| "grad_norm": 1.0719698667526245, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1158, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.06318988961766117, | |
| "grad_norm": 1.4757615327835083, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1158, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.06334986402175652, | |
| "grad_norm": 1.4807626008987427, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1161, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.06350983842585187, | |
| "grad_norm": 1.0825022459030151, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1139, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.06366981282994721, | |
| "grad_norm": 0.9957918524742126, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1152, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.06382978723404255, | |
| "grad_norm": 0.9809920191764832, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1122, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.0639897616381379, | |
| "grad_norm": 1.0636423826217651, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1142, | |
| "step": 400 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 6251, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 400, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 64804028416000.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |