Invalid JSON:Unexpected token 'I', ..."ad_norm": Infinity,
"... is not valid JSON
| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9802814527681807, | |
| "eval_steps": 1000, | |
| "global_step": 47000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0021066823965618943, | |
| "grad_norm": 0.7183173298835754, | |
| "learning_rate": 1.225e-05, | |
| "loss": 3.2731781005859375, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.004213364793123789, | |
| "grad_norm": 0.6987810134887695, | |
| "learning_rate": 2.4750000000000002e-05, | |
| "loss": 3.2310821533203127, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.006320047189685683, | |
| "grad_norm": 0.7160114645957947, | |
| "learning_rate": 3.7250000000000004e-05, | |
| "loss": 3.2143896484375, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.008426729586247577, | |
| "grad_norm": 0.7347292900085449, | |
| "learning_rate": 4.975e-05, | |
| "loss": 3.1949951171875, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.010533411982809472, | |
| "grad_norm": 0.7420850992202759, | |
| "learning_rate": 4.999986741211794e-05, | |
| "loss": 3.1873373413085937, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.012640094379371365, | |
| "grad_norm": 0.6894696354866028, | |
| "learning_rate": 4.9999458771223715e-05, | |
| "loss": 3.1542486572265624, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.01474677677593326, | |
| "grad_norm": 0.7430497407913208, | |
| "learning_rate": 4.9998774026608505e-05, | |
| "loss": 3.1673529052734377, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.016853459172495155, | |
| "grad_norm": 0.7466180920600891, | |
| "learning_rate": 4.999781318583491e-05, | |
| "loss": 3.172335205078125, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.018960141569057048, | |
| "grad_norm": 0.7113528251647949, | |
| "learning_rate": 4.999657625951482e-05, | |
| "loss": 3.1631011962890625, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.021066823965618944, | |
| "grad_norm": 0.7085075974464417, | |
| "learning_rate": 4.9995063261309336e-05, | |
| "loss": 3.15373779296875, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.023173506362180837, | |
| "grad_norm": 0.7303823232650757, | |
| "learning_rate": 4.999327420792859e-05, | |
| "loss": 3.1551800537109376, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.02528018875874273, | |
| "grad_norm": 0.7048414945602417, | |
| "learning_rate": 4.999120911913159e-05, | |
| "loss": 3.144339599609375, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.027386871155304627, | |
| "grad_norm": 0.6965512633323669, | |
| "learning_rate": 4.998886801772597e-05, | |
| "loss": 3.135062255859375, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.02949355355186652, | |
| "grad_norm": 0.7184078097343445, | |
| "learning_rate": 4.9986250929567766e-05, | |
| "loss": 3.1415130615234377, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.031600235948428416, | |
| "grad_norm": 0.7143896222114563, | |
| "learning_rate": 4.998335788356112e-05, | |
| "loss": 3.148109130859375, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.03370691834499031, | |
| "grad_norm": 0.720639705657959, | |
| "learning_rate": 4.9980188911657936e-05, | |
| "loss": 3.1356280517578123, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.0358136007415522, | |
| "grad_norm": 0.7260337471961975, | |
| "learning_rate": 4.997674404885759e-05, | |
| "loss": 3.10700927734375, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.037920283138114096, | |
| "grad_norm": 0.7433862090110779, | |
| "learning_rate": 4.997302333320648e-05, | |
| "loss": 3.129173583984375, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.040026965534675996, | |
| "grad_norm": 0.6927924752235413, | |
| "learning_rate": 4.996902680579762e-05, | |
| "loss": 3.115341796875, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.04213364793123789, | |
| "grad_norm": 0.7047356963157654, | |
| "learning_rate": 4.996475451077022e-05, | |
| "loss": 3.1260260009765624, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.04213364793123789, | |
| "eval_loss": 2.956613063812256, | |
| "eval_runtime": 129.4336, | |
| "eval_samples_per_second": 118.54, | |
| "eval_steps_per_second": 3.708, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.04424033032779978, | |
| "grad_norm": 0.7373839616775513, | |
| "learning_rate": 4.9960206495309156e-05, | |
| "loss": 3.109400634765625, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.046347012724361675, | |
| "grad_norm": 0.700670063495636, | |
| "learning_rate": 4.995538280964449e-05, | |
| "loss": 3.115065612792969, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.04845369512092357, | |
| "grad_norm": 0.6800923943519592, | |
| "learning_rate": 4.995028350705086e-05, | |
| "loss": 3.103834228515625, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.05056037751748546, | |
| "grad_norm": 0.6795762181282043, | |
| "learning_rate": 4.9944908643846954e-05, | |
| "loss": 3.0806964111328123, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.05266705991404736, | |
| "grad_norm": 0.6991699934005737, | |
| "learning_rate": 4.993925827939484e-05, | |
| "loss": 3.1230282592773437, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.054773742310609254, | |
| "grad_norm": 0.7005462646484375, | |
| "learning_rate": 4.9933332476099336e-05, | |
| "loss": 3.0986212158203124, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.05688042470717115, | |
| "grad_norm": 0.7109544277191162, | |
| "learning_rate": 4.992713129940731e-05, | |
| "loss": 3.12240966796875, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.05898710710373304, | |
| "grad_norm": 0.7147893905639648, | |
| "learning_rate": 4.9920654817806975e-05, | |
| "loss": 3.113079833984375, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.06109378950029493, | |
| "grad_norm": 0.7276305556297302, | |
| "learning_rate": 4.99139031028271e-05, | |
| "loss": 3.10232421875, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.06320047189685683, | |
| "grad_norm": 0.6964511871337891, | |
| "learning_rate": 4.9906876229036234e-05, | |
| "loss": 3.091283874511719, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.06530715429341873, | |
| "grad_norm": 0.7357209920883179, | |
| "learning_rate": 4.9899574274041906e-05, | |
| "loss": 3.1123638916015626, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.06741383668998062, | |
| "grad_norm": 0.7081229090690613, | |
| "learning_rate": 4.989199731848975e-05, | |
| "loss": 3.0990875244140623, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.06952051908654251, | |
| "grad_norm": 0.7206143736839294, | |
| "learning_rate": 4.988414544606258e-05, | |
| "loss": 3.106065673828125, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.0716272014831044, | |
| "grad_norm": 0.7044396996498108, | |
| "learning_rate": 4.9876018743479535e-05, | |
| "loss": 3.118277893066406, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.0737338838796663, | |
| "grad_norm": 0.7007234692573547, | |
| "learning_rate": 4.986761730049507e-05, | |
| "loss": 3.107901611328125, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.07584056627622819, | |
| "grad_norm": 0.7111091613769531, | |
| "learning_rate": 4.985894120989798e-05, | |
| "loss": 3.0930078125, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.07794724867279008, | |
| "grad_norm": 0.6917787790298462, | |
| "learning_rate": 4.984999056751037e-05, | |
| "loss": 3.0914599609375, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.08005393106935199, | |
| "grad_norm": 0.7268196940422058, | |
| "learning_rate": 4.9840765472186624e-05, | |
| "loss": 3.096241455078125, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.08216061346591388, | |
| "grad_norm": 0.700495183467865, | |
| "learning_rate": 4.983126602581226e-05, | |
| "loss": 3.0989410400390627, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.08426729586247578, | |
| "grad_norm": 0.697629988193512, | |
| "learning_rate": 4.9821492333302845e-05, | |
| "loss": 3.0934375, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.08426729586247578, | |
| "eval_loss": 2.9592740535736084, | |
| "eval_runtime": 128.7893, | |
| "eval_samples_per_second": 119.133, | |
| "eval_steps_per_second": 3.727, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.08637397825903767, | |
| "grad_norm": 0.7469701170921326, | |
| "learning_rate": 4.981144450260283e-05, | |
| "loss": 3.0795492553710937, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.08848066065559956, | |
| "grad_norm": 0.7168535590171814, | |
| "learning_rate": 4.980112264468437e-05, | |
| "loss": 3.095859375, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.09058734305216146, | |
| "grad_norm": 0.7082321047782898, | |
| "learning_rate": 4.9790526873546056e-05, | |
| "loss": 3.090489501953125, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.09269402544872335, | |
| "grad_norm": 0.7243888974189758, | |
| "learning_rate": 4.9779657306211704e-05, | |
| "loss": 3.06650390625, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.09480070784528524, | |
| "grad_norm": 0.7053064107894897, | |
| "learning_rate": 4.976851406272902e-05, | |
| "loss": 3.0657330322265626, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.09690739024184714, | |
| "grad_norm": 0.6817116141319275, | |
| "learning_rate": 4.975709726616831e-05, | |
| "loss": 3.094239501953125, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.09901407263840903, | |
| "grad_norm": 0.7089473605155945, | |
| "learning_rate": 4.974540704262112e-05, | |
| "loss": 3.0672488403320313, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.10112075503497092, | |
| "grad_norm": 0.7032839059829712, | |
| "learning_rate": 4.97334435211988e-05, | |
| "loss": 3.0685894775390623, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.10322743743153283, | |
| "grad_norm": 0.7137486934661865, | |
| "learning_rate": 4.972120683403112e-05, | |
| "loss": 3.1034426879882813, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.10533411982809472, | |
| "grad_norm": 0.6994624733924866, | |
| "learning_rate": 4.97086971162648e-05, | |
| "loss": 3.079570617675781, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.10744080222465661, | |
| "grad_norm": 0.7199147343635559, | |
| "learning_rate": 4.9695914506062035e-05, | |
| "loss": 3.0958953857421876, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.10954748462121851, | |
| "grad_norm": 0.7233619689941406, | |
| "learning_rate": 4.968285914459891e-05, | |
| "loss": 3.0878173828125, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.1116541670177804, | |
| "grad_norm": 0.7326440811157227, | |
| "learning_rate": 4.966953117606391e-05, | |
| "loss": 3.0890634155273435, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.1137608494143423, | |
| "grad_norm": 0.6974560022354126, | |
| "learning_rate": 4.9655930747656286e-05, | |
| "loss": 3.0775030517578124, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.11586753181090419, | |
| "grad_norm": 0.7033624053001404, | |
| "learning_rate": 4.9642058009584444e-05, | |
| "loss": 3.0849713134765624, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.11797421420746608, | |
| "grad_norm": 0.6946061253547668, | |
| "learning_rate": 4.96279131150643e-05, | |
| "loss": 3.069232177734375, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.12008089660402797, | |
| "grad_norm": 0.7199132442474365, | |
| "learning_rate": 4.9613496220317536e-05, | |
| "loss": 3.0727505493164062, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.12218757900058987, | |
| "grad_norm": 0.7373772263526917, | |
| "learning_rate": 4.959880748456994e-05, | |
| "loss": 3.065733642578125, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.12429426139715176, | |
| "grad_norm": 0.7177034616470337, | |
| "learning_rate": 4.958384707004962e-05, | |
| "loss": 3.07641845703125, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.12640094379371367, | |
| "grad_norm": 0.6991363167762756, | |
| "learning_rate": 4.956861514198518e-05, | |
| "loss": 3.0732962036132814, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.12640094379371367, | |
| "eval_loss": 2.960636615753174, | |
| "eval_runtime": 129.7441, | |
| "eval_samples_per_second": 118.256, | |
| "eval_steps_per_second": 3.7, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.12850762619027556, | |
| "grad_norm": 0.7046240568161011, | |
| "learning_rate": 4.9553111868603944e-05, | |
| "loss": 3.072054748535156, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.13061430858683745, | |
| "grad_norm": 0.7022120356559753, | |
| "learning_rate": 4.953733742113008e-05, | |
| "loss": 3.0696722412109376, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.13272099098339935, | |
| "grad_norm": 0.7111489772796631, | |
| "learning_rate": 4.9521291973782705e-05, | |
| "loss": 3.08820556640625, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.13482767337996124, | |
| "grad_norm": 0.6908293962478638, | |
| "learning_rate": 4.950497570377397e-05, | |
| "loss": 3.0887225341796873, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.13693435577652313, | |
| "grad_norm": 0.6781262159347534, | |
| "learning_rate": 4.9488388791307096e-05, | |
| "loss": 3.068189697265625, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.13904103817308502, | |
| "grad_norm": 0.6849786639213562, | |
| "learning_rate": 4.9471531419574356e-05, | |
| "loss": 3.0767697143554686, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.14114772056964692, | |
| "grad_norm": 0.6951805949211121, | |
| "learning_rate": 4.945440377475513e-05, | |
| "loss": 3.064984130859375, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.1432544029662088, | |
| "grad_norm": 0.7169398069381714, | |
| "learning_rate": 4.943700604601375e-05, | |
| "loss": 3.073297119140625, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.1453610853627707, | |
| "grad_norm": 0.6996468305587769, | |
| "learning_rate": 4.9419338425497494e-05, | |
| "loss": 3.083929443359375, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.1474677677593326, | |
| "grad_norm": 0.7052185535430908, | |
| "learning_rate": 4.940140110833442e-05, | |
| "loss": 3.0852130126953123, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.1495744501558945, | |
| "grad_norm": 0.7242560386657715, | |
| "learning_rate": 4.93831942926312e-05, | |
| "loss": 3.087889404296875, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.15168113255245638, | |
| "grad_norm": 0.7300705909729004, | |
| "learning_rate": 4.936471817947099e-05, | |
| "loss": 3.038572998046875, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.15378781494901828, | |
| "grad_norm": 0.7155873775482178, | |
| "learning_rate": 4.934597297291114e-05, | |
| "loss": 3.0702294921875, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.15589449734558017, | |
| "grad_norm": 0.685499906539917, | |
| "learning_rate": 4.932695887998098e-05, | |
| "loss": 3.0512451171875, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.15800117974214206, | |
| "grad_norm": 0.7155754566192627, | |
| "learning_rate": 4.9307676110679535e-05, | |
| "loss": 3.0650628662109374, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.16010786213870398, | |
| "grad_norm": 0.7290592789649963, | |
| "learning_rate": 4.928812487797319e-05, | |
| "loss": 3.0623654174804686, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.16221454453526588, | |
| "grad_norm": 0.7060722708702087, | |
| "learning_rate": 4.926830539779335e-05, | |
| "loss": 3.0404345703125, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.16432122693182777, | |
| "grad_norm": 0.7160583734512329, | |
| "learning_rate": 4.924821788903402e-05, | |
| "loss": 3.032493896484375, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.16642790932838966, | |
| "grad_norm": 0.7150232195854187, | |
| "learning_rate": 4.922786257354947e-05, | |
| "loss": 3.0439007568359373, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.16853459172495155, | |
| "grad_norm": 0.712369978427887, | |
| "learning_rate": 4.920723967615168e-05, | |
| "loss": 3.04423095703125, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.16853459172495155, | |
| "eval_loss": 2.961803674697876, | |
| "eval_runtime": 124.6859, | |
| "eval_samples_per_second": 123.053, | |
| "eval_steps_per_second": 3.85, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.17064127412151345, | |
| "grad_norm": 0.7305163145065308, | |
| "learning_rate": 4.9186349424607934e-05, | |
| "loss": 3.074200439453125, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.17274795651807534, | |
| "grad_norm": 0.7282182574272156, | |
| "learning_rate": 4.916519204963828e-05, | |
| "loss": 3.0275787353515624, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.17485463891463723, | |
| "grad_norm": 0.7124244570732117, | |
| "learning_rate": 4.9143767784912965e-05, | |
| "loss": 3.058924560546875, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.17696132131119913, | |
| "grad_norm": 0.7103128433227539, | |
| "learning_rate": 4.9122076867049884e-05, | |
| "loss": 3.0691461181640625, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.17906800370776102, | |
| "grad_norm": 0.7060617208480835, | |
| "learning_rate": 4.910011953561195e-05, | |
| "loss": 3.0494329833984377, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.1811746861043229, | |
| "grad_norm": 0.7271001935005188, | |
| "learning_rate": 4.9077896033104455e-05, | |
| "loss": 3.055644226074219, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.1832813685008848, | |
| "grad_norm": 0.7033340334892273, | |
| "learning_rate": 4.905540660497238e-05, | |
| "loss": 3.050093994140625, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.1853880508974467, | |
| "grad_norm": 0.7435086369514465, | |
| "learning_rate": 4.903265149959768e-05, | |
| "loss": 3.0552102661132814, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.1874947332940086, | |
| "grad_norm": 0.7204045057296753, | |
| "learning_rate": 4.9009630968296584e-05, | |
| "loss": 3.0501980590820312, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.18960141569057049, | |
| "grad_norm": 0.7094113230705261, | |
| "learning_rate": 4.898634526531676e-05, | |
| "loss": 3.055870361328125, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.19170809808713238, | |
| "grad_norm": 0.6945235729217529, | |
| "learning_rate": 4.896279464783455e-05, | |
| "loss": 3.05148193359375, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.19381478048369427, | |
| "grad_norm": 0.721178412437439, | |
| "learning_rate": 4.893897937595211e-05, | |
| "loss": 3.0516928100585936, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.19592146288025616, | |
| "grad_norm": 0.7072880864143372, | |
| "learning_rate": 4.891489971269453e-05, | |
| "loss": 3.0531573486328125, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.19802814527681806, | |
| "grad_norm": 0.6985597014427185, | |
| "learning_rate": 4.889055592400697e-05, | |
| "loss": 3.0319134521484377, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.20013482767337995, | |
| "grad_norm": 0.6876893043518066, | |
| "learning_rate": 4.886594827875167e-05, | |
| "loss": 3.05040283203125, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.20224151006994184, | |
| "grad_norm": 0.7180494666099548, | |
| "learning_rate": 4.8841077048705004e-05, | |
| "loss": 3.05064697265625, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.20434819246650374, | |
| "grad_norm": 0.7060564160346985, | |
| "learning_rate": 4.881594250855449e-05, | |
| "loss": 3.060054931640625, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.20645487486306566, | |
| "grad_norm": 0.7116970419883728, | |
| "learning_rate": 4.879054493589574e-05, | |
| "loss": 3.05646484375, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.20856155725962755, | |
| "grad_norm": 0.7381977438926697, | |
| "learning_rate": 4.8764884611229404e-05, | |
| "loss": 3.0426617431640626, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.21066823965618944, | |
| "grad_norm": 0.7227292656898499, | |
| "learning_rate": 4.8738961817958056e-05, | |
| "loss": 3.070029296875, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.21066823965618944, | |
| "eval_loss": 2.959958791732788, | |
| "eval_runtime": 124.7155, | |
| "eval_samples_per_second": 123.024, | |
| "eval_steps_per_second": 3.849, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.21277492205275134, | |
| "grad_norm": 0.7169764637947083, | |
| "learning_rate": 4.871277684238308e-05, | |
| "loss": 3.06226318359375, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.21488160444931323, | |
| "grad_norm": 0.7150042057037354, | |
| "learning_rate": 4.868632997370151e-05, | |
| "loss": 3.05496337890625, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.21698828684587512, | |
| "grad_norm": 0.7077784538269043, | |
| "learning_rate": 4.865962150400282e-05, | |
| "loss": 3.0570599365234377, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.21909496924243702, | |
| "grad_norm": 0.7164002656936646, | |
| "learning_rate": 4.863265172826571e-05, | |
| "loss": 3.0604672241210937, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.2212016516389989, | |
| "grad_norm": 0.717158854007721, | |
| "learning_rate": 4.860542094435485e-05, | |
| "loss": 3.0436383056640626, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.2233083340355608, | |
| "grad_norm": 0.7141779065132141, | |
| "learning_rate": 4.857792945301758e-05, | |
| "loss": 3.042879638671875, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.2254150164321227, | |
| "grad_norm": 0.7105504870414734, | |
| "learning_rate": 4.85501775578806e-05, | |
| "loss": 3.040563049316406, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.2275216988286846, | |
| "grad_norm": 0.7205433249473572, | |
| "learning_rate": 4.852216556544659e-05, | |
| "loss": 3.0424761962890625, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.22962838122524648, | |
| "grad_norm": 0.7419236898422241, | |
| "learning_rate": 4.849389378509088e-05, | |
| "loss": 3.057305908203125, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 0.23173506362180837, | |
| "grad_norm": 0.7136011123657227, | |
| "learning_rate": 4.8465362529057966e-05, | |
| "loss": 3.05100341796875, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.23384174601837027, | |
| "grad_norm": 0.7066657543182373, | |
| "learning_rate": 4.843657211245811e-05, | |
| "loss": 3.0470831298828127, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 0.23594842841493216, | |
| "grad_norm": 0.7254061698913574, | |
| "learning_rate": 4.840752285326385e-05, | |
| "loss": 3.0324981689453123, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.23805511081149405, | |
| "grad_norm": 0.7086924314498901, | |
| "learning_rate": 4.837821507230648e-05, | |
| "loss": 3.0507342529296877, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 0.24016179320805595, | |
| "grad_norm": 0.7063336968421936, | |
| "learning_rate": 4.834864909327248e-05, | |
| "loss": 3.030625, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.24226847560461784, | |
| "grad_norm": 0.7155880928039551, | |
| "learning_rate": 4.83188252427e-05, | |
| "loss": 3.0367218017578126, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 0.24437515800117973, | |
| "grad_norm": 0.7138485908508301, | |
| "learning_rate": 4.8288743849975236e-05, | |
| "loss": 3.023646240234375, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.24648184039774163, | |
| "grad_norm": 0.7371372580528259, | |
| "learning_rate": 4.8258405247328734e-05, | |
| "loss": 3.0523486328125, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 0.24858852279430352, | |
| "grad_norm": 0.7491302490234375, | |
| "learning_rate": 4.82278097698318e-05, | |
| "loss": 3.0256768798828126, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.2506952051908654, | |
| "grad_norm": 0.7265546321868896, | |
| "learning_rate": 4.819695775539276e-05, | |
| "loss": 3.039449462890625, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 0.25280188758742733, | |
| "grad_norm": 0.7091307044029236, | |
| "learning_rate": 4.816584954475323e-05, | |
| "loss": 3.03005615234375, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.25280188758742733, | |
| "eval_loss": 2.9590306282043457, | |
| "eval_runtime": 124.9212, | |
| "eval_samples_per_second": 122.821, | |
| "eval_steps_per_second": 3.842, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.2549085699839892, | |
| "grad_norm": 0.7203435897827148, | |
| "learning_rate": 4.813448548148435e-05, | |
| "loss": 3.0497561645507814, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 0.2570152523805511, | |
| "grad_norm": 0.712683916091919, | |
| "learning_rate": 4.8102865911983e-05, | |
| "loss": 3.0374334716796874, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.259121934777113, | |
| "grad_norm": 0.737154483795166, | |
| "learning_rate": 4.8071631178264684e-05, | |
| "loss": 3.0338262939453124, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 0.2612286171736749, | |
| "grad_norm": 0.7363243103027344, | |
| "learning_rate": 4.8039506739404214e-05, | |
| "loss": 3.038179931640625, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.26333529957023677, | |
| "grad_norm": 0.7131369709968567, | |
| "learning_rate": 4.800712784329335e-05, | |
| "loss": 3.04284912109375, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 0.2654419819667987, | |
| "grad_norm": 0.7210983633995056, | |
| "learning_rate": 4.7974494847537186e-05, | |
| "loss": 3.037200622558594, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.26754866436336056, | |
| "grad_norm": 0.7105332016944885, | |
| "learning_rate": 4.7941608112547176e-05, | |
| "loss": 3.050509338378906, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 0.2696553467599225, | |
| "grad_norm": 0.7372371554374695, | |
| "learning_rate": 4.79084680015372e-05, | |
| "loss": 3.0444134521484374, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.27176202915648434, | |
| "grad_norm": 0.7344969511032104, | |
| "learning_rate": 4.787507488051949e-05, | |
| "loss": 3.0304275512695313, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 0.27386871155304626, | |
| "grad_norm": 0.728776216506958, | |
| "learning_rate": 4.7841429118300625e-05, | |
| "loss": 3.042907409667969, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.2759753939496082, | |
| "grad_norm": 0.7331499457359314, | |
| "learning_rate": 4.780753108647746e-05, | |
| "loss": 3.0423983764648437, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 0.27808207634617005, | |
| "grad_norm": 0.7019347548484802, | |
| "learning_rate": 4.777338115943301e-05, | |
| "loss": 3.0304049682617187, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.28018875874273197, | |
| "grad_norm": 0.7327196598052979, | |
| "learning_rate": 4.773897971433231e-05, | |
| "loss": 3.037353210449219, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 0.28229544113929383, | |
| "grad_norm": 0.724443793296814, | |
| "learning_rate": 4.770432713111828e-05, | |
| "loss": 3.027728271484375, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.28440212353585576, | |
| "grad_norm": 0.7023022174835205, | |
| "learning_rate": 4.7669423792507476e-05, | |
| "loss": 3.039205322265625, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 0.2865088059324176, | |
| "grad_norm": 0.6965822577476501, | |
| "learning_rate": 4.763427008398592e-05, | |
| "loss": 3.04097900390625, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.28861548832897954, | |
| "grad_norm": 0.710150420665741, | |
| "learning_rate": 4.7598866393804804e-05, | |
| "loss": 3.060494384765625, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 0.2907221707255414, | |
| "grad_norm": 0.7383567690849304, | |
| "learning_rate": 4.756321311297621e-05, | |
| "loss": 3.0389471435546875, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.2928288531221033, | |
| "grad_norm": 0.7048441767692566, | |
| "learning_rate": 4.752731063526882e-05, | |
| "loss": 3.03699951171875, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 0.2949355355186652, | |
| "grad_norm": 0.7232741713523865, | |
| "learning_rate": 4.7491159357203516e-05, | |
| "loss": 3.0090789794921875, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.2949355355186652, | |
| "eval_loss": 2.95876407623291, | |
| "eval_runtime": 125.3781, | |
| "eval_samples_per_second": 122.374, | |
| "eval_steps_per_second": 3.828, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.2970422179152271, | |
| "grad_norm": 0.7183651924133301, | |
| "learning_rate": 4.7454759678049034e-05, | |
| "loss": 3.0392718505859375, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 0.299148900311789, | |
| "grad_norm": 0.7372629046440125, | |
| "learning_rate": 4.741811199981755e-05, | |
| "loss": 3.032550048828125, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.3012555827083509, | |
| "grad_norm": 0.7527030110359192, | |
| "learning_rate": 4.738121672726025e-05, | |
| "loss": 3.0378472900390623, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 0.30336226510491276, | |
| "grad_norm": 0.7273029685020447, | |
| "learning_rate": 4.734407426786282e-05, | |
| "loss": 3.0402880859375, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.3054689475014747, | |
| "grad_norm": 0.7418842911720276, | |
| "learning_rate": 4.730668503184102e-05, | |
| "loss": 3.05841064453125, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 0.30757562989803655, | |
| "grad_norm": 0.7132033109664917, | |
| "learning_rate": 4.7269049432136035e-05, | |
| "loss": 3.039431457519531, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 0.30968231229459847, | |
| "grad_norm": 0.7326894998550415, | |
| "learning_rate": 4.723116788441007e-05, | |
| "loss": 3.051656494140625, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 0.31178899469116034, | |
| "grad_norm": 0.732483983039856, | |
| "learning_rate": 4.719304080704161e-05, | |
| "loss": 3.0274822998046873, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.31389567708772226, | |
| "grad_norm": 0.7055999040603638, | |
| "learning_rate": 4.71546686211209e-05, | |
| "loss": 3.014556884765625, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 0.3160023594842841, | |
| "grad_norm": 0.7180454730987549, | |
| "learning_rate": 4.7116051750445245e-05, | |
| "loss": 3.0318145751953125, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.31810904188084604, | |
| "grad_norm": 0.7178027629852295, | |
| "learning_rate": 4.707719062151434e-05, | |
| "loss": 3.0280184936523438, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 0.32021572427740796, | |
| "grad_norm": 0.6923954486846924, | |
| "learning_rate": 4.703808566352557e-05, | |
| "loss": 3.014013671875, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.32232240667396983, | |
| "grad_norm": 0.7218888401985168, | |
| "learning_rate": 4.699873730836926e-05, | |
| "loss": 3.0315927124023436, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 0.32442908907053175, | |
| "grad_norm": 0.7266069650650024, | |
| "learning_rate": 4.695914599062392e-05, | |
| "loss": 3.038184814453125, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 0.3265357714670936, | |
| "grad_norm": 0.71517413854599, | |
| "learning_rate": 4.691931214755141e-05, | |
| "loss": 3.019275817871094, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 0.32864245386365554, | |
| "grad_norm": 0.7297743558883667, | |
| "learning_rate": 4.687923621909216e-05, | |
| "loss": 3.0307427978515626, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.3307491362602174, | |
| "grad_norm": 0.7044832706451416, | |
| "learning_rate": 4.683891864786025e-05, | |
| "loss": 3.0128753662109373, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 0.3328558186567793, | |
| "grad_norm": 0.7655161023139954, | |
| "learning_rate": 4.6798359879138596e-05, | |
| "loss": 3.0257077026367187, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 0.3349625010533412, | |
| "grad_norm": 0.7163241505622864, | |
| "learning_rate": 4.675756036087397e-05, | |
| "loss": 3.0375396728515627, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 0.3370691834499031, | |
| "grad_norm": 0.7311005592346191, | |
| "learning_rate": 4.671652054367207e-05, | |
| "loss": 3.040563049316406, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.3370691834499031, | |
| "eval_loss": 2.9559757709503174, | |
| "eval_runtime": 125.7203, | |
| "eval_samples_per_second": 122.041, | |
| "eval_steps_per_second": 3.818, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.339175865846465, | |
| "grad_norm": 0.7203993201255798, | |
| "learning_rate": 4.667524088079257e-05, | |
| "loss": 3.04523193359375, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 0.3412825482430269, | |
| "grad_norm": 0.7433794736862183, | |
| "learning_rate": 4.663372182814407e-05, | |
| "loss": 3.0348739624023438, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.34338923063958876, | |
| "grad_norm": 0.7451913952827454, | |
| "learning_rate": 4.659196384427911e-05, | |
| "loss": 3.03673583984375, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 0.3454959130361507, | |
| "grad_norm": 0.7146168351173401, | |
| "learning_rate": 4.6550809653477414e-05, | |
| "loss": 3.04627685546875, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.34760259543271255, | |
| "grad_norm": 0.7409563660621643, | |
| "learning_rate": 4.650857994894889e-05, | |
| "loss": 3.04489501953125, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 0.34970927782927447, | |
| "grad_norm": 0.7260156273841858, | |
| "learning_rate": 4.646611269531936e-05, | |
| "loss": 3.0275424194335936, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 0.35181596022583633, | |
| "grad_norm": 0.7213459610939026, | |
| "learning_rate": 4.642340836161364e-05, | |
| "loss": 3.02736572265625, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 0.35392264262239825, | |
| "grad_norm": 0.7113795876502991, | |
| "learning_rate": 4.638046741947497e-05, | |
| "loss": 3.0393060302734374, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.3560293250189601, | |
| "grad_norm": 0.7787646651268005, | |
| "learning_rate": 4.6337290343159766e-05, | |
| "loss": 3.027384033203125, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 0.35813600741552204, | |
| "grad_norm": 0.7401567697525024, | |
| "learning_rate": 4.629387760953241e-05, | |
| "loss": 3.0250250244140626, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.3602426898120839, | |
| "grad_norm": 0.7389817833900452, | |
| "learning_rate": 4.6250229698059976e-05, | |
| "loss": 3.03439697265625, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 0.3623493722086458, | |
| "grad_norm": 0.7421659231185913, | |
| "learning_rate": 4.620634709080692e-05, | |
| "loss": 3.0239453125, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 0.36445605460520775, | |
| "grad_norm": 0.7072033882141113, | |
| "learning_rate": 4.616223027242981e-05, | |
| "loss": 3.0333843994140626, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 0.3665627370017696, | |
| "grad_norm": 0.7407162189483643, | |
| "learning_rate": 4.611787973017187e-05, | |
| "loss": 3.035455627441406, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.36866941939833153, | |
| "grad_norm": 0.7319179773330688, | |
| "learning_rate": 4.60732959538577e-05, | |
| "loss": 3.02468994140625, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 0.3707761017948934, | |
| "grad_norm": 0.7243380546569824, | |
| "learning_rate": 4.602847943588784e-05, | |
| "loss": 3.0051629638671873, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 0.3728827841914553, | |
| "grad_norm": 0.7503135204315186, | |
| "learning_rate": 4.598343067123328e-05, | |
| "loss": 3.004599609375, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 0.3749894665880172, | |
| "grad_norm": 0.7377141714096069, | |
| "learning_rate": 4.593815015743007e-05, | |
| "loss": 3.027547607421875, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 0.3770961489845791, | |
| "grad_norm": 0.6959289312362671, | |
| "learning_rate": 4.589263839457375e-05, | |
| "loss": 3.012759094238281, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 0.37920283138114097, | |
| "grad_norm": 0.7330695390701294, | |
| "learning_rate": 4.584689588531391e-05, | |
| "loss": 3.01300537109375, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.37920283138114097, | |
| "eval_loss": 2.9561047554016113, | |
| "eval_runtime": 125.3383, | |
| "eval_samples_per_second": 122.413, | |
| "eval_steps_per_second": 3.83, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.3813095137777029, | |
| "grad_norm": 0.7340759634971619, | |
| "learning_rate": 4.580092313484855e-05, | |
| "loss": 3.0100698852539063, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 0.38341619617426476, | |
| "grad_norm": 0.7195578813552856, | |
| "learning_rate": 4.575472065091857e-05, | |
| "loss": 3.002470703125, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 0.3855228785708267, | |
| "grad_norm": 0.7334921956062317, | |
| "learning_rate": 4.5708288943802116e-05, | |
| "loss": 3.029892578125, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 0.38762956096738854, | |
| "grad_norm": 0.748664915561676, | |
| "learning_rate": 4.5661628526308984e-05, | |
| "loss": 3.032738037109375, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 0.38973624336395046, | |
| "grad_norm": 0.7079982757568359, | |
| "learning_rate": 4.561473991377492e-05, | |
| "loss": 3.023472900390625, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 0.39184292576051233, | |
| "grad_norm": 0.7562114596366882, | |
| "learning_rate": 4.556762362405595e-05, | |
| "loss": 3.0230252075195314, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 0.39394960815707425, | |
| "grad_norm": 0.7204853892326355, | |
| "learning_rate": 4.5520280177522667e-05, | |
| "loss": 3.030900573730469, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 0.3960562905536361, | |
| "grad_norm": 0.7344993948936462, | |
| "learning_rate": 4.547271009705445e-05, | |
| "loss": 3.015901794433594, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 0.39816297295019804, | |
| "grad_norm": 0.7169850468635559, | |
| "learning_rate": 4.542491390803373e-05, | |
| "loss": 3.005645751953125, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 0.4002696553467599, | |
| "grad_norm": 0.7306743264198303, | |
| "learning_rate": 4.537689213834015e-05, | |
| "loss": 3.0204693603515627, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.4023763377433218, | |
| "grad_norm": 0.7161615490913391, | |
| "learning_rate": 4.532864531834478e-05, | |
| "loss": 3.0287905883789064, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 0.4044830201398837, | |
| "grad_norm": 0.7508752942085266, | |
| "learning_rate": 4.5280173980904194e-05, | |
| "loss": 3.0213604736328126, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 0.4065897025364456, | |
| "grad_norm": 0.7092886567115784, | |
| "learning_rate": 4.523147866135467e-05, | |
| "loss": 3.0126605224609375, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 0.4086963849330075, | |
| "grad_norm": 0.7355290651321411, | |
| "learning_rate": 4.518255989750619e-05, | |
| "loss": 3.04287109375, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 0.4108030673295694, | |
| "grad_norm": 0.749024510383606, | |
| "learning_rate": 4.513341822963657e-05, | |
| "loss": 3.0261370849609377, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 0.4129097497261313, | |
| "grad_norm": 0.7423744201660156, | |
| "learning_rate": 4.5084054200485446e-05, | |
| "loss": 3.0240179443359376, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 0.4150164321226932, | |
| "grad_norm": 0.7616035342216492, | |
| "learning_rate": 4.503446835524832e-05, | |
| "loss": 3.005235595703125, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 0.4171231145192551, | |
| "grad_norm": 0.7451797127723694, | |
| "learning_rate": 4.498466124157049e-05, | |
| "loss": 3.0096875, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 0.41922979691581697, | |
| "grad_norm": 0.701125979423523, | |
| "learning_rate": 4.493463340954105e-05, | |
| "loss": 3.0305224609375, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 0.4213364793123789, | |
| "grad_norm": 0.768416166305542, | |
| "learning_rate": 4.488438541168678e-05, | |
| "loss": 3.00571044921875, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.4213364793123789, | |
| "eval_loss": 2.9549593925476074, | |
| "eval_runtime": 118.9296, | |
| "eval_samples_per_second": 129.009, | |
| "eval_steps_per_second": 4.036, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.42344316170894075, | |
| "grad_norm": 0.7294694185256958, | |
| "learning_rate": 4.4833917802966056e-05, | |
| "loss": 3.018519287109375, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 0.4255498441055027, | |
| "grad_norm": 0.7414897084236145, | |
| "learning_rate": 4.478323114076274e-05, | |
| "loss": 3.031895751953125, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 0.42765652650206454, | |
| "grad_norm": 0.7361171245574951, | |
| "learning_rate": 4.4732325984879984e-05, | |
| "loss": 3.0415203857421873, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 0.42976320889862646, | |
| "grad_norm": 0.7098096609115601, | |
| "learning_rate": 4.468120289753409e-05, | |
| "loss": 2.997996826171875, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 0.4318698912951883, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.4629862443348246e-05, | |
| "loss": 3.0266912841796874, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 0.43397657369175024, | |
| "grad_norm": 0.7545154690742493, | |
| "learning_rate": 4.457933845538554e-05, | |
| "loss": 3.024451904296875, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 0.4360832560883121, | |
| "grad_norm": 0.7749207615852356, | |
| "learning_rate": 4.452756928999812e-05, | |
| "loss": 3.0368328857421876, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 0.43818993848487403, | |
| "grad_norm": 0.7240240573883057, | |
| "learning_rate": 4.447558445455993e-05, | |
| "loss": 3.0295269775390623, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 0.4402966208814359, | |
| "grad_norm": 0.733327329158783, | |
| "learning_rate": 4.4423384523211644e-05, | |
| "loss": 3.02707275390625, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 0.4424033032779978, | |
| "grad_norm": 0.7605792284011841, | |
| "learning_rate": 4.4370970072469575e-05, | |
| "loss": 3.0108306884765623, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.4445099856745597, | |
| "grad_norm": 0.7699744701385498, | |
| "learning_rate": 4.431834168121923e-05, | |
| "loss": 3.0396319580078126, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 0.4466166680711216, | |
| "grad_norm": 0.7354220747947693, | |
| "learning_rate": 4.426549993070899e-05, | |
| "loss": 3.0158978271484376, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 0.44872335046768347, | |
| "grad_norm": 0.7274487614631653, | |
| "learning_rate": 4.4212445404543654e-05, | |
| "loss": 3.0322219848632814, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 0.4508300328642454, | |
| "grad_norm": 0.7341009974479675, | |
| "learning_rate": 4.415917868867797e-05, | |
| "loss": 3.003311767578125, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 0.45293671526080725, | |
| "grad_norm": 0.7555328607559204, | |
| "learning_rate": 4.410570037141023e-05, | |
| "loss": 3.030812072753906, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 0.4550433976573692, | |
| "grad_norm": 0.7357707619667053, | |
| "learning_rate": 4.405201104337572e-05, | |
| "loss": 3.0104705810546877, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 0.4571500800539311, | |
| "grad_norm": 0.7414904236793518, | |
| "learning_rate": 4.399811129754018e-05, | |
| "loss": 3.0151333618164062, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 0.45925676245049296, | |
| "grad_norm": 0.7394636273384094, | |
| "learning_rate": 4.3944001729193345e-05, | |
| "loss": 3.003218688964844, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 0.4613634448470549, | |
| "grad_norm": 0.7744182348251343, | |
| "learning_rate": 4.388968293594226e-05, | |
| "loss": 3.0157513427734375, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 0.46347012724361675, | |
| "grad_norm": 0.7719476819038391, | |
| "learning_rate": 4.3835155517704755e-05, | |
| "loss": 3.027322082519531, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.46347012724361675, | |
| "eval_loss": 2.951545000076294, | |
| "eval_runtime": 118.9579, | |
| "eval_samples_per_second": 128.978, | |
| "eval_steps_per_second": 4.035, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.46557680964017867, | |
| "grad_norm": 0.7172340154647827, | |
| "learning_rate": 4.378042007670279e-05, | |
| "loss": 3.0138421630859376, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 0.46768349203674053, | |
| "grad_norm": 0.7303726673126221, | |
| "learning_rate": 4.372547721745582e-05, | |
| "loss": 3.001175537109375, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 0.46979017443330245, | |
| "grad_norm": 0.7603186964988708, | |
| "learning_rate": 4.3670327546774074e-05, | |
| "loss": 3.0162783813476564, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 0.4718968568298643, | |
| "grad_norm": 0.7521427869796753, | |
| "learning_rate": 4.361497167375193e-05, | |
| "loss": 2.99423095703125, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 0.47400353922642624, | |
| "grad_norm": 0.7517786622047424, | |
| "learning_rate": 4.355941020976111e-05, | |
| "loss": 3.019745178222656, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 0.4761102216229881, | |
| "grad_norm": 0.7296441197395325, | |
| "learning_rate": 4.3503643768443996e-05, | |
| "loss": 3.001739196777344, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 0.47821690401955, | |
| "grad_norm": 0.7638827562332153, | |
| "learning_rate": 4.344767296570678e-05, | |
| "loss": 3.013392028808594, | |
| "step": 11350 | |
| }, | |
| { | |
| "epoch": 0.4803235864161119, | |
| "grad_norm": 0.7276263236999512, | |
| "learning_rate": 4.339149841971274e-05, | |
| "loss": 3.0166070556640623, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 0.4824302688126738, | |
| "grad_norm": 0.7221875786781311, | |
| "learning_rate": 4.333512075087533e-05, | |
| "loss": 2.996725769042969, | |
| "step": 11450 | |
| }, | |
| { | |
| "epoch": 0.4845369512092357, | |
| "grad_norm": 0.7310070395469666, | |
| "learning_rate": 4.327854058185141e-05, | |
| "loss": 3.000296325683594, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.4866436336057976, | |
| "grad_norm": 0.7462596893310547, | |
| "learning_rate": 4.3221758537534324e-05, | |
| "loss": 3.02200439453125, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 0.48875031600235946, | |
| "grad_norm": 0.759739100933075, | |
| "learning_rate": 4.3164775245046986e-05, | |
| "loss": 3.010010986328125, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 0.4908569983989214, | |
| "grad_norm": 0.7589168548583984, | |
| "learning_rate": 4.310759133373499e-05, | |
| "loss": 3.007271728515625, | |
| "step": 11650 | |
| }, | |
| { | |
| "epoch": 0.49296368079548325, | |
| "grad_norm": 0.7378042340278625, | |
| "learning_rate": 4.305020743515963e-05, | |
| "loss": 3.025714111328125, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 0.49507036319204517, | |
| "grad_norm": 0.741910457611084, | |
| "learning_rate": 4.299262418309095e-05, | |
| "loss": 3.006583251953125, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 0.49717704558860704, | |
| "grad_norm": 0.7760137915611267, | |
| "learning_rate": 4.293484221350072e-05, | |
| "loss": 3.01525634765625, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 0.49928372798516896, | |
| "grad_norm": 0.7366864085197449, | |
| "learning_rate": 4.2876862164555434e-05, | |
| "loss": 3.025040283203125, | |
| "step": 11850 | |
| }, | |
| { | |
| "epoch": 0.5013904103817308, | |
| "grad_norm": 0.7541459798812866, | |
| "learning_rate": 4.2818684676609234e-05, | |
| "loss": 3.004857177734375, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 0.5034970927782928, | |
| "grad_norm": 0.751621663570404, | |
| "learning_rate": 4.276031039219688e-05, | |
| "loss": 3.0031988525390627, | |
| "step": 11950 | |
| }, | |
| { | |
| "epoch": 0.5056037751748547, | |
| "grad_norm": 0.7406221628189087, | |
| "learning_rate": 4.270173995602661e-05, | |
| "loss": 3.0164361572265626, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.5056037751748547, | |
| "eval_loss": 2.950573205947876, | |
| "eval_runtime": 118.9842, | |
| "eval_samples_per_second": 128.95, | |
| "eval_steps_per_second": 4.034, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.5077104575714165, | |
| "grad_norm": 0.7495478391647339, | |
| "learning_rate": 4.2642974014973036e-05, | |
| "loss": 3.002237548828125, | |
| "step": 12050 | |
| }, | |
| { | |
| "epoch": 0.5098171399679784, | |
| "grad_norm": 0.7460280060768127, | |
| "learning_rate": 4.258401321807002e-05, | |
| "loss": 3.0048617553710937, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 0.5119238223645404, | |
| "grad_norm": 0.7655878663063049, | |
| "learning_rate": 4.252485821650347e-05, | |
| "loss": 3.003458251953125, | |
| "step": 12150 | |
| }, | |
| { | |
| "epoch": 0.5140305047611022, | |
| "grad_norm": 0.757357656955719, | |
| "learning_rate": 4.2465509663604174e-05, | |
| "loss": 3.004598388671875, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 0.5161371871576641, | |
| "grad_norm": 0.7318490743637085, | |
| "learning_rate": 4.2405968214840575e-05, | |
| "loss": 3.0216680908203126, | |
| "step": 12250 | |
| }, | |
| { | |
| "epoch": 0.518243869554226, | |
| "grad_norm": 0.7530907392501831, | |
| "learning_rate": 4.2346234527811516e-05, | |
| "loss": 3.007520751953125, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 0.5203505519507879, | |
| "grad_norm": 0.7272404432296753, | |
| "learning_rate": 4.228750964074293e-05, | |
| "loss": 3.0171267700195314, | |
| "step": 12350 | |
| }, | |
| { | |
| "epoch": 0.5224572343473498, | |
| "grad_norm": 0.7328222393989563, | |
| "learning_rate": 4.222739727029933e-05, | |
| "loss": 3.0320773315429688, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 0.5245639167439117, | |
| "grad_norm": 0.7220742702484131, | |
| "learning_rate": 4.216709463379709e-05, | |
| "loss": 3.0238848876953126, | |
| "step": 12450 | |
| }, | |
| { | |
| "epoch": 0.5266705991404735, | |
| "grad_norm": 0.7630252242088318, | |
| "learning_rate": 4.210660239724189e-05, | |
| "loss": 2.9925677490234377, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.5287772815370355, | |
| "grad_norm": 0.7353333830833435, | |
| "learning_rate": 4.204592122873347e-05, | |
| "loss": 3.0201007080078126, | |
| "step": 12550 | |
| }, | |
| { | |
| "epoch": 0.5308839639335974, | |
| "grad_norm": 0.7585554122924805, | |
| "learning_rate": 4.1985051798458194e-05, | |
| "loss": 3.0046511840820314, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 0.5329906463301592, | |
| "grad_norm": 0.7346253991127014, | |
| "learning_rate": 4.1923994778681675e-05, | |
| "loss": 3.0252825927734377, | |
| "step": 12650 | |
| }, | |
| { | |
| "epoch": 0.5350973287267211, | |
| "grad_norm": 0.7465812563896179, | |
| "learning_rate": 4.186275084374131e-05, | |
| "loss": 2.9988339233398436, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 0.5372040111232831, | |
| "grad_norm": 0.7695534229278564, | |
| "learning_rate": 4.180132067003887e-05, | |
| "loss": 3.013394775390625, | |
| "step": 12750 | |
| }, | |
| { | |
| "epoch": 0.539310693519845, | |
| "grad_norm": 0.756037712097168, | |
| "learning_rate": 4.1739704936033044e-05, | |
| "loss": 3.00047119140625, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 0.5414173759164068, | |
| "grad_norm": 0.7378901243209839, | |
| "learning_rate": 4.1677904322231886e-05, | |
| "loss": 2.9985260009765624, | |
| "step": 12850 | |
| }, | |
| { | |
| "epoch": 0.5435240583129687, | |
| "grad_norm": 0.7396430969238281, | |
| "learning_rate": 4.161591951118536e-05, | |
| "loss": 3.01582275390625, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 0.5456307407095307, | |
| "grad_norm": 0.7484205365180969, | |
| "learning_rate": 4.1553751187477774e-05, | |
| "loss": 3.00086669921875, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 0.5477374231060925, | |
| "grad_norm": 0.7418805956840515, | |
| "learning_rate": 4.1491400037720194e-05, | |
| "loss": 2.985556640625, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.5477374231060925, | |
| "eval_loss": 2.9498443603515625, | |
| "eval_runtime": 118.9076, | |
| "eval_samples_per_second": 129.033, | |
| "eval_steps_per_second": 4.037, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.5498441055026544, | |
| "grad_norm": 0.7477849721908569, | |
| "learning_rate": 4.142886675054291e-05, | |
| "loss": 3.006175537109375, | |
| "step": 13050 | |
| }, | |
| { | |
| "epoch": 0.5519507878992164, | |
| "grad_norm": 0.7623112201690674, | |
| "learning_rate": 4.136615201658781e-05, | |
| "loss": 3.0172186279296875, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 0.5540574702957782, | |
| "grad_norm": 0.7711979746818542, | |
| "learning_rate": 4.130325652850075e-05, | |
| "loss": 2.9972796630859375, | |
| "step": 13150 | |
| }, | |
| { | |
| "epoch": 0.5561641526923401, | |
| "grad_norm": 0.7240932583808899, | |
| "learning_rate": 4.124018098092388e-05, | |
| "loss": 2.9975146484375, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 0.558270835088902, | |
| "grad_norm": 0.7590881586074829, | |
| "learning_rate": 4.117692607048804e-05, | |
| "loss": 3.007953796386719, | |
| "step": 13250 | |
| }, | |
| { | |
| "epoch": 0.5603775174854639, | |
| "grad_norm": 0.7551672458648682, | |
| "learning_rate": 4.1113492495804986e-05, | |
| "loss": 3.0047311401367187, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 0.5624841998820258, | |
| "grad_norm": 0.7241302132606506, | |
| "learning_rate": 4.104988095745974e-05, | |
| "loss": 2.968081359863281, | |
| "step": 13350 | |
| }, | |
| { | |
| "epoch": 0.5645908822785877, | |
| "grad_norm": 0.7135413289070129, | |
| "learning_rate": 4.0986092158002784e-05, | |
| "loss": 3.018740234375, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 0.5666975646751495, | |
| "grad_norm": 0.7576137781143188, | |
| "learning_rate": 4.092212680194239e-05, | |
| "loss": 2.992593994140625, | |
| "step": 13450 | |
| }, | |
| { | |
| "epoch": 0.5688042470717115, | |
| "grad_norm": 0.727250337600708, | |
| "learning_rate": 4.085798559573676e-05, | |
| "loss": 3.0134197998046877, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.5709109294682734, | |
| "grad_norm": 0.7550929188728333, | |
| "learning_rate": 4.0793669247786245e-05, | |
| "loss": 2.98953857421875, | |
| "step": 13550 | |
| }, | |
| { | |
| "epoch": 0.5730176118648352, | |
| "grad_norm": 0.7384030818939209, | |
| "learning_rate": 4.072917846842556e-05, | |
| "loss": 3.01269287109375, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 0.5751242942613971, | |
| "grad_norm": 0.7186519503593445, | |
| "learning_rate": 4.066451396991588e-05, | |
| "loss": 3.01514404296875, | |
| "step": 13650 | |
| }, | |
| { | |
| "epoch": 0.5772309766579591, | |
| "grad_norm": 0.7442272901535034, | |
| "learning_rate": 4.059967646643701e-05, | |
| "loss": 2.9887322998046875, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 0.579337659054521, | |
| "grad_norm": 0.7435423731803894, | |
| "learning_rate": 4.05346666740795e-05, | |
| "loss": 3.0130267333984375, | |
| "step": 13750 | |
| }, | |
| { | |
| "epoch": 0.5814443414510828, | |
| "grad_norm": 0.7321817874908447, | |
| "learning_rate": 4.046948531083672e-05, | |
| "loss": 2.9902178955078127, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 0.5835510238476447, | |
| "grad_norm": 0.7547968029975891, | |
| "learning_rate": 4.0404133096596917e-05, | |
| "loss": 3.0112127685546874, | |
| "step": 13850 | |
| }, | |
| { | |
| "epoch": 0.5856577062442067, | |
| "grad_norm": 0.7581467032432556, | |
| "learning_rate": 4.033861075313531e-05, | |
| "loss": 2.987550048828125, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 0.5877643886407685, | |
| "grad_norm": 0.7501505613327026, | |
| "learning_rate": 4.027291900410607e-05, | |
| "loss": 2.976295166015625, | |
| "step": 13950 | |
| }, | |
| { | |
| "epoch": 0.5898710710373304, | |
| "grad_norm": 0.7598941326141357, | |
| "learning_rate": 4.0207058575034366e-05, | |
| "loss": 2.9963720703125, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.5898710710373304, | |
| "eval_loss": 2.9478862285614014, | |
| "eval_runtime": 119.1263, | |
| "eval_samples_per_second": 128.796, | |
| "eval_steps_per_second": 4.029, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.5919777534338923, | |
| "grad_norm": 0.7242246270179749, | |
| "learning_rate": 4.0141030193308324e-05, | |
| "loss": 3.017314453125, | |
| "step": 14050 | |
| }, | |
| { | |
| "epoch": 0.5940844358304542, | |
| "grad_norm": 0.7500660419464111, | |
| "learning_rate": 4.007483458817101e-05, | |
| "loss": 3.000011291503906, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 0.5961911182270161, | |
| "grad_norm": 0.750382661819458, | |
| "learning_rate": 4.0008472490712355e-05, | |
| "loss": 3.004837646484375, | |
| "step": 14150 | |
| }, | |
| { | |
| "epoch": 0.598297800623578, | |
| "grad_norm": 0.7483128309249878, | |
| "learning_rate": 3.9941944633861126e-05, | |
| "loss": 3.0029196166992187, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 0.6004044830201399, | |
| "grad_norm": 0.7319640517234802, | |
| "learning_rate": 3.9875251752376765e-05, | |
| "loss": 2.9899969482421875, | |
| "step": 14250 | |
| }, | |
| { | |
| "epoch": 0.6025111654167018, | |
| "grad_norm": 0.7555121779441833, | |
| "learning_rate": 3.9808394582841345e-05, | |
| "loss": 2.984419860839844, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 0.6046178478132637, | |
| "grad_norm": 0.7269479036331177, | |
| "learning_rate": 3.9741373863651366e-05, | |
| "loss": 2.9930014038085937, | |
| "step": 14350 | |
| }, | |
| { | |
| "epoch": 0.6067245302098255, | |
| "grad_norm": 0.7246697545051575, | |
| "learning_rate": 3.967419033500966e-05, | |
| "loss": 3.01332763671875, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 0.6088312126063875, | |
| "grad_norm": 0.7506778240203857, | |
| "learning_rate": 3.960684473891719e-05, | |
| "loss": 3.005030517578125, | |
| "step": 14450 | |
| }, | |
| { | |
| "epoch": 0.6109378950029494, | |
| "grad_norm": 0.7472758293151855, | |
| "learning_rate": 3.954068953371231e-05, | |
| "loss": 2.998134765625, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.6130445773995112, | |
| "grad_norm": 0.7576263546943665, | |
| "learning_rate": 3.9473025240116325e-05, | |
| "loss": 2.966005859375, | |
| "step": 14550 | |
| }, | |
| { | |
| "epoch": 0.6151512597960731, | |
| "grad_norm": 0.73203045129776, | |
| "learning_rate": 3.9405201100814915e-05, | |
| "loss": 2.988453369140625, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 0.6172579421926351, | |
| "grad_norm": 0.7592427730560303, | |
| "learning_rate": 3.933721786488418e-05, | |
| "loss": 3.00134765625, | |
| "step": 14650 | |
| }, | |
| { | |
| "epoch": 0.6193646245891969, | |
| "grad_norm": 0.7523754239082336, | |
| "learning_rate": 3.9269076283157355e-05, | |
| "loss": 2.987010498046875, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 0.6214713069857588, | |
| "grad_norm": 0.7665632367134094, | |
| "learning_rate": 3.9200777108216505e-05, | |
| "loss": 2.9885418701171873, | |
| "step": 14750 | |
| }, | |
| { | |
| "epoch": 0.6235779893823207, | |
| "grad_norm": 0.7370322942733765, | |
| "learning_rate": 3.913232109438422e-05, | |
| "loss": 2.9893780517578126, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 0.6256846717788827, | |
| "grad_norm": 0.7885867953300476, | |
| "learning_rate": 3.906370899771526e-05, | |
| "loss": 2.986861572265625, | |
| "step": 14850 | |
| }, | |
| { | |
| "epoch": 0.6277913541754445, | |
| "grad_norm": 0.7267576456069946, | |
| "learning_rate": 3.8996318441698775e-05, | |
| "loss": 2.979769287109375, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 0.6298980365720064, | |
| "grad_norm": 0.7535339593887329, | |
| "learning_rate": 3.8927399538264976e-05, | |
| "loss": 3.0014642333984374, | |
| "step": 14950 | |
| }, | |
| { | |
| "epoch": 0.6320047189685682, | |
| "grad_norm": 0.7546215057373047, | |
| "learning_rate": 3.885832681522767e-05, | |
| "loss": 3.002267150878906, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.6320047189685682, | |
| "eval_loss": 2.947200059890747, | |
| "eval_runtime": 119.0412, | |
| "eval_samples_per_second": 128.888, | |
| "eval_steps_per_second": 4.032, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.6341114013651302, | |
| "grad_norm": 0.7570881843566895, | |
| "learning_rate": 3.878910103545281e-05, | |
| "loss": 2.9890081787109377, | |
| "step": 15050 | |
| }, | |
| { | |
| "epoch": 0.6362180837616921, | |
| "grad_norm": 0.7803136706352234, | |
| "learning_rate": 3.871972296349676e-05, | |
| "loss": 2.991712646484375, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 0.638324766158254, | |
| "grad_norm": 0.7539452910423279, | |
| "learning_rate": 3.8650193365597864e-05, | |
| "loss": 3.0010919189453125, | |
| "step": 15150 | |
| }, | |
| { | |
| "epoch": 0.6404314485548159, | |
| "grad_norm": 0.7635869383811951, | |
| "learning_rate": 3.8580513009667976e-05, | |
| "loss": 2.9745663452148436, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 0.6425381309513778, | |
| "grad_norm": 0.7659322023391724, | |
| "learning_rate": 3.8510682665283984e-05, | |
| "loss": 3.023065185546875, | |
| "step": 15250 | |
| }, | |
| { | |
| "epoch": 0.6446448133479397, | |
| "grad_norm": 0.7653759121894836, | |
| "learning_rate": 3.844070310367927e-05, | |
| "loss": 2.991380615234375, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 0.6467514957445015, | |
| "grad_norm": 0.7359285354614258, | |
| "learning_rate": 3.83705750977353e-05, | |
| "loss": 3.0079351806640626, | |
| "step": 15350 | |
| }, | |
| { | |
| "epoch": 0.6488581781410635, | |
| "grad_norm": 0.7472555637359619, | |
| "learning_rate": 3.830029942197297e-05, | |
| "loss": 2.9910882568359374, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 0.6509648605376254, | |
| "grad_norm": 0.7436686158180237, | |
| "learning_rate": 3.822987685254409e-05, | |
| "loss": 2.9968963623046876, | |
| "step": 15450 | |
| }, | |
| { | |
| "epoch": 0.6530715429341872, | |
| "grad_norm": 0.7502424120903015, | |
| "learning_rate": 3.8159308167222845e-05, | |
| "loss": 2.982164001464844, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.6551782253307491, | |
| "grad_norm": 0.7380735278129578, | |
| "learning_rate": 3.808859414539717e-05, | |
| "loss": 2.9850360107421876, | |
| "step": 15550 | |
| }, | |
| { | |
| "epoch": 0.6572849077273111, | |
| "grad_norm": 0.7413625121116638, | |
| "learning_rate": 3.8017735568060155e-05, | |
| "loss": 2.9962612915039064, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 0.6593915901238729, | |
| "grad_norm": 0.7653988003730774, | |
| "learning_rate": 3.7946733217801414e-05, | |
| "loss": 3.0060968017578125, | |
| "step": 15650 | |
| }, | |
| { | |
| "epoch": 0.6614982725204348, | |
| "grad_norm": 0.7521942257881165, | |
| "learning_rate": 3.787558787879846e-05, | |
| "loss": 3.0134130859375, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 0.6636049549169967, | |
| "grad_norm": 0.7369142770767212, | |
| "learning_rate": 3.780430033680801e-05, | |
| "loss": 2.991086730957031, | |
| "step": 15750 | |
| }, | |
| { | |
| "epoch": 0.6657116373135586, | |
| "grad_norm": 0.7535198330879211, | |
| "learning_rate": 3.773287137915732e-05, | |
| "loss": 2.99282470703125, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 0.6678183197101205, | |
| "grad_norm": 0.7708492875099182, | |
| "learning_rate": 3.766130179473553e-05, | |
| "loss": 3.0071200561523437, | |
| "step": 15850 | |
| }, | |
| { | |
| "epoch": 0.6699250021066824, | |
| "grad_norm": 0.7468549609184265, | |
| "learning_rate": 3.758959237398488e-05, | |
| "loss": 2.9987628173828127, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 0.6720316845032442, | |
| "grad_norm": 0.7619309425354004, | |
| "learning_rate": 3.751774390889205e-05, | |
| "loss": 3.015542297363281, | |
| "step": 15950 | |
| }, | |
| { | |
| "epoch": 0.6741383668998062, | |
| "grad_norm": 0.7726176977157593, | |
| "learning_rate": 3.7445757192979336e-05, | |
| "loss": 3.007193298339844, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.6741383668998062, | |
| "eval_loss": 2.9454808235168457, | |
| "eval_runtime": 118.9703, | |
| "eval_samples_per_second": 128.965, | |
| "eval_steps_per_second": 4.035, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.6762450492963681, | |
| "grad_norm": 0.7481387257575989, | |
| "learning_rate": 3.7373633021295975e-05, | |
| "loss": 2.9904840087890623, | |
| "step": 16050 | |
| }, | |
| { | |
| "epoch": 0.67835173169293, | |
| "grad_norm": 0.7400880455970764, | |
| "learning_rate": 3.730137219040929e-05, | |
| "loss": 2.991521301269531, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 0.6804584140894918, | |
| "grad_norm": 0.7606403827667236, | |
| "learning_rate": 3.722897549839593e-05, | |
| "loss": 2.99654541015625, | |
| "step": 16150 | |
| }, | |
| { | |
| "epoch": 0.6825650964860538, | |
| "grad_norm": 0.7475659847259521, | |
| "learning_rate": 3.715644374483305e-05, | |
| "loss": 2.98114990234375, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 0.6846717788826157, | |
| "grad_norm": 0.7718997001647949, | |
| "learning_rate": 3.708377773078947e-05, | |
| "loss": 2.9824609375, | |
| "step": 16250 | |
| }, | |
| { | |
| "epoch": 0.6867784612791775, | |
| "grad_norm": 0.7612332105636597, | |
| "learning_rate": 3.7010978258816855e-05, | |
| "loss": 2.9789178466796873, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 0.6888851436757395, | |
| "grad_norm": 0.73729407787323, | |
| "learning_rate": 3.693804613294079e-05, | |
| "loss": 2.992366943359375, | |
| "step": 16350 | |
| }, | |
| { | |
| "epoch": 0.6909918260723014, | |
| "grad_norm": 0.7615039944648743, | |
| "learning_rate": 3.6864982158651986e-05, | |
| "loss": 3.0145358276367187, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 0.6930985084688632, | |
| "grad_norm": 0.7860649824142456, | |
| "learning_rate": 3.679178714289732e-05, | |
| "loss": 2.9808953857421874, | |
| "step": 16450 | |
| }, | |
| { | |
| "epoch": 0.6952051908654251, | |
| "grad_norm": 0.7659120559692383, | |
| "learning_rate": 3.671846189407093e-05, | |
| "loss": 3.0128125, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.6973118732619871, | |
| "grad_norm": 0.7645251154899597, | |
| "learning_rate": 3.664500722200531e-05, | |
| "loss": 3.0033697509765624, | |
| "step": 16550 | |
| }, | |
| { | |
| "epoch": 0.6994185556585489, | |
| "grad_norm": 0.7724819779396057, | |
| "learning_rate": 3.657142393796236e-05, | |
| "loss": 3.01952880859375, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 0.7015252380551108, | |
| "grad_norm": 0.7801803946495056, | |
| "learning_rate": 3.649771285462441e-05, | |
| "loss": 2.9940887451171876, | |
| "step": 16650 | |
| }, | |
| { | |
| "epoch": 0.7036319204516727, | |
| "grad_norm": 0.7721547484397888, | |
| "learning_rate": 3.642387478608527e-05, | |
| "loss": 2.99277587890625, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 0.7057386028482346, | |
| "grad_norm": 0.7836876511573792, | |
| "learning_rate": 3.634991054784121e-05, | |
| "loss": 2.9915655517578124, | |
| "step": 16750 | |
| }, | |
| { | |
| "epoch": 0.7078452852447965, | |
| "grad_norm": 0.7785339951515198, | |
| "learning_rate": 3.627582095678197e-05, | |
| "loss": 2.98688232421875, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 0.7099519676413584, | |
| "grad_norm": 0.7511887550354004, | |
| "learning_rate": 3.620160683118172e-05, | |
| "loss": 2.9928399658203126, | |
| "step": 16850 | |
| }, | |
| { | |
| "epoch": 0.7120586500379202, | |
| "grad_norm": 0.7539852857589722, | |
| "learning_rate": 3.6127268990690075e-05, | |
| "loss": 2.9907171630859377, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 0.7141653324344822, | |
| "grad_norm": 0.7648764252662659, | |
| "learning_rate": 3.605280825632299e-05, | |
| "loss": 2.9792535400390623, | |
| "step": 16950 | |
| }, | |
| { | |
| "epoch": 0.7162720148310441, | |
| "grad_norm": 0.7702951431274414, | |
| "learning_rate": 3.5978225450453675e-05, | |
| "loss": 2.990997314453125, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.7162720148310441, | |
| "eval_loss": 2.943453788757324, | |
| "eval_runtime": 118.9878, | |
| "eval_samples_per_second": 128.946, | |
| "eval_steps_per_second": 4.034, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.718378697227606, | |
| "grad_norm": 0.7571023106575012, | |
| "learning_rate": 3.5903521396803586e-05, | |
| "loss": 3.00954345703125, | |
| "step": 17050 | |
| }, | |
| { | |
| "epoch": 0.7204853796241678, | |
| "grad_norm": 0.7798113226890564, | |
| "learning_rate": 3.582869692043328e-05, | |
| "loss": 2.991300048828125, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 0.7225920620207298, | |
| "grad_norm": 0.757226824760437, | |
| "learning_rate": 3.5753752847733286e-05, | |
| "loss": 2.976229248046875, | |
| "step": 17150 | |
| }, | |
| { | |
| "epoch": 0.7246987444172917, | |
| "grad_norm": 0.7685642838478088, | |
| "learning_rate": 3.567869000641503e-05, | |
| "loss": 2.970736083984375, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 0.7268054268138535, | |
| "grad_norm": 0.7552573084831238, | |
| "learning_rate": 3.5603509225501635e-05, | |
| "loss": 3.0016534423828123, | |
| "step": 17250 | |
| }, | |
| { | |
| "epoch": 0.7289121092104155, | |
| "grad_norm": 0.7572063207626343, | |
| "learning_rate": 3.552821133531883e-05, | |
| "loss": 2.974124755859375, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 0.7310187916069774, | |
| "grad_norm": 0.7647002935409546, | |
| "learning_rate": 3.545279716748572e-05, | |
| "loss": 2.9755426025390626, | |
| "step": 17350 | |
| }, | |
| { | |
| "epoch": 0.7331254740035392, | |
| "grad_norm": 0.7891484498977661, | |
| "learning_rate": 3.537726755490562e-05, | |
| "loss": 2.987391357421875, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 0.7352321564001011, | |
| "grad_norm": 0.7644563913345337, | |
| "learning_rate": 3.530162333175688e-05, | |
| "loss": 3.000482177734375, | |
| "step": 17450 | |
| }, | |
| { | |
| "epoch": 0.7373388387966631, | |
| "grad_norm": 0.7742002010345459, | |
| "learning_rate": 3.5225865333483656e-05, | |
| "loss": 2.993226318359375, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.7394455211932249, | |
| "grad_norm": 0.760245144367218, | |
| "learning_rate": 3.514999439678666e-05, | |
| "loss": 2.986939697265625, | |
| "step": 17550 | |
| }, | |
| { | |
| "epoch": 0.7415522035897868, | |
| "grad_norm": 0.774545431137085, | |
| "learning_rate": 3.5074011359613936e-05, | |
| "loss": 2.995418701171875, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 0.7436588859863487, | |
| "grad_norm": 0.7995437383651733, | |
| "learning_rate": 3.4997917061151644e-05, | |
| "loss": 2.9880685424804687, | |
| "step": 17650 | |
| }, | |
| { | |
| "epoch": 0.7457655683829106, | |
| "grad_norm": 0.7373849749565125, | |
| "learning_rate": 3.492171234181474e-05, | |
| "loss": 3.004322509765625, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 0.7478722507794725, | |
| "grad_norm": 0.7646380662918091, | |
| "learning_rate": 3.484539804323769e-05, | |
| "loss": 2.99695556640625, | |
| "step": 17750 | |
| }, | |
| { | |
| "epoch": 0.7499789331760344, | |
| "grad_norm": 0.7866349220275879, | |
| "learning_rate": 3.476897500826522e-05, | |
| "loss": 2.988021240234375, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 0.7520856155725962, | |
| "grad_norm": 0.7696439623832703, | |
| "learning_rate": 3.469244408094299e-05, | |
| "loss": 2.9982611083984376, | |
| "step": 17850 | |
| }, | |
| { | |
| "epoch": 0.7541922979691582, | |
| "grad_norm": 0.7706558108329773, | |
| "learning_rate": 3.461580610650822e-05, | |
| "loss": 2.9865411376953124, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 0.7562989803657201, | |
| "grad_norm": 0.7493193745613098, | |
| "learning_rate": 3.453906193138046e-05, | |
| "loss": 2.9728173828125, | |
| "step": 17950 | |
| }, | |
| { | |
| "epoch": 0.7584056627622819, | |
| "grad_norm": 0.7709448933601379, | |
| "learning_rate": 3.446221240315213e-05, | |
| "loss": 2.99363525390625, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.7584056627622819, | |
| "eval_loss": 2.943455457687378, | |
| "eval_runtime": 118.6655, | |
| "eval_samples_per_second": 129.296, | |
| "eval_steps_per_second": 4.045, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.7605123451588438, | |
| "grad_norm": 0.7643709182739258, | |
| "learning_rate": 3.438525837057924e-05, | |
| "loss": 2.97694580078125, | |
| "step": 18050 | |
| }, | |
| { | |
| "epoch": 0.7626190275554058, | |
| "grad_norm": 0.7416406869888306, | |
| "learning_rate": 3.430820068357199e-05, | |
| "loss": 2.9764605712890626, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 0.7647257099519676, | |
| "grad_norm": 0.7561569809913635, | |
| "learning_rate": 3.423104019318537e-05, | |
| "loss": 2.9853375244140623, | |
| "step": 18150 | |
| }, | |
| { | |
| "epoch": 0.7668323923485295, | |
| "grad_norm": 0.752541720867157, | |
| "learning_rate": 3.415377775160975e-05, | |
| "loss": 2.9886016845703125, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 0.7689390747450914, | |
| "grad_norm": 0.7720414400100708, | |
| "learning_rate": 3.407641421216152e-05, | |
| "loss": 2.973192138671875, | |
| "step": 18250 | |
| }, | |
| { | |
| "epoch": 0.7710457571416534, | |
| "grad_norm": 0.7412005662918091, | |
| "learning_rate": 3.399895042927362e-05, | |
| "loss": 2.9741815185546874, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 0.7731524395382152, | |
| "grad_norm": 0.7541579604148865, | |
| "learning_rate": 3.3921387258486125e-05, | |
| "loss": 2.984752197265625, | |
| "step": 18350 | |
| }, | |
| { | |
| "epoch": 0.7752591219347771, | |
| "grad_norm": 0.7787860631942749, | |
| "learning_rate": 3.3843725556436784e-05, | |
| "loss": 2.98060791015625, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 0.7773658043313391, | |
| "grad_norm": 0.7794627547264099, | |
| "learning_rate": 3.376596618085155e-05, | |
| "loss": 2.996499328613281, | |
| "step": 18450 | |
| }, | |
| { | |
| "epoch": 0.7794724867279009, | |
| "grad_norm": 0.7562053799629211, | |
| "learning_rate": 3.368810999053514e-05, | |
| "loss": 2.996988525390625, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 0.7815791691244628, | |
| "grad_norm": 0.7913422584533691, | |
| "learning_rate": 3.361015784536153e-05, | |
| "loss": 2.9749346923828126, | |
| "step": 18550 | |
| }, | |
| { | |
| "epoch": 0.7836858515210247, | |
| "grad_norm": 0.7462142705917358, | |
| "learning_rate": 3.353211060626444e-05, | |
| "loss": 2.949781494140625, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 0.7857925339175866, | |
| "grad_norm": 0.7861297130584717, | |
| "learning_rate": 3.345396913522786e-05, | |
| "loss": 2.981492919921875, | |
| "step": 18650 | |
| }, | |
| { | |
| "epoch": 0.7878992163141485, | |
| "grad_norm": 0.7630501985549927, | |
| "learning_rate": 3.337573429527652e-05, | |
| "loss": 2.97427978515625, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 0.7900058987107104, | |
| "grad_norm": 0.782783567905426, | |
| "learning_rate": 3.329740695046632e-05, | |
| "loss": 2.9570343017578127, | |
| "step": 18750 | |
| }, | |
| { | |
| "epoch": 0.7921125811072722, | |
| "grad_norm": 0.7517935633659363, | |
| "learning_rate": 3.321898796587488e-05, | |
| "loss": 2.9745233154296873, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 0.7942192635038342, | |
| "grad_norm": 0.7612478733062744, | |
| "learning_rate": 3.314047820759185e-05, | |
| "loss": 2.9816546630859375, | |
| "step": 18850 | |
| }, | |
| { | |
| "epoch": 0.7963259459003961, | |
| "grad_norm": 0.7646007537841797, | |
| "learning_rate": 3.306187854270948e-05, | |
| "loss": 3.006817626953125, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 0.7984326282969579, | |
| "grad_norm": 0.7654311060905457, | |
| "learning_rate": 3.298318983931295e-05, | |
| "loss": 2.9936016845703124, | |
| "step": 18950 | |
| }, | |
| { | |
| "epoch": 0.8005393106935198, | |
| "grad_norm": 0.7550582885742188, | |
| "learning_rate": 3.290441296647083e-05, | |
| "loss": 2.9723480224609373, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.8005393106935198, | |
| "eval_loss": 2.9402308464050293, | |
| "eval_runtime": 119.1534, | |
| "eval_samples_per_second": 128.767, | |
| "eval_steps_per_second": 4.028, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.8026459930900818, | |
| "grad_norm": 0.754814863204956, | |
| "learning_rate": 3.2827126927572445e-05, | |
| "loss": 2.9910968017578123, | |
| "step": 19050 | |
| }, | |
| { | |
| "epoch": 0.8047526754866436, | |
| "grad_norm": 0.7622116804122925, | |
| "learning_rate": 3.274817804695623e-05, | |
| "loss": 2.9905078125, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 0.8068593578832055, | |
| "grad_norm": 0.7846131920814514, | |
| "learning_rate": 3.266914359245583e-05, | |
| "loss": 2.9857449340820312, | |
| "step": 19150 | |
| }, | |
| { | |
| "epoch": 0.8089660402797674, | |
| "grad_norm": 0.7798764705657959, | |
| "learning_rate": 3.2590024436958436e-05, | |
| "loss": 2.9960977172851564, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 0.8110727226763293, | |
| "grad_norm": 0.7347792387008667, | |
| "learning_rate": 3.2510821454286665e-05, | |
| "loss": 2.9629473876953125, | |
| "step": 19250 | |
| }, | |
| { | |
| "epoch": 0.8131794050728912, | |
| "grad_norm": 0.7740408182144165, | |
| "learning_rate": 3.243153551918901e-05, | |
| "loss": 2.989219055175781, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 0.8152860874694531, | |
| "grad_norm": 0.7646519541740417, | |
| "learning_rate": 3.235216750733009e-05, | |
| "loss": 2.983004150390625, | |
| "step": 19350 | |
| }, | |
| { | |
| "epoch": 0.817392769866015, | |
| "grad_norm": 0.7489182353019714, | |
| "learning_rate": 3.2272718295281e-05, | |
| "loss": 2.97353515625, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 0.8194994522625769, | |
| "grad_norm": 0.7974159717559814, | |
| "learning_rate": 3.2193188760509686e-05, | |
| "loss": 2.984364013671875, | |
| "step": 19450 | |
| }, | |
| { | |
| "epoch": 0.8216061346591388, | |
| "grad_norm": 0.7829204797744751, | |
| "learning_rate": 3.211357978137117e-05, | |
| "loss": 2.973402099609375, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 0.8237128170557007, | |
| "grad_norm": 0.7689785361289978, | |
| "learning_rate": 3.2033892237097895e-05, | |
| "loss": 2.9742327880859376, | |
| "step": 19550 | |
| }, | |
| { | |
| "epoch": 0.8258194994522626, | |
| "grad_norm": 0.743333101272583, | |
| "learning_rate": 3.195572306799289e-05, | |
| "loss": 2.98500732421875, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 0.8279261818488245, | |
| "grad_norm": 0.7535820007324219, | |
| "learning_rate": 3.1875882562051194e-05, | |
| "loss": 2.96677978515625, | |
| "step": 19650 | |
| }, | |
| { | |
| "epoch": 0.8300328642453864, | |
| "grad_norm": 0.755489706993103, | |
| "learning_rate": 3.179596611619505e-05, | |
| "loss": 2.9761566162109374, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 0.8321395466419482, | |
| "grad_norm": 0.7670351266860962, | |
| "learning_rate": 3.171597461305269e-05, | |
| "loss": 2.9710345458984375, | |
| "step": 19750 | |
| }, | |
| { | |
| "epoch": 0.8342462290385102, | |
| "grad_norm": 0.7711771726608276, | |
| "learning_rate": 3.1635908936081306e-05, | |
| "loss": 2.988819885253906, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 0.8363529114350721, | |
| "grad_norm": 0.7788257002830505, | |
| "learning_rate": 3.155576996955729e-05, | |
| "loss": 2.9836825561523437, | |
| "step": 19850 | |
| }, | |
| { | |
| "epoch": 0.8384595938316339, | |
| "grad_norm": 0.8039634823799133, | |
| "learning_rate": 3.147555859856648e-05, | |
| "loss": 2.9674588012695313, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 0.8405662762281958, | |
| "grad_norm": 0.7871184349060059, | |
| "learning_rate": 3.139527570899438e-05, | |
| "loss": 2.993527526855469, | |
| "step": 19950 | |
| }, | |
| { | |
| "epoch": 0.8426729586247578, | |
| "grad_norm": 0.7455061078071594, | |
| "learning_rate": 3.131492218751635e-05, | |
| "loss": 2.9716851806640623, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.8426729586247578, | |
| "eval_loss": 2.938898801803589, | |
| "eval_runtime": 118.9917, | |
| "eval_samples_per_second": 128.942, | |
| "eval_steps_per_second": 4.034, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.8447796410213196, | |
| "grad_norm": 0.7597523927688599, | |
| "learning_rate": 3.123449892158788e-05, | |
| "loss": 2.984842529296875, | |
| "step": 20050 | |
| }, | |
| { | |
| "epoch": 0.8468863234178815, | |
| "grad_norm": 0.7505768537521362, | |
| "learning_rate": 3.115400679943469e-05, | |
| "loss": 2.9934478759765626, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 0.8489930058144434, | |
| "grad_norm": 0.7709312438964844, | |
| "learning_rate": 3.107344671004303e-05, | |
| "loss": 2.96940673828125, | |
| "step": 20150 | |
| }, | |
| { | |
| "epoch": 0.8510996882110053, | |
| "grad_norm": 0.7617827653884888, | |
| "learning_rate": 3.0992819543149765e-05, | |
| "loss": 2.983475341796875, | |
| "step": 20200 | |
| }, | |
| { | |
| "epoch": 0.8532063706075672, | |
| "grad_norm": 0.7898280024528503, | |
| "learning_rate": 3.091212618923261e-05, | |
| "loss": 2.9768841552734373, | |
| "step": 20250 | |
| }, | |
| { | |
| "epoch": 0.8553130530041291, | |
| "grad_norm": 0.7669104337692261, | |
| "learning_rate": 3.083136753950026e-05, | |
| "loss": 2.9685321044921875, | |
| "step": 20300 | |
| }, | |
| { | |
| "epoch": 0.8574197354006909, | |
| "grad_norm": 0.7699618339538574, | |
| "learning_rate": 3.0750544485882585e-05, | |
| "loss": 2.9848779296875, | |
| "step": 20350 | |
| }, | |
| { | |
| "epoch": 0.8595264177972529, | |
| "grad_norm": 0.7880122661590576, | |
| "learning_rate": 3.066965792102073e-05, | |
| "loss": 2.985040283203125, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 0.8616331001938148, | |
| "grad_norm": 0.760508120059967, | |
| "learning_rate": 3.0588708738257324e-05, | |
| "loss": 2.9629864501953125, | |
| "step": 20450 | |
| }, | |
| { | |
| "epoch": 0.8637397825903766, | |
| "grad_norm": 0.7696555852890015, | |
| "learning_rate": 3.0507697831626507e-05, | |
| "loss": 2.988545227050781, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 0.8658464649869386, | |
| "grad_norm": 0.7843043804168701, | |
| "learning_rate": 3.042662609584418e-05, | |
| "loss": 2.995951843261719, | |
| "step": 20550 | |
| }, | |
| { | |
| "epoch": 0.8679531473835005, | |
| "grad_norm": 0.7505366206169128, | |
| "learning_rate": 3.0345494426298045e-05, | |
| "loss": 2.9777276611328123, | |
| "step": 20600 | |
| }, | |
| { | |
| "epoch": 0.8700598297800624, | |
| "grad_norm": 0.7813105583190918, | |
| "learning_rate": 3.0264303719037733e-05, | |
| "loss": 2.98427734375, | |
| "step": 20650 | |
| }, | |
| { | |
| "epoch": 0.8721665121766242, | |
| "grad_norm": 0.7776883244514465, | |
| "learning_rate": 3.018305487076491e-05, | |
| "loss": 2.9737237548828124, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 0.8742731945731862, | |
| "grad_norm": 0.7753885388374329, | |
| "learning_rate": 3.0101748778823374e-05, | |
| "loss": 3.0072528076171876, | |
| "step": 20750 | |
| }, | |
| { | |
| "epoch": 0.8763798769697481, | |
| "grad_norm": 0.7814272046089172, | |
| "learning_rate": 3.0020386341189154e-05, | |
| "loss": 2.9858108520507813, | |
| "step": 20800 | |
| }, | |
| { | |
| "epoch": 0.8784865593663099, | |
| "grad_norm": 0.7478072047233582, | |
| "learning_rate": 2.993896845646057e-05, | |
| "loss": 3.0002532958984376, | |
| "step": 20850 | |
| }, | |
| { | |
| "epoch": 0.8805932417628718, | |
| "grad_norm": 0.7745850086212158, | |
| "learning_rate": 2.9857496023848323e-05, | |
| "loss": 2.971799621582031, | |
| "step": 20900 | |
| }, | |
| { | |
| "epoch": 0.8826999241594338, | |
| "grad_norm": 0.7583919763565063, | |
| "learning_rate": 2.9775969943165572e-05, | |
| "loss": 2.975789794921875, | |
| "step": 20950 | |
| }, | |
| { | |
| "epoch": 0.8848066065559956, | |
| "grad_norm": 0.757747232913971, | |
| "learning_rate": 2.9694391114817966e-05, | |
| "loss": 2.957315673828125, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.8848066065559956, | |
| "eval_loss": 2.9376776218414307, | |
| "eval_runtime": 119.1894, | |
| "eval_samples_per_second": 128.728, | |
| "eval_steps_per_second": 4.027, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.8869132889525575, | |
| "grad_norm": 0.7598838210105896, | |
| "learning_rate": 2.9612760439793753e-05, | |
| "loss": 2.96072998046875, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 0.8890199713491194, | |
| "grad_norm": 0.7845181822776794, | |
| "learning_rate": 2.953107881965376e-05, | |
| "loss": 2.9771969604492186, | |
| "step": 21100 | |
| }, | |
| { | |
| "epoch": 0.8911266537456813, | |
| "grad_norm": 0.7929947376251221, | |
| "learning_rate": 2.9449347156521502e-05, | |
| "loss": 2.9737921142578125, | |
| "step": 21150 | |
| }, | |
| { | |
| "epoch": 0.8932333361422432, | |
| "grad_norm": 0.7893011569976807, | |
| "learning_rate": 2.9367566353073155e-05, | |
| "loss": 2.9805047607421873, | |
| "step": 21200 | |
| }, | |
| { | |
| "epoch": 0.8953400185388051, | |
| "grad_norm": 0.7605388760566711, | |
| "learning_rate": 2.928573731252766e-05, | |
| "loss": 2.981062927246094, | |
| "step": 21250 | |
| }, | |
| { | |
| "epoch": 0.8974467009353669, | |
| "grad_norm": 0.7496638298034668, | |
| "learning_rate": 2.920386093863666e-05, | |
| "loss": 2.977832946777344, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 0.8995533833319289, | |
| "grad_norm": 0.7931700348854065, | |
| "learning_rate": 2.9121938135674614e-05, | |
| "loss": 2.9760317993164063, | |
| "step": 21350 | |
| }, | |
| { | |
| "epoch": 0.9016600657284908, | |
| "grad_norm": 0.7766400575637817, | |
| "learning_rate": 2.903996980842873e-05, | |
| "loss": 2.959849548339844, | |
| "step": 21400 | |
| }, | |
| { | |
| "epoch": 0.9037667481250526, | |
| "grad_norm": 0.7383178472518921, | |
| "learning_rate": 2.8957956862189012e-05, | |
| "loss": 2.972950134277344, | |
| "step": 21450 | |
| }, | |
| { | |
| "epoch": 0.9058734305216145, | |
| "grad_norm": 0.7554975748062134, | |
| "learning_rate": 2.8875900202738253e-05, | |
| "loss": 2.9869000244140627, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 0.9079801129181765, | |
| "grad_norm": 0.7777841687202454, | |
| "learning_rate": 2.8793800736342037e-05, | |
| "loss": 2.9726498413085936, | |
| "step": 21550 | |
| }, | |
| { | |
| "epoch": 0.9100867953147384, | |
| "grad_norm": 0.7751466035842896, | |
| "learning_rate": 2.871165936973872e-05, | |
| "loss": 2.9812335205078124, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 0.9121934777113002, | |
| "grad_norm": 0.7738240957260132, | |
| "learning_rate": 2.862947701012943e-05, | |
| "loss": 2.9768447875976562, | |
| "step": 21650 | |
| }, | |
| { | |
| "epoch": 0.9143001601078622, | |
| "grad_norm": 0.7454856038093567, | |
| "learning_rate": 2.8547254565168007e-05, | |
| "loss": 2.9890667724609377, | |
| "step": 21700 | |
| }, | |
| { | |
| "epoch": 0.9164068425044241, | |
| "grad_norm": 0.7908093333244324, | |
| "learning_rate": 2.846499294295104e-05, | |
| "loss": 2.965343017578125, | |
| "step": 21750 | |
| }, | |
| { | |
| "epoch": 0.9185135249009859, | |
| "grad_norm": 0.7969614863395691, | |
| "learning_rate": 2.838269305200781e-05, | |
| "loss": 2.983328857421875, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 0.9206202072975478, | |
| "grad_norm": 0.7719734311103821, | |
| "learning_rate": 2.8300355801290218e-05, | |
| "loss": 2.9759671020507814, | |
| "step": 21850 | |
| }, | |
| { | |
| "epoch": 0.9227268896941098, | |
| "grad_norm": 0.770438015460968, | |
| "learning_rate": 2.8217982100162822e-05, | |
| "loss": 2.9574295043945313, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 0.9248335720906716, | |
| "grad_norm": 0.7770126461982727, | |
| "learning_rate": 2.8135572858392722e-05, | |
| "loss": 2.9908953857421876, | |
| "step": 21950 | |
| }, | |
| { | |
| "epoch": 0.9269402544872335, | |
| "grad_norm": 0.7843868732452393, | |
| "learning_rate": 2.805312898613956e-05, | |
| "loss": 2.98980224609375, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.9269402544872335, | |
| "eval_loss": 2.9365010261535645, | |
| "eval_runtime": 119.2163, | |
| "eval_samples_per_second": 128.699, | |
| "eval_steps_per_second": 4.026, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.9290469368837954, | |
| "grad_norm": 0.771686315536499, | |
| "learning_rate": 2.797065139394544e-05, | |
| "loss": 2.9799072265625, | |
| "step": 22050 | |
| }, | |
| { | |
| "epoch": 0.9311536192803573, | |
| "grad_norm": 0.7738041877746582, | |
| "learning_rate": 2.78881409927249e-05, | |
| "loss": 2.9672747802734376, | |
| "step": 22100 | |
| }, | |
| { | |
| "epoch": 0.9332603016769192, | |
| "grad_norm": 0.7904272079467773, | |
| "learning_rate": 2.7805598693754798e-05, | |
| "loss": 2.9740606689453126, | |
| "step": 22150 | |
| }, | |
| { | |
| "epoch": 0.9353669840734811, | |
| "grad_norm": 0.7471991181373596, | |
| "learning_rate": 2.7723025408664317e-05, | |
| "loss": 2.9706402587890626, | |
| "step": 22200 | |
| }, | |
| { | |
| "epoch": 0.9374736664700429, | |
| "grad_norm": 0.7692591547966003, | |
| "learning_rate": 2.764042204942485e-05, | |
| "loss": 2.969313659667969, | |
| "step": 22250 | |
| }, | |
| { | |
| "epoch": 0.9395803488666049, | |
| "grad_norm": 0.7923398613929749, | |
| "learning_rate": 2.755778952833995e-05, | |
| "loss": 2.9702215576171875, | |
| "step": 22300 | |
| }, | |
| { | |
| "epoch": 0.9416870312631668, | |
| "grad_norm": 0.7700166702270508, | |
| "learning_rate": 2.7475128758035223e-05, | |
| "loss": 2.9864556884765623, | |
| "step": 22350 | |
| }, | |
| { | |
| "epoch": 0.9437937136597286, | |
| "grad_norm": 0.7593397498130798, | |
| "learning_rate": 2.7392440651448288e-05, | |
| "loss": 2.9608709716796877, | |
| "step": 22400 | |
| }, | |
| { | |
| "epoch": 0.9459003960562905, | |
| "grad_norm": 0.788577139377594, | |
| "learning_rate": 2.7309726121818674e-05, | |
| "loss": 2.9901824951171876, | |
| "step": 22450 | |
| }, | |
| { | |
| "epoch": 0.9480070784528525, | |
| "grad_norm": 0.7750556468963623, | |
| "learning_rate": 2.7226986082677734e-05, | |
| "loss": 2.9944754028320313, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.9501137608494143, | |
| "grad_norm": 0.7848498225212097, | |
| "learning_rate": 2.7144221447838565e-05, | |
| "loss": 2.9570849609375, | |
| "step": 22550 | |
| }, | |
| { | |
| "epoch": 0.9522204432459762, | |
| "grad_norm": 0.7730944752693176, | |
| "learning_rate": 2.7061433131385888e-05, | |
| "loss": 2.976864013671875, | |
| "step": 22600 | |
| }, | |
| { | |
| "epoch": 0.9543271256425381, | |
| "grad_norm": 0.7930340766906738, | |
| "learning_rate": 2.6978622047665996e-05, | |
| "loss": 2.96012451171875, | |
| "step": 22650 | |
| }, | |
| { | |
| "epoch": 0.9564338080391, | |
| "grad_norm": 0.788257896900177, | |
| "learning_rate": 2.6895789111276627e-05, | |
| "loss": 2.957564697265625, | |
| "step": 22700 | |
| }, | |
| { | |
| "epoch": 0.9585404904356619, | |
| "grad_norm": 0.7802656292915344, | |
| "learning_rate": 2.6812935237056852e-05, | |
| "loss": 2.955384521484375, | |
| "step": 22750 | |
| }, | |
| { | |
| "epoch": 0.9606471728322238, | |
| "grad_norm": 0.7437043190002441, | |
| "learning_rate": 2.6730061340077013e-05, | |
| "loss": 2.973957824707031, | |
| "step": 22800 | |
| }, | |
| { | |
| "epoch": 0.9627538552287858, | |
| "grad_norm": 0.7454256415367126, | |
| "learning_rate": 2.6647168335628562e-05, | |
| "loss": 2.960787048339844, | |
| "step": 22850 | |
| }, | |
| { | |
| "epoch": 0.9648605376253476, | |
| "grad_norm": 0.7836768627166748, | |
| "learning_rate": 2.656425713921402e-05, | |
| "loss": 2.96308349609375, | |
| "step": 22900 | |
| }, | |
| { | |
| "epoch": 0.9669672200219095, | |
| "grad_norm": 0.7823575139045715, | |
| "learning_rate": 2.6481328666536792e-05, | |
| "loss": 3.0113314819335937, | |
| "step": 22950 | |
| }, | |
| { | |
| "epoch": 0.9690739024184714, | |
| "grad_norm": 0.8028575778007507, | |
| "learning_rate": 2.6398383833491104e-05, | |
| "loss": 2.96595458984375, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.9690739024184714, | |
| "eval_loss": 2.933638334274292, | |
| "eval_runtime": 119.4147, | |
| "eval_samples_per_second": 128.485, | |
| "eval_steps_per_second": 4.02, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.9711805848150333, | |
| "grad_norm": 0.7629191875457764, | |
| "learning_rate": 2.6315423556151864e-05, | |
| "loss": 2.973958740234375, | |
| "step": 23050 | |
| }, | |
| { | |
| "epoch": 0.9732872672115952, | |
| "grad_norm": 0.7664568424224854, | |
| "learning_rate": 2.623244875076457e-05, | |
| "loss": 2.965931396484375, | |
| "step": 23100 | |
| }, | |
| { | |
| "epoch": 0.9753939496081571, | |
| "grad_norm": 0.7835934162139893, | |
| "learning_rate": 2.6149460333735154e-05, | |
| "loss": 2.9625234985351563, | |
| "step": 23150 | |
| }, | |
| { | |
| "epoch": 0.9775006320047189, | |
| "grad_norm": 0.7699547410011292, | |
| "learning_rate": 2.606645922161989e-05, | |
| "loss": 2.963145446777344, | |
| "step": 23200 | |
| }, | |
| { | |
| "epoch": 0.9796073144012809, | |
| "grad_norm": 0.7824780941009521, | |
| "learning_rate": 2.5983446331115257e-05, | |
| "loss": 2.9785943603515626, | |
| "step": 23250 | |
| }, | |
| { | |
| "epoch": 0.9817139967978428, | |
| "grad_norm": 0.7860645055770874, | |
| "learning_rate": 2.5900422579047824e-05, | |
| "loss": 2.9981890869140626, | |
| "step": 23300 | |
| }, | |
| { | |
| "epoch": 0.9838206791944046, | |
| "grad_norm": 0.7866495251655579, | |
| "learning_rate": 2.581738888236412e-05, | |
| "loss": 2.9776904296875, | |
| "step": 23350 | |
| }, | |
| { | |
| "epoch": 0.9859273615909665, | |
| "grad_norm": 0.7753951549530029, | |
| "learning_rate": 2.5734346158120497e-05, | |
| "loss": 2.9835089111328124, | |
| "step": 23400 | |
| }, | |
| { | |
| "epoch": 0.9880340439875285, | |
| "grad_norm": 0.7717190384864807, | |
| "learning_rate": 2.5651295323473025e-05, | |
| "loss": 2.954573974609375, | |
| "step": 23450 | |
| }, | |
| { | |
| "epoch": 0.9901407263840903, | |
| "grad_norm": 0.825840950012207, | |
| "learning_rate": 2.5568237295667335e-05, | |
| "loss": 2.984539794921875, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 0.9922474087806522, | |
| "grad_norm": 0.7789013981819153, | |
| "learning_rate": 2.548517299202851e-05, | |
| "loss": 2.97105224609375, | |
| "step": 23550 | |
| }, | |
| { | |
| "epoch": 0.9943540911772141, | |
| "grad_norm": 0.7816980481147766, | |
| "learning_rate": 2.5402103329950944e-05, | |
| "loss": 2.9745220947265625, | |
| "step": 23600 | |
| }, | |
| { | |
| "epoch": 0.996460773573776, | |
| "grad_norm": 0.7796857953071594, | |
| "learning_rate": 2.5320690746536783e-05, | |
| "loss": 2.957999267578125, | |
| "step": 23650 | |
| }, | |
| { | |
| "epoch": 0.9985674559703379, | |
| "grad_norm": 0.7736715078353882, | |
| "learning_rate": 2.5237613181469354e-05, | |
| "loss": 2.9646710205078124, | |
| "step": 23700 | |
| }, | |
| { | |
| "epoch": 1.0006741383668998, | |
| "grad_norm": 0.7790223956108093, | |
| "learning_rate": 2.5154532992109774e-05, | |
| "loss": 2.9590081787109375, | |
| "step": 23750 | |
| }, | |
| { | |
| "epoch": 1.0027808207634616, | |
| "grad_norm": 0.7563338279724121, | |
| "learning_rate": 2.5071451096027883e-05, | |
| "loss": 2.96490234375, | |
| "step": 23800 | |
| }, | |
| { | |
| "epoch": 1.0048875031600235, | |
| "grad_norm": 0.7708901166915894, | |
| "learning_rate": 2.4988368410812367e-05, | |
| "loss": 2.9566876220703127, | |
| "step": 23850 | |
| }, | |
| { | |
| "epoch": 1.0069941855565856, | |
| "grad_norm": 0.779563844203949, | |
| "learning_rate": 2.4905285854060623e-05, | |
| "loss": 2.967255859375, | |
| "step": 23900 | |
| }, | |
| { | |
| "epoch": 1.0091008679531475, | |
| "grad_norm": 0.7678179740905762, | |
| "learning_rate": 2.4822204343368666e-05, | |
| "loss": 2.9493267822265623, | |
| "step": 23950 | |
| }, | |
| { | |
| "epoch": 1.0112075503497093, | |
| "grad_norm": 0.7816394567489624, | |
| "learning_rate": 2.47391247963209e-05, | |
| "loss": 2.951517333984375, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.0112075503497093, | |
| "eval_loss": 2.931684970855713, | |
| "eval_runtime": 119.3409, | |
| "eval_samples_per_second": 128.564, | |
| "eval_steps_per_second": 4.022, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.0133142327462712, | |
| "grad_norm": 0.7653071880340576, | |
| "learning_rate": 2.46560481304801e-05, | |
| "loss": 2.9472750854492187, | |
| "step": 24050 | |
| }, | |
| { | |
| "epoch": 1.015420915142833, | |
| "grad_norm": 0.7852833867073059, | |
| "learning_rate": 2.4572975263377175e-05, | |
| "loss": 2.9459237670898437, | |
| "step": 24100 | |
| }, | |
| { | |
| "epoch": 1.017527597539395, | |
| "grad_norm": 0.8014306426048279, | |
| "learning_rate": 2.4489907112501097e-05, | |
| "loss": 2.950411071777344, | |
| "step": 24150 | |
| }, | |
| { | |
| "epoch": 1.0196342799359568, | |
| "grad_norm": 0.7756820917129517, | |
| "learning_rate": 2.440850578448941e-05, | |
| "loss": 2.969849853515625, | |
| "step": 24200 | |
| }, | |
| { | |
| "epoch": 1.0217409623325187, | |
| "grad_norm": 0.7699014544487, | |
| "learning_rate": 2.4325449678304885e-05, | |
| "loss": 2.9526519775390625, | |
| "step": 24250 | |
| }, | |
| { | |
| "epoch": 1.0238476447290807, | |
| "grad_norm": 0.788263738155365, | |
| "learning_rate": 2.424240102211579e-05, | |
| "loss": 2.9390045166015626, | |
| "step": 24300 | |
| }, | |
| { | |
| "epoch": 1.0259543271256426, | |
| "grad_norm": 0.782202959060669, | |
| "learning_rate": 2.415936073314372e-05, | |
| "loss": 2.9497976684570313, | |
| "step": 24350 | |
| }, | |
| { | |
| "epoch": 1.0280610095222045, | |
| "grad_norm": 0.7703037261962891, | |
| "learning_rate": 2.4076329728517826e-05, | |
| "loss": 2.9414105224609375, | |
| "step": 24400 | |
| }, | |
| { | |
| "epoch": 1.0301676919187663, | |
| "grad_norm": 0.7838400602340698, | |
| "learning_rate": 2.3993308925264746e-05, | |
| "loss": 2.939110107421875, | |
| "step": 24450 | |
| }, | |
| { | |
| "epoch": 1.0322743743153282, | |
| "grad_norm": 0.795155942440033, | |
| "learning_rate": 2.391029924029842e-05, | |
| "loss": 2.9515594482421874, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 1.03438105671189, | |
| "grad_norm": 0.7806478142738342, | |
| "learning_rate": 2.3827301590410032e-05, | |
| "loss": 2.9430780029296875, | |
| "step": 24550 | |
| }, | |
| { | |
| "epoch": 1.036487739108452, | |
| "grad_norm": 0.7851158380508423, | |
| "learning_rate": 2.3744316892257814e-05, | |
| "loss": 2.9436154174804687, | |
| "step": 24600 | |
| }, | |
| { | |
| "epoch": 1.0385944215050138, | |
| "grad_norm": 0.7663733959197998, | |
| "learning_rate": 2.3661346062356965e-05, | |
| "loss": 2.9495138549804687, | |
| "step": 24650 | |
| }, | |
| { | |
| "epoch": 1.0407011039015759, | |
| "grad_norm": 0.7458925247192383, | |
| "learning_rate": 2.357839001706953e-05, | |
| "loss": 2.954429931640625, | |
| "step": 24700 | |
| }, | |
| { | |
| "epoch": 1.0428077862981378, | |
| "grad_norm": 0.7798863649368286, | |
| "learning_rate": 2.3495449672594243e-05, | |
| "loss": 2.94943603515625, | |
| "step": 24750 | |
| }, | |
| { | |
| "epoch": 1.0449144686946996, | |
| "grad_norm": 0.7786155939102173, | |
| "learning_rate": 2.341252594495645e-05, | |
| "loss": 2.934886474609375, | |
| "step": 24800 | |
| }, | |
| { | |
| "epoch": 1.0470211510912615, | |
| "grad_norm": 0.7668699026107788, | |
| "learning_rate": 2.3329619749997954e-05, | |
| "loss": 2.9515509033203124, | |
| "step": 24850 | |
| }, | |
| { | |
| "epoch": 1.0491278334878233, | |
| "grad_norm": 0.793624997138977, | |
| "learning_rate": 2.3246732003366962e-05, | |
| "loss": 2.9429592895507812, | |
| "step": 24900 | |
| }, | |
| { | |
| "epoch": 1.0512345158843852, | |
| "grad_norm": 0.78849196434021, | |
| "learning_rate": 2.3163863620507886e-05, | |
| "loss": 2.961234130859375, | |
| "step": 24950 | |
| }, | |
| { | |
| "epoch": 1.053341198280947, | |
| "grad_norm": 0.8046838045120239, | |
| "learning_rate": 2.30810155166513e-05, | |
| "loss": 2.965606994628906, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.053341198280947, | |
| "eval_loss": 2.930619239807129, | |
| "eval_runtime": 119.2511, | |
| "eval_samples_per_second": 128.661, | |
| "eval_steps_per_second": 4.025, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.0554478806775092, | |
| "grad_norm": 0.7959691286087036, | |
| "learning_rate": 2.299818860680379e-05, | |
| "loss": 2.95442626953125, | |
| "step": 25050 | |
| }, | |
| { | |
| "epoch": 1.057554563074071, | |
| "grad_norm": 0.79048752784729, | |
| "learning_rate": 2.2915383805737886e-05, | |
| "loss": 2.945472412109375, | |
| "step": 25100 | |
| }, | |
| { | |
| "epoch": 1.059661245470633, | |
| "grad_norm": 0.802245020866394, | |
| "learning_rate": 2.2832602027981955e-05, | |
| "loss": 2.949869384765625, | |
| "step": 25150 | |
| }, | |
| { | |
| "epoch": 1.0617679278671948, | |
| "grad_norm": 0.7903581857681274, | |
| "learning_rate": 2.2749844187810038e-05, | |
| "loss": 2.9403094482421874, | |
| "step": 25200 | |
| }, | |
| { | |
| "epoch": 1.0638746102637566, | |
| "grad_norm": 0.7770790457725525, | |
| "learning_rate": 2.2667111199231848e-05, | |
| "loss": 2.9359564208984374, | |
| "step": 25250 | |
| }, | |
| { | |
| "epoch": 1.0659812926603185, | |
| "grad_norm": 0.7775770425796509, | |
| "learning_rate": 2.2584403975982582e-05, | |
| "loss": 2.951751403808594, | |
| "step": 25300 | |
| }, | |
| { | |
| "epoch": 1.0680879750568804, | |
| "grad_norm": 0.8019676208496094, | |
| "learning_rate": 2.2501723431512923e-05, | |
| "loss": 2.928594665527344, | |
| "step": 25350 | |
| }, | |
| { | |
| "epoch": 1.0701946574534422, | |
| "grad_norm": 0.790306568145752, | |
| "learning_rate": 2.2419070478978868e-05, | |
| "loss": 2.9543621826171873, | |
| "step": 25400 | |
| }, | |
| { | |
| "epoch": 1.0723013398500043, | |
| "grad_norm": 0.7839942574501038, | |
| "learning_rate": 2.2336446031231695e-05, | |
| "loss": 2.9492529296875, | |
| "step": 25450 | |
| }, | |
| { | |
| "epoch": 1.0744080222465662, | |
| "grad_norm": 0.790597677230835, | |
| "learning_rate": 2.225385100080784e-05, | |
| "loss": 2.9704766845703126, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 1.076514704643128, | |
| "grad_norm": 0.7874144911766052, | |
| "learning_rate": 2.2171286299918884e-05, | |
| "loss": 2.9508425903320314, | |
| "step": 25550 | |
| }, | |
| { | |
| "epoch": 1.07862138703969, | |
| "grad_norm": 0.7846702337265015, | |
| "learning_rate": 2.2088752840441408e-05, | |
| "loss": 2.939859619140625, | |
| "step": 25600 | |
| }, | |
| { | |
| "epoch": 1.0807280694362518, | |
| "grad_norm": 0.7921671271324158, | |
| "learning_rate": 2.200625153390694e-05, | |
| "loss": 2.940799560546875, | |
| "step": 25650 | |
| }, | |
| { | |
| "epoch": 1.0828347518328136, | |
| "grad_norm": 0.792601466178894, | |
| "learning_rate": 2.1923783291491944e-05, | |
| "loss": 2.9628338623046875, | |
| "step": 25700 | |
| }, | |
| { | |
| "epoch": 1.0849414342293755, | |
| "grad_norm": 0.7859891057014465, | |
| "learning_rate": 2.1841349024007652e-05, | |
| "loss": 2.942427062988281, | |
| "step": 25750 | |
| }, | |
| { | |
| "epoch": 1.0870481166259376, | |
| "grad_norm": 0.7772258520126343, | |
| "learning_rate": 2.175894964189012e-05, | |
| "loss": 2.9303094482421876, | |
| "step": 25800 | |
| }, | |
| { | |
| "epoch": 1.0891547990224995, | |
| "grad_norm": 0.7993125915527344, | |
| "learning_rate": 2.167658605519007e-05, | |
| "loss": 2.9263409423828124, | |
| "step": 25850 | |
| }, | |
| { | |
| "epoch": 1.0912614814190613, | |
| "grad_norm": 0.7829107046127319, | |
| "learning_rate": 2.159425917356292e-05, | |
| "loss": 2.9401812744140625, | |
| "step": 25900 | |
| }, | |
| { | |
| "epoch": 1.0933681638156232, | |
| "grad_norm": 0.79358971118927, | |
| "learning_rate": 2.1511969906258675e-05, | |
| "loss": 2.951754455566406, | |
| "step": 25950 | |
| }, | |
| { | |
| "epoch": 1.095474846212185, | |
| "grad_norm": 0.7937335968017578, | |
| "learning_rate": 2.142971916211194e-05, | |
| "loss": 2.9429083251953125, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 1.095474846212185, | |
| "eval_loss": 2.929333209991455, | |
| "eval_runtime": 119.3254, | |
| "eval_samples_per_second": 128.581, | |
| "eval_steps_per_second": 4.023, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 1.097581528608747, | |
| "grad_norm": 0.7786532640457153, | |
| "learning_rate": 2.1347507849531834e-05, | |
| "loss": 2.9413037109375, | |
| "step": 26050 | |
| }, | |
| { | |
| "epoch": 1.0996882110053088, | |
| "grad_norm": 0.7673649191856384, | |
| "learning_rate": 2.126533687649198e-05, | |
| "loss": 2.9261761474609376, | |
| "step": 26100 | |
| }, | |
| { | |
| "epoch": 1.1017948934018706, | |
| "grad_norm": 0.785383403301239, | |
| "learning_rate": 2.1183207150520502e-05, | |
| "loss": 2.941805114746094, | |
| "step": 26150 | |
| }, | |
| { | |
| "epoch": 1.1039015757984325, | |
| "grad_norm": 0.786217451095581, | |
| "learning_rate": 2.1101119578689937e-05, | |
| "loss": 2.9692947387695314, | |
| "step": 26200 | |
| }, | |
| { | |
| "epoch": 1.1060082581949946, | |
| "grad_norm": 0.7452182173728943, | |
| "learning_rate": 2.1019075067607285e-05, | |
| "loss": 2.9329632568359374, | |
| "step": 26250 | |
| }, | |
| { | |
| "epoch": 1.1081149405915565, | |
| "grad_norm": 0.7984647154808044, | |
| "learning_rate": 2.0937074523403937e-05, | |
| "loss": 2.9429800415039065, | |
| "step": 26300 | |
| }, | |
| { | |
| "epoch": 1.1102216229881183, | |
| "grad_norm": 0.7744592428207397, | |
| "learning_rate": 2.085511885172574e-05, | |
| "loss": 2.94785400390625, | |
| "step": 26350 | |
| }, | |
| { | |
| "epoch": 1.1123283053846802, | |
| "grad_norm": 0.7572144865989685, | |
| "learning_rate": 2.07732089577229e-05, | |
| "loss": 2.9520443725585936, | |
| "step": 26400 | |
| }, | |
| { | |
| "epoch": 1.114434987781242, | |
| "grad_norm": 0.7689366936683655, | |
| "learning_rate": 2.0691345746040077e-05, | |
| "loss": 2.9589193725585936, | |
| "step": 26450 | |
| }, | |
| { | |
| "epoch": 1.116541670177804, | |
| "grad_norm": 0.780914306640625, | |
| "learning_rate": 2.060953012080634e-05, | |
| "loss": 2.940218505859375, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 1.1186483525743658, | |
| "grad_norm": 0.787312924861908, | |
| "learning_rate": 2.0527762985625186e-05, | |
| "loss": 2.941663818359375, | |
| "step": 26550 | |
| }, | |
| { | |
| "epoch": 1.1207550349709279, | |
| "grad_norm": 0.7888234853744507, | |
| "learning_rate": 2.044604524356459e-05, | |
| "loss": 2.9333465576171873, | |
| "step": 26600 | |
| }, | |
| { | |
| "epoch": 1.1228617173674897, | |
| "grad_norm": 0.7970293760299683, | |
| "learning_rate": 2.036437779714698e-05, | |
| "loss": 2.947113037109375, | |
| "step": 26650 | |
| }, | |
| { | |
| "epoch": 1.1249683997640516, | |
| "grad_norm": 0.8074147701263428, | |
| "learning_rate": 2.0282761548339333e-05, | |
| "loss": 2.9566986083984377, | |
| "step": 26700 | |
| }, | |
| { | |
| "epoch": 1.1270750821606135, | |
| "grad_norm": 0.7938132882118225, | |
| "learning_rate": 2.020119739854314e-05, | |
| "loss": 2.9510516357421874, | |
| "step": 26750 | |
| }, | |
| { | |
| "epoch": 1.1291817645571753, | |
| "grad_norm": 0.7867845296859741, | |
| "learning_rate": 2.0119686248584544e-05, | |
| "loss": 2.926278076171875, | |
| "step": 26800 | |
| }, | |
| { | |
| "epoch": 1.1312884469537372, | |
| "grad_norm": 0.7708490490913391, | |
| "learning_rate": 2.0038228998704284e-05, | |
| "loss": 2.9552883911132812, | |
| "step": 26850 | |
| }, | |
| { | |
| "epoch": 1.133395129350299, | |
| "grad_norm": 0.7819187045097351, | |
| "learning_rate": 1.9956826548547826e-05, | |
| "loss": 2.9469345092773436, | |
| "step": 26900 | |
| }, | |
| { | |
| "epoch": 1.135501811746861, | |
| "grad_norm": 0.7773310542106628, | |
| "learning_rate": 1.987547979715543e-05, | |
| "loss": 2.954520263671875, | |
| "step": 26950 | |
| }, | |
| { | |
| "epoch": 1.137608494143423, | |
| "grad_norm": 0.7851362228393555, | |
| "learning_rate": 1.979418964295216e-05, | |
| "loss": 2.93560791015625, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 1.137608494143423, | |
| "eval_loss": 2.927694320678711, | |
| "eval_runtime": 119.2829, | |
| "eval_samples_per_second": 128.627, | |
| "eval_steps_per_second": 4.024, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 1.1397151765399849, | |
| "grad_norm": 0.7849646806716919, | |
| "learning_rate": 1.9712956983738023e-05, | |
| "loss": 2.932006530761719, | |
| "step": 27050 | |
| }, | |
| { | |
| "epoch": 1.1418218589365468, | |
| "grad_norm": 0.7975698709487915, | |
| "learning_rate": 1.9631782716678004e-05, | |
| "loss": 2.9447900390625, | |
| "step": 27100 | |
| }, | |
| { | |
| "epoch": 1.1439285413331086, | |
| "grad_norm": 0.7575515508651733, | |
| "learning_rate": 1.9550667738292225e-05, | |
| "loss": 2.9597747802734373, | |
| "step": 27150 | |
| }, | |
| { | |
| "epoch": 1.1460352237296705, | |
| "grad_norm": 0.8183218836784363, | |
| "learning_rate": 1.9469612944445948e-05, | |
| "loss": 2.9584356689453126, | |
| "step": 27200 | |
| }, | |
| { | |
| "epoch": 1.1481419061262323, | |
| "grad_norm": 0.7996168732643127, | |
| "learning_rate": 1.9388619230339793e-05, | |
| "loss": 2.9178704833984375, | |
| "step": 27250 | |
| }, | |
| { | |
| "epoch": 1.1502485885227942, | |
| "grad_norm": 0.7881982326507568, | |
| "learning_rate": 1.9307687490499723e-05, | |
| "loss": 2.9272723388671875, | |
| "step": 27300 | |
| }, | |
| { | |
| "epoch": 1.1523552709193563, | |
| "grad_norm": 0.8170751929283142, | |
| "learning_rate": 1.9226818618767298e-05, | |
| "loss": 2.9390142822265624, | |
| "step": 27350 | |
| }, | |
| { | |
| "epoch": 1.1544619533159182, | |
| "grad_norm": 0.7902806997299194, | |
| "learning_rate": 1.914601350828969e-05, | |
| "loss": 2.9641290283203126, | |
| "step": 27400 | |
| }, | |
| { | |
| "epoch": 1.15656863571248, | |
| "grad_norm": 0.7922801971435547, | |
| "learning_rate": 1.9065273051509898e-05, | |
| "loss": 2.9279544067382814, | |
| "step": 27450 | |
| }, | |
| { | |
| "epoch": 1.158675318109042, | |
| "grad_norm": 0.79641193151474, | |
| "learning_rate": 1.8984598140156853e-05, | |
| "loss": 2.93123291015625, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 1.1607820005056038, | |
| "grad_norm": 0.7968478202819824, | |
| "learning_rate": 1.8903989665235554e-05, | |
| "loss": 2.9517864990234375, | |
| "step": 27550 | |
| }, | |
| { | |
| "epoch": 1.1628886829021656, | |
| "grad_norm": 0.8142721652984619, | |
| "learning_rate": 1.8823448517017298e-05, | |
| "loss": 2.9202032470703125, | |
| "step": 27600 | |
| }, | |
| { | |
| "epoch": 1.1649953652987275, | |
| "grad_norm": 0.7819079756736755, | |
| "learning_rate": 1.8742975585029748e-05, | |
| "loss": 2.9275604248046876, | |
| "step": 27650 | |
| }, | |
| { | |
| "epoch": 1.1671020476952894, | |
| "grad_norm": 0.7684738636016846, | |
| "learning_rate": 1.8662571758047193e-05, | |
| "loss": 2.9260235595703126, | |
| "step": 27700 | |
| }, | |
| { | |
| "epoch": 1.1692087300918514, | |
| "grad_norm": 0.772836446762085, | |
| "learning_rate": 1.858223792408068e-05, | |
| "loss": 2.9181884765625, | |
| "step": 27750 | |
| }, | |
| { | |
| "epoch": 1.1713154124884133, | |
| "grad_norm": 0.7890585064888, | |
| "learning_rate": 1.8503579529081115e-05, | |
| "loss": 2.9714337158203126, | |
| "step": 27800 | |
| }, | |
| { | |
| "epoch": 1.1734220948849752, | |
| "grad_norm": 0.7656514048576355, | |
| "learning_rate": 1.842338689806163e-05, | |
| "loss": 2.944093017578125, | |
| "step": 27850 | |
| }, | |
| { | |
| "epoch": 1.175528777281537, | |
| "grad_norm": 0.8071841597557068, | |
| "learning_rate": 1.8343266901708566e-05, | |
| "loss": 2.9365826416015626, | |
| "step": 27900 | |
| }, | |
| { | |
| "epoch": 1.177635459678099, | |
| "grad_norm": 0.7989082336425781, | |
| "learning_rate": 1.8263220424898265e-05, | |
| "loss": 2.9464288330078126, | |
| "step": 27950 | |
| }, | |
| { | |
| "epoch": 1.1797421420746608, | |
| "grad_norm": 0.7741073966026306, | |
| "learning_rate": 1.8183248351695063e-05, | |
| "loss": 2.9338690185546876, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 1.1797421420746608, | |
| "eval_loss": 2.9263670444488525, | |
| "eval_runtime": 119.1397, | |
| "eval_samples_per_second": 128.782, | |
| "eval_steps_per_second": 4.029, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 1.1818488244712226, | |
| "grad_norm": 0.782497227191925, | |
| "learning_rate": 1.810335156534155e-05, | |
| "loss": 2.9569940185546875, | |
| "step": 28050 | |
| }, | |
| { | |
| "epoch": 1.1839555068677847, | |
| "grad_norm": 0.8218105435371399, | |
| "learning_rate": 1.8023530948248848e-05, | |
| "loss": 2.9056085205078124, | |
| "step": 28100 | |
| }, | |
| { | |
| "epoch": 1.1860621892643466, | |
| "grad_norm": 0.7782419919967651, | |
| "learning_rate": 1.7943787381986808e-05, | |
| "loss": 2.9361358642578126, | |
| "step": 28150 | |
| }, | |
| { | |
| "epoch": 1.1881688716609085, | |
| "grad_norm": 0.8223652839660645, | |
| "learning_rate": 1.7864121747274314e-05, | |
| "loss": 2.932328186035156, | |
| "step": 28200 | |
| }, | |
| { | |
| "epoch": 1.1902755540574703, | |
| "grad_norm": 0.8126109838485718, | |
| "learning_rate": 1.778453492396953e-05, | |
| "loss": 2.948155517578125, | |
| "step": 28250 | |
| }, | |
| { | |
| "epoch": 1.1923822364540322, | |
| "grad_norm": 0.7880218029022217, | |
| "learning_rate": 1.7705027791060232e-05, | |
| "loss": 2.9344052124023436, | |
| "step": 28300 | |
| }, | |
| { | |
| "epoch": 1.194488918850594, | |
| "grad_norm": 0.7809317111968994, | |
| "learning_rate": 1.7625601226654015e-05, | |
| "loss": 2.9467529296875, | |
| "step": 28350 | |
| }, | |
| { | |
| "epoch": 1.196595601247156, | |
| "grad_norm": 0.7934635877609253, | |
| "learning_rate": 1.7546256107968682e-05, | |
| "loss": 2.9407623291015623, | |
| "step": 28400 | |
| }, | |
| { | |
| "epoch": 1.1987022836437178, | |
| "grad_norm": 0.7794665694236755, | |
| "learning_rate": 1.746699331132251e-05, | |
| "loss": 2.9444644165039064, | |
| "step": 28450 | |
| }, | |
| { | |
| "epoch": 1.2008089660402796, | |
| "grad_norm": 0.8297936916351318, | |
| "learning_rate": 1.738781371212455e-05, | |
| "loss": 2.9417449951171877, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 1.2029156484368417, | |
| "grad_norm": 0.8044422268867493, | |
| "learning_rate": 1.7308718184865048e-05, | |
| "loss": 2.9425592041015625, | |
| "step": 28550 | |
| }, | |
| { | |
| "epoch": 1.2050223308334036, | |
| "grad_norm": 0.8233380913734436, | |
| "learning_rate": 1.7229707603105667e-05, | |
| "loss": 2.9383099365234373, | |
| "step": 28600 | |
| }, | |
| { | |
| "epoch": 1.2071290132299655, | |
| "grad_norm": 0.7789031267166138, | |
| "learning_rate": 1.715078283946993e-05, | |
| "loss": 2.932237548828125, | |
| "step": 28650 | |
| }, | |
| { | |
| "epoch": 1.2092356956265273, | |
| "grad_norm": 0.7915503978729248, | |
| "learning_rate": 1.7071944765633534e-05, | |
| "loss": 2.9309442138671873, | |
| "step": 28700 | |
| }, | |
| { | |
| "epoch": 1.2113423780230892, | |
| "grad_norm": 0.7797256112098694, | |
| "learning_rate": 1.6993194252314776e-05, | |
| "loss": 2.927774658203125, | |
| "step": 28750 | |
| }, | |
| { | |
| "epoch": 1.213449060419651, | |
| "grad_norm": 0.7625344395637512, | |
| "learning_rate": 1.6914532169264853e-05, | |
| "loss": 2.9445306396484376, | |
| "step": 28800 | |
| }, | |
| { | |
| "epoch": 1.2155557428162131, | |
| "grad_norm": 0.7801703214645386, | |
| "learning_rate": 1.6835959385258327e-05, | |
| "loss": 2.9501434326171876, | |
| "step": 28850 | |
| }, | |
| { | |
| "epoch": 1.217662425212775, | |
| "grad_norm": 0.7768624424934387, | |
| "learning_rate": 1.6757476768083525e-05, | |
| "loss": 2.9339080810546876, | |
| "step": 28900 | |
| }, | |
| { | |
| "epoch": 1.2197691076093369, | |
| "grad_norm": 0.7735788226127625, | |
| "learning_rate": 1.66790851845329e-05, | |
| "loss": 2.929815368652344, | |
| "step": 28950 | |
| }, | |
| { | |
| "epoch": 1.2218757900058987, | |
| "grad_norm": 0.7853214740753174, | |
| "learning_rate": 1.6600785500393527e-05, | |
| "loss": 2.95386962890625, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 1.2218757900058987, | |
| "eval_loss": 2.9254322052001953, | |
| "eval_runtime": 119.1358, | |
| "eval_samples_per_second": 128.786, | |
| "eval_steps_per_second": 4.029, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 1.2239824724024606, | |
| "grad_norm": 0.7844828963279724, | |
| "learning_rate": 1.652257858043747e-05, | |
| "loss": 2.937286376953125, | |
| "step": 29050 | |
| }, | |
| { | |
| "epoch": 1.2260891547990225, | |
| "grad_norm": 0.7900869846343994, | |
| "learning_rate": 1.6444465288412315e-05, | |
| "loss": 2.9779071044921874, | |
| "step": 29100 | |
| }, | |
| { | |
| "epoch": 1.2281958371955843, | |
| "grad_norm": 0.7900272011756897, | |
| "learning_rate": 1.6366446487031546e-05, | |
| "loss": 2.9597027587890623, | |
| "step": 29150 | |
| }, | |
| { | |
| "epoch": 1.2303025195921462, | |
| "grad_norm": 0.8046372532844543, | |
| "learning_rate": 1.6288523037965092e-05, | |
| "loss": 2.9388983154296877, | |
| "step": 29200 | |
| }, | |
| { | |
| "epoch": 1.232409201988708, | |
| "grad_norm": 0.7946169972419739, | |
| "learning_rate": 1.6210695801829737e-05, | |
| "loss": 2.9342681884765627, | |
| "step": 29250 | |
| }, | |
| { | |
| "epoch": 1.2345158843852702, | |
| "grad_norm": 0.7965327501296997, | |
| "learning_rate": 1.613296563817967e-05, | |
| "loss": 2.953227233886719, | |
| "step": 29300 | |
| }, | |
| { | |
| "epoch": 1.236622566781832, | |
| "grad_norm": 0.7878540754318237, | |
| "learning_rate": 1.6055333405496985e-05, | |
| "loss": 2.958277587890625, | |
| "step": 29350 | |
| }, | |
| { | |
| "epoch": 1.2387292491783939, | |
| "grad_norm": 0.8178039789199829, | |
| "learning_rate": 1.5977799961182165e-05, | |
| "loss": 2.938063659667969, | |
| "step": 29400 | |
| }, | |
| { | |
| "epoch": 1.2408359315749558, | |
| "grad_norm": 0.7918407320976257, | |
| "learning_rate": 1.590036616154466e-05, | |
| "loss": 2.93583251953125, | |
| "step": 29450 | |
| }, | |
| { | |
| "epoch": 1.2429426139715176, | |
| "grad_norm": 0.7985025644302368, | |
| "learning_rate": 1.582303286179337e-05, | |
| "loss": 2.931334228515625, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 1.2450492963680795, | |
| "grad_norm": 0.7899501323699951, | |
| "learning_rate": 1.574580091602729e-05, | |
| "loss": 2.9419769287109374, | |
| "step": 29550 | |
| }, | |
| { | |
| "epoch": 1.2471559787646413, | |
| "grad_norm": 0.807771623134613, | |
| "learning_rate": 1.566867117722596e-05, | |
| "loss": 2.925600891113281, | |
| "step": 29600 | |
| }, | |
| { | |
| "epoch": 1.2492626611612034, | |
| "grad_norm": 0.7778540253639221, | |
| "learning_rate": 1.5591644497240153e-05, | |
| "loss": 2.9385748291015625, | |
| "step": 29650 | |
| }, | |
| { | |
| "epoch": 1.2513693435577653, | |
| "grad_norm": 0.8029125332832336, | |
| "learning_rate": 1.55147217267824e-05, | |
| "loss": 2.931204833984375, | |
| "step": 29700 | |
| }, | |
| { | |
| "epoch": 1.2534760259543272, | |
| "grad_norm": 0.767647922039032, | |
| "learning_rate": 1.5437903715417622e-05, | |
| "loss": 2.9307843017578126, | |
| "step": 29750 | |
| }, | |
| { | |
| "epoch": 1.255582708350889, | |
| "grad_norm": 0.8017622232437134, | |
| "learning_rate": 1.536119131155374e-05, | |
| "loss": 2.9459600830078125, | |
| "step": 29800 | |
| }, | |
| { | |
| "epoch": 1.257689390747451, | |
| "grad_norm": 0.7804931998252869, | |
| "learning_rate": 1.528458536243229e-05, | |
| "loss": 2.9380758666992186, | |
| "step": 29850 | |
| }, | |
| { | |
| "epoch": 1.2597960731440128, | |
| "grad_norm": 0.7753841876983643, | |
| "learning_rate": 1.5208086714119113e-05, | |
| "loss": 2.9388946533203124, | |
| "step": 29900 | |
| }, | |
| { | |
| "epoch": 1.2619027555405746, | |
| "grad_norm": 0.7983092665672302, | |
| "learning_rate": 1.5131696211494933e-05, | |
| "loss": 2.936572265625, | |
| "step": 29950 | |
| }, | |
| { | |
| "epoch": 1.2640094379371365, | |
| "grad_norm": 0.7694459557533264, | |
| "learning_rate": 1.5055414698246123e-05, | |
| "loss": 2.9460601806640625, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 1.2640094379371365, | |
| "eval_loss": 2.9237051010131836, | |
| "eval_runtime": 119.3333, | |
| "eval_samples_per_second": 128.573, | |
| "eval_steps_per_second": 4.022, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 1.2661161203336984, | |
| "grad_norm": 0.8067724108695984, | |
| "learning_rate": 1.4979243016855288e-05, | |
| "loss": 2.941827392578125, | |
| "step": 30050 | |
| }, | |
| { | |
| "epoch": 1.2682228027302604, | |
| "grad_norm": 0.7834928631782532, | |
| "learning_rate": 1.4903182008592037e-05, | |
| "loss": 2.9663165283203123, | |
| "step": 30100 | |
| }, | |
| { | |
| "epoch": 1.2703294851268223, | |
| "grad_norm": 0.8025876879692078, | |
| "learning_rate": 1.4827232513503636e-05, | |
| "loss": 2.9382012939453124, | |
| "step": 30150 | |
| }, | |
| { | |
| "epoch": 1.2724361675233842, | |
| "grad_norm": 0.8243033289909363, | |
| "learning_rate": 1.4751395370405796e-05, | |
| "loss": 2.928963928222656, | |
| "step": 30200 | |
| }, | |
| { | |
| "epoch": 1.274542849919946, | |
| "grad_norm": 0.7702295780181885, | |
| "learning_rate": 1.4675671416873338e-05, | |
| "loss": 2.9373565673828126, | |
| "step": 30250 | |
| }, | |
| { | |
| "epoch": 1.276649532316508, | |
| "grad_norm": 0.7878978848457336, | |
| "learning_rate": 1.4600061489230971e-05, | |
| "loss": 2.9364288330078123, | |
| "step": 30300 | |
| }, | |
| { | |
| "epoch": 1.27875621471307, | |
| "grad_norm": 0.7720517516136169, | |
| "learning_rate": 1.4524566422544082e-05, | |
| "loss": 2.9434030151367185, | |
| "step": 30350 | |
| }, | |
| { | |
| "epoch": 1.2808628971096319, | |
| "grad_norm": 0.7778488993644714, | |
| "learning_rate": 1.4449187050609458e-05, | |
| "loss": 2.944191589355469, | |
| "step": 30400 | |
| }, | |
| { | |
| "epoch": 1.2829695795061937, | |
| "grad_norm": 0.7929774522781372, | |
| "learning_rate": 1.437392420594615e-05, | |
| "loss": 2.9433721923828124, | |
| "step": 30450 | |
| }, | |
| { | |
| "epoch": 1.2850762619027556, | |
| "grad_norm": 0.8068262934684753, | |
| "learning_rate": 1.4298778719786194e-05, | |
| "loss": 2.9357684326171873, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 1.2871829442993175, | |
| "grad_norm": 0.80678790807724, | |
| "learning_rate": 1.4223751422065496e-05, | |
| "loss": 2.9380233764648436, | |
| "step": 30550 | |
| }, | |
| { | |
| "epoch": 1.2892896266958793, | |
| "grad_norm": 0.769245982170105, | |
| "learning_rate": 1.4148843141414627e-05, | |
| "loss": 2.9365478515625, | |
| "step": 30600 | |
| }, | |
| { | |
| "epoch": 1.2913963090924412, | |
| "grad_norm": 0.8063893914222717, | |
| "learning_rate": 1.407405470514972e-05, | |
| "loss": 2.949693603515625, | |
| "step": 30650 | |
| }, | |
| { | |
| "epoch": 1.293502991489003, | |
| "grad_norm": 0.8096292614936829, | |
| "learning_rate": 1.3999386939263256e-05, | |
| "loss": 2.93534423828125, | |
| "step": 30700 | |
| }, | |
| { | |
| "epoch": 1.295609673885565, | |
| "grad_norm": 0.7904254198074341, | |
| "learning_rate": 1.392484066841502e-05, | |
| "loss": 2.9535675048828125, | |
| "step": 30750 | |
| }, | |
| { | |
| "epoch": 1.2977163562821268, | |
| "grad_norm": 0.7776016592979431, | |
| "learning_rate": 1.3850416715922956e-05, | |
| "loss": 2.930771484375, | |
| "step": 30800 | |
| }, | |
| { | |
| "epoch": 1.2998230386786889, | |
| "grad_norm": 0.8131218552589417, | |
| "learning_rate": 1.377611590375405e-05, | |
| "loss": 2.925182189941406, | |
| "step": 30850 | |
| }, | |
| { | |
| "epoch": 1.3019297210752507, | |
| "grad_norm": 0.7831723093986511, | |
| "learning_rate": 1.370193905251531e-05, | |
| "loss": 2.9488519287109374, | |
| "step": 30900 | |
| }, | |
| { | |
| "epoch": 1.3040364034718126, | |
| "grad_norm": 0.7675512433052063, | |
| "learning_rate": 1.3627886981444652e-05, | |
| "loss": 2.9460296630859375, | |
| "step": 30950 | |
| }, | |
| { | |
| "epoch": 1.3061430858683745, | |
| "grad_norm": 0.7997528314590454, | |
| "learning_rate": 1.3553960508401875e-05, | |
| "loss": 2.953719482421875, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 1.3061430858683745, | |
| "eval_loss": 2.9219441413879395, | |
| "eval_runtime": 133.5018, | |
| "eval_samples_per_second": 114.927, | |
| "eval_steps_per_second": 3.595, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 1.3082497682649363, | |
| "grad_norm": 0.7918239831924438, | |
| "learning_rate": 1.3480160449859616e-05, | |
| "loss": 2.932680969238281, | |
| "step": 31050 | |
| }, | |
| { | |
| "epoch": 1.3103564506614982, | |
| "grad_norm": 0.7773085832595825, | |
| "learning_rate": 1.3406487620894375e-05, | |
| "loss": 2.925394287109375, | |
| "step": 31100 | |
| }, | |
| { | |
| "epoch": 1.3124631330580603, | |
| "grad_norm": 0.7864688038825989, | |
| "learning_rate": 1.3332942835177431e-05, | |
| "loss": 2.925726318359375, | |
| "step": 31150 | |
| }, | |
| { | |
| "epoch": 1.3145698154546221, | |
| "grad_norm": 0.7832627892494202, | |
| "learning_rate": 1.325952690496594e-05, | |
| "loss": 2.947120361328125, | |
| "step": 31200 | |
| }, | |
| { | |
| "epoch": 1.316676497851184, | |
| "grad_norm": 0.7992543578147888, | |
| "learning_rate": 1.3186240641093927e-05, | |
| "loss": 2.9400369262695314, | |
| "step": 31250 | |
| }, | |
| { | |
| "epoch": 1.3187831802477459, | |
| "grad_norm": 0.8116305470466614, | |
| "learning_rate": 1.3113084852963304e-05, | |
| "loss": 2.9369207763671876, | |
| "step": 31300 | |
| }, | |
| { | |
| "epoch": 1.3208898626443077, | |
| "grad_norm": 0.7965041995048523, | |
| "learning_rate": 1.3040060348534987e-05, | |
| "loss": 2.948853759765625, | |
| "step": 31350 | |
| }, | |
| { | |
| "epoch": 1.3229965450408696, | |
| "grad_norm": 0.8035820126533508, | |
| "learning_rate": 1.2967167934319935e-05, | |
| "loss": 2.950780029296875, | |
| "step": 31400 | |
| }, | |
| { | |
| "epoch": 1.3251032274374315, | |
| "grad_norm": 0.7838873267173767, | |
| "learning_rate": 1.2894408415370234e-05, | |
| "loss": 2.916210021972656, | |
| "step": 31450 | |
| }, | |
| { | |
| "epoch": 1.3272099098339933, | |
| "grad_norm": 0.834223210811615, | |
| "learning_rate": 1.2821782595270238e-05, | |
| "loss": 2.9532537841796875, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 1.3293165922305552, | |
| "grad_norm": 0.807707667350769, | |
| "learning_rate": 1.274929127612769e-05, | |
| "loss": 2.9319525146484375, | |
| "step": 31550 | |
| }, | |
| { | |
| "epoch": 1.3314232746271173, | |
| "grad_norm": 0.7963380217552185, | |
| "learning_rate": 1.2676935258564832e-05, | |
| "loss": 2.950096435546875, | |
| "step": 31600 | |
| }, | |
| { | |
| "epoch": 1.3335299570236792, | |
| "grad_norm": 0.7738616466522217, | |
| "learning_rate": 1.2604715341709592e-05, | |
| "loss": 2.9467239379882812, | |
| "step": 31650 | |
| }, | |
| { | |
| "epoch": 1.335636639420241, | |
| "grad_norm": 0.7763379812240601, | |
| "learning_rate": 1.2532632323186761e-05, | |
| "loss": 2.9259683227539064, | |
| "step": 31700 | |
| }, | |
| { | |
| "epoch": 1.3377433218168029, | |
| "grad_norm": 0.7878997325897217, | |
| "learning_rate": 1.2460686999109154e-05, | |
| "loss": 2.9317642211914063, | |
| "step": 31750 | |
| }, | |
| { | |
| "epoch": 1.3398500042133648, | |
| "grad_norm": 0.8008918762207031, | |
| "learning_rate": 1.2388880164068836e-05, | |
| "loss": 2.926044921875, | |
| "step": 31800 | |
| }, | |
| { | |
| "epoch": 1.3419566866099266, | |
| "grad_norm": 0.7763420939445496, | |
| "learning_rate": 1.2317212611128343e-05, | |
| "loss": 2.9527883911132813, | |
| "step": 31850 | |
| }, | |
| { | |
| "epoch": 1.3440633690064887, | |
| "grad_norm": 0.783769428730011, | |
| "learning_rate": 1.2245685131811951e-05, | |
| "loss": 2.9466986083984374, | |
| "step": 31900 | |
| }, | |
| { | |
| "epoch": 1.3461700514030506, | |
| "grad_norm": 0.7776080369949341, | |
| "learning_rate": 1.2174298516096869e-05, | |
| "loss": 2.95081787109375, | |
| "step": 31950 | |
| }, | |
| { | |
| "epoch": 1.3482767337996124, | |
| "grad_norm": 0.78815096616745, | |
| "learning_rate": 1.2104477058396635e-05, | |
| "loss": 2.9214215087890625, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 1.3482767337996124, | |
| "eval_loss": 2.9208405017852783, | |
| "eval_runtime": 133.39, | |
| "eval_samples_per_second": 115.024, | |
| "eval_steps_per_second": 3.598, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 1.3503834161961743, | |
| "grad_norm": 0.7701238989830017, | |
| "learning_rate": 1.2033371677105546e-05, | |
| "loss": 2.93294677734375, | |
| "step": 32050 | |
| }, | |
| { | |
| "epoch": 1.3524900985927362, | |
| "grad_norm": 0.7919666767120361, | |
| "learning_rate": 1.1962409504287925e-05, | |
| "loss": 2.9343310546875, | |
| "step": 32100 | |
| }, | |
| { | |
| "epoch": 1.354596780989298, | |
| "grad_norm": 0.785975456237793, | |
| "learning_rate": 1.189159132367755e-05, | |
| "loss": 2.9307794189453125, | |
| "step": 32150 | |
| }, | |
| { | |
| "epoch": 1.35670346338586, | |
| "grad_norm": 0.827540397644043, | |
| "learning_rate": 1.1820917917417889e-05, | |
| "loss": 2.9201791381835935, | |
| "step": 32200 | |
| }, | |
| { | |
| "epoch": 1.3588101457824218, | |
| "grad_norm": 0.7883698344230652, | |
| "learning_rate": 1.1750390066053452e-05, | |
| "loss": 2.94340576171875, | |
| "step": 32250 | |
| }, | |
| { | |
| "epoch": 1.3609168281789836, | |
| "grad_norm": 0.7953295111656189, | |
| "learning_rate": 1.1680008548521185e-05, | |
| "loss": 2.93116455078125, | |
| "step": 32300 | |
| }, | |
| { | |
| "epoch": 1.3630235105755455, | |
| "grad_norm": 0.7758684754371643, | |
| "learning_rate": 1.1609774142141867e-05, | |
| "loss": 2.943220520019531, | |
| "step": 32350 | |
| }, | |
| { | |
| "epoch": 1.3651301929721076, | |
| "grad_norm": 0.8139975666999817, | |
| "learning_rate": 1.1539687622611543e-05, | |
| "loss": 2.94186767578125, | |
| "step": 32400 | |
| }, | |
| { | |
| "epoch": 1.3672368753686694, | |
| "grad_norm": 0.8057471513748169, | |
| "learning_rate": 1.1469749763992904e-05, | |
| "loss": 2.9390771484375, | |
| "step": 32450 | |
| }, | |
| { | |
| "epoch": 1.3693435577652313, | |
| "grad_norm": 0.820174515247345, | |
| "learning_rate": 1.139996133870682e-05, | |
| "loss": 2.944349365234375, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 1.3714502401617932, | |
| "grad_norm": 0.8091501593589783, | |
| "learning_rate": 1.1330323117523706e-05, | |
| "loss": 2.9372442626953124, | |
| "step": 32550 | |
| }, | |
| { | |
| "epoch": 1.373556922558355, | |
| "grad_norm": 0.8005126714706421, | |
| "learning_rate": 1.1260835869555106e-05, | |
| "loss": 2.9346054077148436, | |
| "step": 32600 | |
| }, | |
| { | |
| "epoch": 1.3756636049549171, | |
| "grad_norm": 0.8052904605865479, | |
| "learning_rate": 1.1191500362245152e-05, | |
| "loss": 2.9285906982421874, | |
| "step": 32650 | |
| }, | |
| { | |
| "epoch": 1.377770287351479, | |
| "grad_norm": 0.780090868473053, | |
| "learning_rate": 1.1122317361362075e-05, | |
| "loss": 2.9148419189453123, | |
| "step": 32700 | |
| }, | |
| { | |
| "epoch": 1.3798769697480409, | |
| "grad_norm": 0.7812498807907104, | |
| "learning_rate": 1.1053287630989775e-05, | |
| "loss": 2.91372314453125, | |
| "step": 32750 | |
| }, | |
| { | |
| "epoch": 1.3819836521446027, | |
| "grad_norm": 0.7818878889083862, | |
| "learning_rate": 1.0984411933519368e-05, | |
| "loss": 2.922346496582031, | |
| "step": 32800 | |
| }, | |
| { | |
| "epoch": 1.3840903345411646, | |
| "grad_norm": 0.7950148582458496, | |
| "learning_rate": 1.091569102964079e-05, | |
| "loss": 2.93470458984375, | |
| "step": 32850 | |
| }, | |
| { | |
| "epoch": 1.3861970169377265, | |
| "grad_norm": 0.8164188265800476, | |
| "learning_rate": 1.0847125678334347e-05, | |
| "loss": 2.9304373168945315, | |
| "step": 32900 | |
| }, | |
| { | |
| "epoch": 1.3883036993342883, | |
| "grad_norm": 0.7998409271240234, | |
| "learning_rate": 1.0778716636862395e-05, | |
| "loss": 2.9494195556640626, | |
| "step": 32950 | |
| }, | |
| { | |
| "epoch": 1.3904103817308502, | |
| "grad_norm": 0.7881671786308289, | |
| "learning_rate": 1.0710464660760914e-05, | |
| "loss": 2.924466552734375, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 1.3904103817308502, | |
| "eval_loss": 2.9189743995666504, | |
| "eval_runtime": 133.7151, | |
| "eval_samples_per_second": 114.744, | |
| "eval_steps_per_second": 3.59, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 1.392517064127412, | |
| "grad_norm": 0.7967885136604309, | |
| "learning_rate": 1.0642370503831217e-05, | |
| "loss": 2.9310653686523436, | |
| "step": 33050 | |
| }, | |
| { | |
| "epoch": 1.394623746523974, | |
| "grad_norm": 0.7761057019233704, | |
| "learning_rate": 1.0574434918131585e-05, | |
| "loss": 2.922852783203125, | |
| "step": 33100 | |
| }, | |
| { | |
| "epoch": 1.396730428920536, | |
| "grad_norm": 0.8000385165214539, | |
| "learning_rate": 1.0506658653968967e-05, | |
| "loss": 2.93267822265625, | |
| "step": 33150 | |
| }, | |
| { | |
| "epoch": 1.3988371113170979, | |
| "grad_norm": 0.7959194779396057, | |
| "learning_rate": 1.0439042459890741e-05, | |
| "loss": 2.95590087890625, | |
| "step": 33200 | |
| }, | |
| { | |
| "epoch": 1.4009437937136597, | |
| "grad_norm": 0.7842496633529663, | |
| "learning_rate": 1.0371587082676367e-05, | |
| "loss": 2.9364727783203124, | |
| "step": 33250 | |
| }, | |
| { | |
| "epoch": 1.4030504761102216, | |
| "grad_norm": 0.8106942176818848, | |
| "learning_rate": 1.0304293267329219e-05, | |
| "loss": 2.944447021484375, | |
| "step": 33300 | |
| }, | |
| { | |
| "epoch": 1.4051571585067835, | |
| "grad_norm": 0.8099303245544434, | |
| "learning_rate": 1.0237161757068273e-05, | |
| "loss": 2.914239807128906, | |
| "step": 33350 | |
| }, | |
| { | |
| "epoch": 1.4072638409033453, | |
| "grad_norm": 0.791988730430603, | |
| "learning_rate": 1.0170193293319999e-05, | |
| "loss": 2.9319525146484375, | |
| "step": 33400 | |
| }, | |
| { | |
| "epoch": 1.4093705232999074, | |
| "grad_norm": 0.7973111271858215, | |
| "learning_rate": 1.0103388615710068e-05, | |
| "loss": 2.93970703125, | |
| "step": 33450 | |
| }, | |
| { | |
| "epoch": 1.4114772056964693, | |
| "grad_norm": 0.8064714074134827, | |
| "learning_rate": 1.0036748462055276e-05, | |
| "loss": 2.9383502197265625, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 1.4135838880930311, | |
| "grad_norm": 0.8036237359046936, | |
| "learning_rate": 9.970273568355321e-06, | |
| "loss": 2.9164556884765624, | |
| "step": 33550 | |
| }, | |
| { | |
| "epoch": 1.415690570489593, | |
| "grad_norm": 0.8388088345527649, | |
| "learning_rate": 9.90396466878471e-06, | |
| "loss": 2.9235543823242187, | |
| "step": 33600 | |
| }, | |
| { | |
| "epoch": 1.4177972528861549, | |
| "grad_norm": 0.7946650981903076, | |
| "learning_rate": 9.837822495684671e-06, | |
| "loss": 2.9373529052734373, | |
| "step": 33650 | |
| }, | |
| { | |
| "epoch": 1.4199039352827167, | |
| "grad_norm": 0.8091840744018555, | |
| "learning_rate": 9.771847779555005e-06, | |
| "loss": 2.957337646484375, | |
| "step": 33700 | |
| }, | |
| { | |
| "epoch": 1.4220106176792786, | |
| "grad_norm": 0.8072313666343689, | |
| "learning_rate": 9.706041249046088e-06, | |
| "loss": 2.932064208984375, | |
| "step": 33750 | |
| }, | |
| { | |
| "epoch": 1.4241173000758405, | |
| "grad_norm": 0.786514937877655, | |
| "learning_rate": 9.640403630950756e-06, | |
| "loss": 2.9320501708984374, | |
| "step": 33800 | |
| }, | |
| { | |
| "epoch": 1.4262239824724023, | |
| "grad_norm": 0.7859194874763489, | |
| "learning_rate": 9.574935650196346e-06, | |
| "loss": 2.952900390625, | |
| "step": 33850 | |
| }, | |
| { | |
| "epoch": 1.4283306648689644, | |
| "grad_norm": 0.7723864912986755, | |
| "learning_rate": 9.509638029836612e-06, | |
| "loss": 2.926435546875, | |
| "step": 33900 | |
| }, | |
| { | |
| "epoch": 1.4304373472655263, | |
| "grad_norm": 0.7824335694313049, | |
| "learning_rate": 9.444511491043828e-06, | |
| "loss": 2.9314031982421875, | |
| "step": 33950 | |
| }, | |
| { | |
| "epoch": 1.4325440296620882, | |
| "grad_norm": 0.8018701672554016, | |
| "learning_rate": 9.38085415956809e-06, | |
| "loss": 2.92718017578125, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 1.4325440296620882, | |
| "eval_loss": 2.918320417404175, | |
| "eval_runtime": 132.6789, | |
| "eval_samples_per_second": 115.64, | |
| "eval_steps_per_second": 3.618, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 1.43465071205865, | |
| "grad_norm": 0.7768269181251526, | |
| "learning_rate": 9.316068482477261e-06, | |
| "loss": 2.924657897949219, | |
| "step": 34050 | |
| }, | |
| { | |
| "epoch": 1.436757394455212, | |
| "grad_norm": 0.8097816109657288, | |
| "learning_rate": 9.251456024810538e-06, | |
| "loss": 2.927236328125, | |
| "step": 34100 | |
| }, | |
| { | |
| "epoch": 1.4388640768517738, | |
| "grad_norm": 0.8051019906997681, | |
| "learning_rate": 9.187017500172984e-06, | |
| "loss": 2.944189453125, | |
| "step": 34150 | |
| }, | |
| { | |
| "epoch": 1.4409707592483358, | |
| "grad_norm": 0.8041905760765076, | |
| "learning_rate": 9.122753620248647e-06, | |
| "loss": 2.9382794189453123, | |
| "step": 34200 | |
| }, | |
| { | |
| "epoch": 1.4430774416448977, | |
| "grad_norm": 0.7908426523208618, | |
| "learning_rate": 9.058665094792768e-06, | |
| "loss": 2.9181817626953124, | |
| "step": 34250 | |
| }, | |
| { | |
| "epoch": 1.4451841240414596, | |
| "grad_norm": 0.8021414875984192, | |
| "learning_rate": 8.994752631623862e-06, | |
| "loss": 2.9588043212890627, | |
| "step": 34300 | |
| }, | |
| { | |
| "epoch": 1.4472908064380214, | |
| "grad_norm": 0.8225296139717102, | |
| "learning_rate": 8.93101693661599e-06, | |
| "loss": 2.93732666015625, | |
| "step": 34350 | |
| }, | |
| { | |
| "epoch": 1.4493974888345833, | |
| "grad_norm": 0.7841650247573853, | |
| "learning_rate": 8.86745871369087e-06, | |
| "loss": 2.9336138916015626, | |
| "step": 34400 | |
| }, | |
| { | |
| "epoch": 1.4515041712311452, | |
| "grad_norm": 0.7821727991104126, | |
| "learning_rate": 8.804078664810193e-06, | |
| "loss": 2.942218017578125, | |
| "step": 34450 | |
| }, | |
| { | |
| "epoch": 1.453610853627707, | |
| "grad_norm": 0.796100914478302, | |
| "learning_rate": 8.740877489967797e-06, | |
| "loss": 2.93341064453125, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 1.455717536024269, | |
| "grad_norm": 0.763460636138916, | |
| "learning_rate": 8.677855887181967e-06, | |
| "loss": 2.932227783203125, | |
| "step": 34550 | |
| }, | |
| { | |
| "epoch": 1.4578242184208308, | |
| "grad_norm": 0.8215290904045105, | |
| "learning_rate": 8.615014552487751e-06, | |
| "loss": 2.930343017578125, | |
| "step": 34600 | |
| }, | |
| { | |
| "epoch": 1.4599309008173929, | |
| "grad_norm": 0.799164891242981, | |
| "learning_rate": 8.552354179929212e-06, | |
| "loss": 2.922919921875, | |
| "step": 34650 | |
| }, | |
| { | |
| "epoch": 1.4620375832139547, | |
| "grad_norm": 0.8057267069816589, | |
| "learning_rate": 8.489875461551827e-06, | |
| "loss": 2.9110009765625, | |
| "step": 34700 | |
| }, | |
| { | |
| "epoch": 1.4641442656105166, | |
| "grad_norm": 0.7977777719497681, | |
| "learning_rate": 8.427579087394782e-06, | |
| "loss": 2.947958984375, | |
| "step": 34750 | |
| }, | |
| { | |
| "epoch": 1.4662509480070784, | |
| "grad_norm": 0.7965235710144043, | |
| "learning_rate": 8.365465745483422e-06, | |
| "loss": 2.940159912109375, | |
| "step": 34800 | |
| }, | |
| { | |
| "epoch": 1.4683576304036403, | |
| "grad_norm": 0.786981463432312, | |
| "learning_rate": 8.303536121821579e-06, | |
| "loss": 2.9247445678710937, | |
| "step": 34850 | |
| }, | |
| { | |
| "epoch": 1.4704643128002022, | |
| "grad_norm": 0.7833293676376343, | |
| "learning_rate": 8.241790900384041e-06, | |
| "loss": 2.933876647949219, | |
| "step": 34900 | |
| }, | |
| { | |
| "epoch": 1.4725709951967643, | |
| "grad_norm": 0.8161737322807312, | |
| "learning_rate": 8.180230763108984e-06, | |
| "loss": 2.9441455078125, | |
| "step": 34950 | |
| }, | |
| { | |
| "epoch": 1.4746776775933261, | |
| "grad_norm": 0.8140338659286499, | |
| "learning_rate": 8.11885638989045e-06, | |
| "loss": 2.9499853515625, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 1.4746776775933261, | |
| "eval_loss": 2.9171628952026367, | |
| "eval_runtime": 132.8636, | |
| "eval_samples_per_second": 115.479, | |
| "eval_steps_per_second": 3.613, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 1.476784359989888, | |
| "grad_norm": 0.8166197538375854, | |
| "learning_rate": 8.057668458570836e-06, | |
| "loss": 2.9236883544921874, | |
| "step": 35050 | |
| }, | |
| { | |
| "epoch": 1.4788910423864499, | |
| "grad_norm": 0.8071980476379395, | |
| "learning_rate": 7.996667644933375e-06, | |
| "loss": 2.9362152099609373, | |
| "step": 35100 | |
| }, | |
| { | |
| "epoch": 1.4809977247830117, | |
| "grad_norm": 0.7924031615257263, | |
| "learning_rate": 7.935854622694735e-06, | |
| "loss": 2.9321023559570314, | |
| "step": 35150 | |
| }, | |
| { | |
| "epoch": 1.4831044071795736, | |
| "grad_norm": 0.7746392488479614, | |
| "learning_rate": 7.87523006349751e-06, | |
| "loss": 2.9222271728515623, | |
| "step": 35200 | |
| }, | |
| { | |
| "epoch": 1.4852110895761355, | |
| "grad_norm": 0.8078408241271973, | |
| "learning_rate": 7.814794636902864e-06, | |
| "loss": 2.943782958984375, | |
| "step": 35250 | |
| }, | |
| { | |
| "epoch": 1.4873177719726973, | |
| "grad_norm": 0.8177636861801147, | |
| "learning_rate": 7.754549010383086e-06, | |
| "loss": 2.9366082763671875, | |
| "step": 35300 | |
| }, | |
| { | |
| "epoch": 1.4894244543692592, | |
| "grad_norm": 0.8233819007873535, | |
| "learning_rate": 7.694493849314249e-06, | |
| "loss": 2.949815368652344, | |
| "step": 35350 | |
| }, | |
| { | |
| "epoch": 1.491531136765821, | |
| "grad_norm": 0.7855923175811768, | |
| "learning_rate": 7.634629816968843e-06, | |
| "loss": 2.9289474487304688, | |
| "step": 35400 | |
| }, | |
| { | |
| "epoch": 1.4936378191623831, | |
| "grad_norm": 0.7993690967559814, | |
| "learning_rate": 7.5749575745084704e-06, | |
| "loss": 2.9322943115234374, | |
| "step": 35450 | |
| }, | |
| { | |
| "epoch": 1.495744501558945, | |
| "grad_norm": 0.8059825897216797, | |
| "learning_rate": 7.515477780976532e-06, | |
| "loss": 2.93564697265625, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 1.4978511839555069, | |
| "grad_norm": 0.806198239326477, | |
| "learning_rate": 7.456191093290932e-06, | |
| "loss": 2.91688720703125, | |
| "step": 35550 | |
| }, | |
| { | |
| "epoch": 1.4999578663520687, | |
| "grad_norm": 0.7923563122749329, | |
| "learning_rate": 7.39709816623686e-06, | |
| "loss": 2.9082913208007812, | |
| "step": 35600 | |
| }, | |
| { | |
| "epoch": 1.5020645487486306, | |
| "grad_norm": 0.8135302066802979, | |
| "learning_rate": 7.338199652459521e-06, | |
| "loss": 2.9541253662109375, | |
| "step": 35650 | |
| }, | |
| { | |
| "epoch": 1.5041712311451927, | |
| "grad_norm": 0.7841634154319763, | |
| "learning_rate": 7.279496202456948e-06, | |
| "loss": 2.9492755126953125, | |
| "step": 35700 | |
| }, | |
| { | |
| "epoch": 1.5062779135417546, | |
| "grad_norm": 0.8015527725219727, | |
| "learning_rate": 7.220988464572812e-06, | |
| "loss": 2.9421246337890623, | |
| "step": 35750 | |
| }, | |
| { | |
| "epoch": 1.5083845959383164, | |
| "grad_norm": 0.8087325692176819, | |
| "learning_rate": 7.162677084989283e-06, | |
| "loss": 2.94099609375, | |
| "step": 35800 | |
| }, | |
| { | |
| "epoch": 1.5104912783348783, | |
| "grad_norm": 0.7978482246398926, | |
| "learning_rate": 7.104562707719839e-06, | |
| "loss": 2.943082275390625, | |
| "step": 35850 | |
| }, | |
| { | |
| "epoch": 1.5125979607314401, | |
| "grad_norm": 0.802884578704834, | |
| "learning_rate": 7.04664597460222e-06, | |
| "loss": 2.915927734375, | |
| "step": 35900 | |
| }, | |
| { | |
| "epoch": 1.514704643128002, | |
| "grad_norm": 0.8194138407707214, | |
| "learning_rate": 6.9889275252913e-06, | |
| "loss": 2.93603515625, | |
| "step": 35950 | |
| }, | |
| { | |
| "epoch": 1.5168113255245639, | |
| "grad_norm": 0.7744643092155457, | |
| "learning_rate": 6.93140799725201e-06, | |
| "loss": 2.9371746826171874, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 1.5168113255245639, | |
| "eval_loss": 2.9163427352905273, | |
| "eval_runtime": 132.434, | |
| "eval_samples_per_second": 115.854, | |
| "eval_steps_per_second": 3.624, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 1.5189180079211257, | |
| "grad_norm": 0.7795388102531433, | |
| "learning_rate": 6.874088025752346e-06, | |
| "loss": 2.9265447998046876, | |
| "step": 36050 | |
| }, | |
| { | |
| "epoch": 1.5210246903176876, | |
| "grad_norm": 0.8050419092178345, | |
| "learning_rate": 6.8169682438563006e-06, | |
| "loss": 2.9376104736328124, | |
| "step": 36100 | |
| }, | |
| { | |
| "epoch": 1.5231313727142495, | |
| "grad_norm": 0.790391743183136, | |
| "learning_rate": 6.760049282416911e-06, | |
| "loss": 2.943501281738281, | |
| "step": 36150 | |
| }, | |
| { | |
| "epoch": 1.5252380551108113, | |
| "grad_norm": 0.8023267984390259, | |
| "learning_rate": 6.703331770069251e-06, | |
| "loss": 2.934488525390625, | |
| "step": 36200 | |
| }, | |
| { | |
| "epoch": 1.5273447375073734, | |
| "grad_norm": 0.808557391166687, | |
| "learning_rate": 6.646816333223557e-06, | |
| "loss": 2.925758056640625, | |
| "step": 36250 | |
| }, | |
| { | |
| "epoch": 1.5294514199039353, | |
| "grad_norm": 0.7777003049850464, | |
| "learning_rate": 6.59162786031825e-06, | |
| "loss": 2.923520202636719, | |
| "step": 36300 | |
| }, | |
| { | |
| "epoch": 1.5315581023004972, | |
| "grad_norm": 0.8037623763084412, | |
| "learning_rate": 6.535514372260088e-06, | |
| "loss": 2.946346435546875, | |
| "step": 36350 | |
| }, | |
| { | |
| "epoch": 1.533664784697059, | |
| "grad_norm": 0.8090129494667053, | |
| "learning_rate": 6.479604813144327e-06, | |
| "loss": 2.93824462890625, | |
| "step": 36400 | |
| }, | |
| { | |
| "epoch": 1.5357714670936211, | |
| "grad_norm": 0.818238377571106, | |
| "learning_rate": 6.423899800457817e-06, | |
| "loss": 2.9253659057617187, | |
| "step": 36450 | |
| }, | |
| { | |
| "epoch": 1.537878149490183, | |
| "grad_norm": 0.8148744106292725, | |
| "learning_rate": 6.368399949428333e-06, | |
| "loss": 2.91279052734375, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 1.5399848318867448, | |
| "grad_norm": 0.781205952167511, | |
| "learning_rate": 6.313105873017769e-06, | |
| "loss": 2.940987854003906, | |
| "step": 36550 | |
| }, | |
| { | |
| "epoch": 1.5420915142833067, | |
| "grad_norm": 0.784307062625885, | |
| "learning_rate": 6.258018181915351e-06, | |
| "loss": 2.9270831298828126, | |
| "step": 36600 | |
| }, | |
| { | |
| "epoch": 1.5441981966798686, | |
| "grad_norm": 0.777840793132782, | |
| "learning_rate": 6.203137484530905e-06, | |
| "loss": 2.9171002197265623, | |
| "step": 36650 | |
| }, | |
| { | |
| "epoch": 1.5463048790764304, | |
| "grad_norm": 0.789941668510437, | |
| "learning_rate": 6.14846438698814e-06, | |
| "loss": 2.932503356933594, | |
| "step": 36700 | |
| }, | |
| { | |
| "epoch": 1.5484115614729923, | |
| "grad_norm": 0.7811296582221985, | |
| "learning_rate": 6.093999493117969e-06, | |
| "loss": 2.9210369873046873, | |
| "step": 36750 | |
| }, | |
| { | |
| "epoch": 1.5505182438695542, | |
| "grad_norm": 0.7956785559654236, | |
| "learning_rate": 6.039743404451795e-06, | |
| "loss": 2.935350341796875, | |
| "step": 36800 | |
| }, | |
| { | |
| "epoch": 1.552624926266116, | |
| "grad_norm": 0.8070129752159119, | |
| "learning_rate": 5.985696720214923e-06, | |
| "loss": 2.936260070800781, | |
| "step": 36850 | |
| }, | |
| { | |
| "epoch": 1.554731608662678, | |
| "grad_norm": 0.808233380317688, | |
| "learning_rate": 5.931860037319886e-06, | |
| "loss": 2.9369403076171876, | |
| "step": 36900 | |
| }, | |
| { | |
| "epoch": 1.5568382910592398, | |
| "grad_norm": 0.8150511384010315, | |
| "learning_rate": 5.878233950359901e-06, | |
| "loss": 2.9321359252929686, | |
| "step": 36950 | |
| }, | |
| { | |
| "epoch": 1.5589449734558019, | |
| "grad_norm": 0.79117351770401, | |
| "learning_rate": 5.824819051602287e-06, | |
| "loss": 2.921833190917969, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 1.5589449734558019, | |
| "eval_loss": 2.9154181480407715, | |
| "eval_runtime": 110.7687, | |
| "eval_samples_per_second": 138.514, | |
| "eval_steps_per_second": 4.333, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 1.5610516558523637, | |
| "grad_norm": 0.7824537754058838, | |
| "learning_rate": 5.7716159309819e-06, | |
| "loss": 2.912900390625, | |
| "step": 37050 | |
| }, | |
| { | |
| "epoch": 1.5631583382489256, | |
| "grad_norm": 0.7897117137908936, | |
| "learning_rate": 5.718625176094639e-06, | |
| "loss": 2.9198175048828126, | |
| "step": 37100 | |
| }, | |
| { | |
| "epoch": 1.5652650206454874, | |
| "grad_norm": 0.8109549283981323, | |
| "learning_rate": 5.665847372190955e-06, | |
| "loss": 2.9246038818359374, | |
| "step": 37150 | |
| }, | |
| { | |
| "epoch": 1.5673717030420495, | |
| "grad_norm": 0.7990185022354126, | |
| "learning_rate": 5.613283102169406e-06, | |
| "loss": 2.9372125244140626, | |
| "step": 37200 | |
| }, | |
| { | |
| "epoch": 1.5694783854386114, | |
| "grad_norm": 0.790890097618103, | |
| "learning_rate": 5.56093294657016e-06, | |
| "loss": 2.933094482421875, | |
| "step": 37250 | |
| }, | |
| { | |
| "epoch": 1.5715850678351733, | |
| "grad_norm": 0.8145341277122498, | |
| "learning_rate": 5.508797483568665e-06, | |
| "loss": 2.9228143310546875, | |
| "step": 37300 | |
| }, | |
| { | |
| "epoch": 1.5736917502317351, | |
| "grad_norm": 0.7912366986274719, | |
| "learning_rate": 5.4568772889691836e-06, | |
| "loss": 2.929857177734375, | |
| "step": 37350 | |
| }, | |
| { | |
| "epoch": 1.575798432628297, | |
| "grad_norm": 0.8049252033233643, | |
| "learning_rate": 5.405172936198502e-06, | |
| "loss": 2.94720947265625, | |
| "step": 37400 | |
| }, | |
| { | |
| "epoch": 1.5779051150248589, | |
| "grad_norm": 0.7895810604095459, | |
| "learning_rate": 5.353684996299541e-06, | |
| "loss": 2.940379638671875, | |
| "step": 37450 | |
| }, | |
| { | |
| "epoch": 1.5800117974214207, | |
| "grad_norm": 0.805823802947998, | |
| "learning_rate": 5.302414037925077e-06, | |
| "loss": 2.916343994140625, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 1.5821184798179826, | |
| "grad_norm": 0.8166496157646179, | |
| "learning_rate": 5.251360627331478e-06, | |
| "loss": 2.93001220703125, | |
| "step": 37550 | |
| }, | |
| { | |
| "epoch": 1.5842251622145445, | |
| "grad_norm": 0.81240314245224, | |
| "learning_rate": 5.200525328372405e-06, | |
| "loss": 2.906513671875, | |
| "step": 37600 | |
| }, | |
| { | |
| "epoch": 1.5863318446111063, | |
| "grad_norm": 0.8053615689277649, | |
| "learning_rate": 5.149908702492631e-06, | |
| "loss": 2.9357623291015624, | |
| "step": 37650 | |
| }, | |
| { | |
| "epoch": 1.5884385270076682, | |
| "grad_norm": 0.802976131439209, | |
| "learning_rate": 5.0995113087217966e-06, | |
| "loss": 2.9226223754882814, | |
| "step": 37700 | |
| }, | |
| { | |
| "epoch": 1.59054520940423, | |
| "grad_norm": 0.7880696058273315, | |
| "learning_rate": 5.049333703668285e-06, | |
| "loss": 2.92947265625, | |
| "step": 37750 | |
| }, | |
| { | |
| "epoch": 1.5926518918007921, | |
| "grad_norm": 0.8215526342391968, | |
| "learning_rate": 4.999376441513015e-06, | |
| "loss": 2.9145855712890625, | |
| "step": 37800 | |
| }, | |
| { | |
| "epoch": 1.594758574197354, | |
| "grad_norm": 0.7832110524177551, | |
| "learning_rate": 4.949640074003389e-06, | |
| "loss": 2.9458853149414064, | |
| "step": 37850 | |
| }, | |
| { | |
| "epoch": 1.5968652565939159, | |
| "grad_norm": 0.8030633330345154, | |
| "learning_rate": 4.9001251504471325e-06, | |
| "loss": 2.918009948730469, | |
| "step": 37900 | |
| }, | |
| { | |
| "epoch": 1.598971938990478, | |
| "grad_norm": 0.80305415391922, | |
| "learning_rate": 4.850832217706267e-06, | |
| "loss": 2.928319396972656, | |
| "step": 37950 | |
| }, | |
| { | |
| "epoch": 1.6010786213870398, | |
| "grad_norm": 0.7899791598320007, | |
| "learning_rate": 4.801761820191072e-06, | |
| "loss": 2.9024557495117187, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 1.6010786213870398, | |
| "eval_loss": 2.9146671295166016, | |
| "eval_runtime": 110.9544, | |
| "eval_samples_per_second": 138.282, | |
| "eval_steps_per_second": 4.326, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 1.6031853037836017, | |
| "grad_norm": 0.8066175580024719, | |
| "learning_rate": 4.75291449985405e-06, | |
| "loss": 2.9117437744140626, | |
| "step": 38050 | |
| }, | |
| { | |
| "epoch": 1.6052919861801636, | |
| "grad_norm": 0.7874099612236023, | |
| "learning_rate": 4.704290796183963e-06, | |
| "loss": 2.945263671875, | |
| "step": 38100 | |
| }, | |
| { | |
| "epoch": 1.6073986685767254, | |
| "grad_norm": 0.8120174407958984, | |
| "learning_rate": 4.655891246199845e-06, | |
| "loss": 2.9389337158203124, | |
| "step": 38150 | |
| }, | |
| { | |
| "epoch": 1.6095053509732873, | |
| "grad_norm": 0.7942776679992676, | |
| "learning_rate": 4.6077163844451184e-06, | |
| "loss": 2.943665771484375, | |
| "step": 38200 | |
| }, | |
| { | |
| "epoch": 1.6116120333698492, | |
| "grad_norm": 0.797516405582428, | |
| "learning_rate": 4.559766742981639e-06, | |
| "loss": 2.928199462890625, | |
| "step": 38250 | |
| }, | |
| { | |
| "epoch": 1.613718715766411, | |
| "grad_norm": 0.7932865023612976, | |
| "learning_rate": 4.512042851383835e-06, | |
| "loss": 2.9346612548828124, | |
| "step": 38300 | |
| }, | |
| { | |
| "epoch": 1.6158253981629729, | |
| "grad_norm": 0.7972375154495239, | |
| "learning_rate": 4.465492968114715e-06, | |
| "loss": 2.9239166259765623, | |
| "step": 38350 | |
| }, | |
| { | |
| "epoch": 1.6179320805595347, | |
| "grad_norm": 0.790519118309021, | |
| "learning_rate": 4.418217613837431e-06, | |
| "loss": 2.920374755859375, | |
| "step": 38400 | |
| }, | |
| { | |
| "epoch": 1.6200387629560966, | |
| "grad_norm": 0.7743920087814331, | |
| "learning_rate": 4.3711695727493225e-06, | |
| "loss": 2.9266815185546875, | |
| "step": 38450 | |
| }, | |
| { | |
| "epoch": 1.6221454453526585, | |
| "grad_norm": 0.7905875444412231, | |
| "learning_rate": 4.324349364467223e-06, | |
| "loss": 2.918923645019531, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 1.6242521277492206, | |
| "grad_norm": 0.795048713684082, | |
| "learning_rate": 4.277757506091659e-06, | |
| "loss": 2.91831787109375, | |
| "step": 38550 | |
| }, | |
| { | |
| "epoch": 1.6263588101457824, | |
| "grad_norm": 0.8047122955322266, | |
| "learning_rate": 4.231394512201204e-06, | |
| "loss": 2.917353515625, | |
| "step": 38600 | |
| }, | |
| { | |
| "epoch": 1.6284654925423443, | |
| "grad_norm": 0.7855550646781921, | |
| "learning_rate": 4.185260894846732e-06, | |
| "loss": 2.9392071533203126, | |
| "step": 38650 | |
| }, | |
| { | |
| "epoch": 1.6305721749389062, | |
| "grad_norm": 0.8067252039909363, | |
| "learning_rate": 4.139357163545829e-06, | |
| "loss": 2.926513671875, | |
| "step": 38700 | |
| }, | |
| { | |
| "epoch": 1.6326788573354682, | |
| "grad_norm": 0.8326886296272278, | |
| "learning_rate": 4.093683825277098e-06, | |
| "loss": 2.9286224365234377, | |
| "step": 38750 | |
| }, | |
| { | |
| "epoch": 1.6347855397320301, | |
| "grad_norm": 0.811650812625885, | |
| "learning_rate": 4.048241384474613e-06, | |
| "loss": 2.90252685546875, | |
| "step": 38800 | |
| }, | |
| { | |
| "epoch": 1.636892222128592, | |
| "grad_norm": 0.8090134263038635, | |
| "learning_rate": 4.0030303430223186e-06, | |
| "loss": 2.906116943359375, | |
| "step": 38850 | |
| }, | |
| { | |
| "epoch": 1.6389989045251538, | |
| "grad_norm": 0.7756823897361755, | |
| "learning_rate": 3.958051200248483e-06, | |
| "loss": 2.9276080322265625, | |
| "step": 38900 | |
| }, | |
| { | |
| "epoch": 1.6411055869217157, | |
| "grad_norm": 0.7710340023040771, | |
| "learning_rate": 3.9133044529202275e-06, | |
| "loss": 2.9258453369140627, | |
| "step": 38950 | |
| }, | |
| { | |
| "epoch": 1.6432122693182776, | |
| "grad_norm": 0.7959493398666382, | |
| "learning_rate": 3.868790595237973e-06, | |
| "loss": 2.937237548828125, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 1.6432122693182776, | |
| "eval_loss": 2.913990020751953, | |
| "eval_runtime": 109.3408, | |
| "eval_samples_per_second": 140.323, | |
| "eval_steps_per_second": 4.39, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 1.6453189517148394, | |
| "grad_norm": 0.7863781452178955, | |
| "learning_rate": 3.824510118830047e-06, | |
| "loss": 2.9218255615234376, | |
| "step": 39050 | |
| }, | |
| { | |
| "epoch": 1.6474256341114013, | |
| "grad_norm": 0.7998809814453125, | |
| "learning_rate": 3.7804635127471926e-06, | |
| "loss": 2.937727355957031, | |
| "step": 39100 | |
| }, | |
| { | |
| "epoch": 1.6495323165079632, | |
| "grad_norm": 0.7936866283416748, | |
| "learning_rate": 3.7366512634572306e-06, | |
| "loss": 2.952947998046875, | |
| "step": 39150 | |
| }, | |
| { | |
| "epoch": 1.651638998904525, | |
| "grad_norm": 0.7903639078140259, | |
| "learning_rate": 3.693073854839629e-06, | |
| "loss": 2.93330078125, | |
| "step": 39200 | |
| }, | |
| { | |
| "epoch": 1.653745681301087, | |
| "grad_norm": 0.7985586524009705, | |
| "learning_rate": 3.6497317681802008e-06, | |
| "loss": 2.925899658203125, | |
| "step": 39250 | |
| }, | |
| { | |
| "epoch": 1.655852363697649, | |
| "grad_norm": 0.7784944176673889, | |
| "learning_rate": 3.6066254821657514e-06, | |
| "loss": 2.956893005371094, | |
| "step": 39300 | |
| }, | |
| { | |
| "epoch": 1.6579590460942109, | |
| "grad_norm": 0.8026953339576721, | |
| "learning_rate": 3.563755472878838e-06, | |
| "loss": 2.928978271484375, | |
| "step": 39350 | |
| }, | |
| { | |
| "epoch": 1.6600657284907727, | |
| "grad_norm": 0.788847804069519, | |
| "learning_rate": 3.521122213792477e-06, | |
| "loss": 2.9353936767578124, | |
| "step": 39400 | |
| }, | |
| { | |
| "epoch": 1.6621724108873346, | |
| "grad_norm": 0.8241669535636902, | |
| "learning_rate": 3.4787261757649097e-06, | |
| "loss": 2.934281311035156, | |
| "step": 39450 | |
| }, | |
| { | |
| "epoch": 1.6642790932838967, | |
| "grad_norm": 0.7957674264907837, | |
| "learning_rate": 3.4365678270344396e-06, | |
| "loss": 2.9192974853515623, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 1.6663857756804585, | |
| "grad_norm": 0.7936614751815796, | |
| "learning_rate": 3.394647633214215e-06, | |
| "loss": 2.9286676025390626, | |
| "step": 39550 | |
| }, | |
| { | |
| "epoch": 1.6684924580770204, | |
| "grad_norm": 0.7859033346176147, | |
| "learning_rate": 3.3529660572871267e-06, | |
| "loss": 2.9341552734375, | |
| "step": 39600 | |
| }, | |
| { | |
| "epoch": 1.6705991404735823, | |
| "grad_norm": 0.7885980010032654, | |
| "learning_rate": 3.3115235596006654e-06, | |
| "loss": 2.912452392578125, | |
| "step": 39650 | |
| }, | |
| { | |
| "epoch": 1.6727058228701441, | |
| "grad_norm": 0.7956350445747375, | |
| "learning_rate": 3.2703205978618534e-06, | |
| "loss": 2.9545343017578123, | |
| "step": 39700 | |
| }, | |
| { | |
| "epoch": 1.674812505266706, | |
| "grad_norm": 0.832351565361023, | |
| "learning_rate": 3.229357627132176e-06, | |
| "loss": 2.9165264892578127, | |
| "step": 39750 | |
| }, | |
| { | |
| "epoch": 1.6769191876632679, | |
| "grad_norm": 0.7851858139038086, | |
| "learning_rate": 3.1886350998225844e-06, | |
| "loss": 2.921520080566406, | |
| "step": 39800 | |
| }, | |
| { | |
| "epoch": 1.6790258700598297, | |
| "grad_norm": 0.8115505576133728, | |
| "learning_rate": 3.148153465688469e-06, | |
| "loss": 2.9139031982421875, | |
| "step": 39850 | |
| }, | |
| { | |
| "epoch": 1.6811325524563916, | |
| "grad_norm": 0.7883580923080444, | |
| "learning_rate": 3.107913171824692e-06, | |
| "loss": 2.9267724609375, | |
| "step": 39900 | |
| }, | |
| { | |
| "epoch": 1.6832392348529535, | |
| "grad_norm": 0.8143483996391296, | |
| "learning_rate": 3.0679146626606784e-06, | |
| "loss": 2.9394415283203124, | |
| "step": 39950 | |
| }, | |
| { | |
| "epoch": 1.6853459172495153, | |
| "grad_norm": 0.811782717704773, | |
| "learning_rate": 3.0281583799554786e-06, | |
| "loss": 2.9455914306640625, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 1.6853459172495153, | |
| "eval_loss": 2.9129562377929688, | |
| "eval_runtime": 109.4554, | |
| "eval_samples_per_second": 140.176, | |
| "eval_steps_per_second": 4.385, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 1.6874525996460774, | |
| "grad_norm": 0.7961781620979309, | |
| "learning_rate": 2.9886447627929e-06, | |
| "loss": 2.9425677490234374, | |
| "step": 40050 | |
| }, | |
| { | |
| "epoch": 1.6895592820426393, | |
| "grad_norm": 0.80986487865448, | |
| "learning_rate": 2.9493742475766557e-06, | |
| "loss": 2.95297119140625, | |
| "step": 40100 | |
| }, | |
| { | |
| "epoch": 1.6916659644392011, | |
| "grad_norm": 0.8379704356193542, | |
| "learning_rate": 2.9103472680255595e-06, | |
| "loss": 2.94047119140625, | |
| "step": 40150 | |
| }, | |
| { | |
| "epoch": 1.693772646835763, | |
| "grad_norm": 0.8274521827697754, | |
| "learning_rate": 2.8715642551687016e-06, | |
| "loss": 2.9236444091796874, | |
| "step": 40200 | |
| }, | |
| { | |
| "epoch": 1.695879329232325, | |
| "grad_norm": 0.8155441284179688, | |
| "learning_rate": 2.8330256373407304e-06, | |
| "loss": 2.91578369140625, | |
| "step": 40250 | |
| }, | |
| { | |
| "epoch": 1.697986011628887, | |
| "grad_norm": 0.7939282655715942, | |
| "learning_rate": 2.794731840177095e-06, | |
| "loss": 2.9297991943359376, | |
| "step": 40300 | |
| }, | |
| { | |
| "epoch": 1.7000926940254488, | |
| "grad_norm": 0.8066977858543396, | |
| "learning_rate": 2.756683286609335e-06, | |
| "loss": 2.928181457519531, | |
| "step": 40350 | |
| }, | |
| { | |
| "epoch": 1.7021993764220107, | |
| "grad_norm": 0.8126586079597473, | |
| "learning_rate": 2.7188803968604516e-06, | |
| "loss": 2.935025634765625, | |
| "step": 40400 | |
| }, | |
| { | |
| "epoch": 1.7043060588185726, | |
| "grad_norm": 0.8066821694374084, | |
| "learning_rate": 2.682072310324235e-06, | |
| "loss": 2.93519775390625, | |
| "step": 40450 | |
| }, | |
| { | |
| "epoch": 1.7064127412151344, | |
| "grad_norm": 0.811872661113739, | |
| "learning_rate": 2.644757064054837e-06, | |
| "loss": 2.939276123046875, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 1.7085194236116963, | |
| "grad_norm": 0.7744365930557251, | |
| "learning_rate": 2.6076887177609283e-06, | |
| "loss": 2.94591796875, | |
| "step": 40550 | |
| }, | |
| { | |
| "epoch": 1.7106261060082582, | |
| "grad_norm": 0.7963656187057495, | |
| "learning_rate": 2.570867680839703e-06, | |
| "loss": 2.9264068603515625, | |
| "step": 40600 | |
| }, | |
| { | |
| "epoch": 1.71273278840482, | |
| "grad_norm": 0.8092783093452454, | |
| "learning_rate": 2.53429435995699e-06, | |
| "loss": 2.9164804077148436, | |
| "step": 40650 | |
| }, | |
| { | |
| "epoch": 1.7148394708013819, | |
| "grad_norm": 0.8073092699050903, | |
| "learning_rate": 2.497969159042718e-06, | |
| "loss": 2.9387576293945314, | |
| "step": 40700 | |
| }, | |
| { | |
| "epoch": 1.7169461531979437, | |
| "grad_norm": 0.8129700422286987, | |
| "learning_rate": 2.4618924792865093e-06, | |
| "loss": 2.9286749267578127, | |
| "step": 40750 | |
| }, | |
| { | |
| "epoch": 1.7190528355945056, | |
| "grad_norm": 0.7984380722045898, | |
| "learning_rate": 2.426064719133203e-06, | |
| "loss": 2.9281085205078123, | |
| "step": 40800 | |
| }, | |
| { | |
| "epoch": 1.7211595179910677, | |
| "grad_norm": 0.7834457159042358, | |
| "learning_rate": 2.3904862742784835e-06, | |
| "loss": 2.936934509277344, | |
| "step": 40850 | |
| }, | |
| { | |
| "epoch": 1.7232662003876296, | |
| "grad_norm": 0.8101183176040649, | |
| "learning_rate": 2.355157537664507e-06, | |
| "loss": 2.9396533203125, | |
| "step": 40900 | |
| }, | |
| { | |
| "epoch": 1.7253728827841914, | |
| "grad_norm": 0.7983947992324829, | |
| "learning_rate": 2.3200788994755364e-06, | |
| "loss": 2.9327899169921876, | |
| "step": 40950 | |
| }, | |
| { | |
| "epoch": 1.7274795651807533, | |
| "grad_norm": 0.8090130686759949, | |
| "learning_rate": 2.285250747133663e-06, | |
| "loss": 2.9282513427734376, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 1.7274795651807533, | |
| "eval_loss": 2.9124667644500732, | |
| "eval_runtime": 109.1218, | |
| "eval_samples_per_second": 140.604, | |
| "eval_steps_per_second": 4.399, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 1.7295862475773154, | |
| "grad_norm": 0.8032718300819397, | |
| "learning_rate": 2.250673465294509e-06, | |
| "loss": 2.913149719238281, | |
| "step": 41050 | |
| }, | |
| { | |
| "epoch": 1.7316929299738772, | |
| "grad_norm": 0.8181055784225464, | |
| "learning_rate": 2.2163474358429944e-06, | |
| "loss": 2.9339990234375, | |
| "step": 41100 | |
| }, | |
| { | |
| "epoch": 1.7337996123704391, | |
| "grad_norm": 0.8013983368873596, | |
| "learning_rate": 2.182273037889099e-06, | |
| "loss": 2.9100967407226563, | |
| "step": 41150 | |
| }, | |
| { | |
| "epoch": 1.735906294767001, | |
| "grad_norm": 0.836780846118927, | |
| "learning_rate": 2.1484506477637002e-06, | |
| "loss": 2.925872802734375, | |
| "step": 41200 | |
| }, | |
| { | |
| "epoch": 1.7380129771635628, | |
| "grad_norm": 0.7955629229545593, | |
| "learning_rate": 2.114880639014388e-06, | |
| "loss": 2.9159527587890626, | |
| "step": 41250 | |
| }, | |
| { | |
| "epoch": 1.7401196595601247, | |
| "grad_norm": 0.8006013035774231, | |
| "learning_rate": 2.081563382401366e-06, | |
| "loss": 2.90957763671875, | |
| "step": 41300 | |
| }, | |
| { | |
| "epoch": 1.7422263419566866, | |
| "grad_norm": 0.8277612924575806, | |
| "learning_rate": 2.048499245893343e-06, | |
| "loss": 2.928201904296875, | |
| "step": 41350 | |
| }, | |
| { | |
| "epoch": 1.7443330243532484, | |
| "grad_norm": 0.7843561768531799, | |
| "learning_rate": 2.0156885946634703e-06, | |
| "loss": 2.9217242431640624, | |
| "step": 41400 | |
| }, | |
| { | |
| "epoch": 1.7464397067498103, | |
| "grad_norm": 0.7751513123512268, | |
| "learning_rate": 1.9831317910853046e-06, | |
| "loss": 2.9333505249023437, | |
| "step": 41450 | |
| }, | |
| { | |
| "epoch": 1.7485463891463722, | |
| "grad_norm": 0.7870394587516785, | |
| "learning_rate": 1.950829194728812e-06, | |
| "loss": 2.9505987548828125, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 1.750653071542934, | |
| "grad_norm": 0.7850804924964905, | |
| "learning_rate": 1.918781162356409e-06, | |
| "loss": 2.90233154296875, | |
| "step": 41550 | |
| }, | |
| { | |
| "epoch": 1.7527597539394961, | |
| "grad_norm": 0.794102668762207, | |
| "learning_rate": 1.8869880479189834e-06, | |
| "loss": 2.94618408203125, | |
| "step": 41600 | |
| }, | |
| { | |
| "epoch": 1.754866436336058, | |
| "grad_norm": 0.8069368600845337, | |
| "learning_rate": 1.8554502025520394e-06, | |
| "loss": 2.9230780029296874, | |
| "step": 41650 | |
| }, | |
| { | |
| "epoch": 1.7569731187326199, | |
| "grad_norm": 0.7900083065032959, | |
| "learning_rate": 1.8241679745717639e-06, | |
| "loss": 2.9422222900390627, | |
| "step": 41700 | |
| }, | |
| { | |
| "epoch": 1.7590798011291817, | |
| "grad_norm": 0.828949511051178, | |
| "learning_rate": 1.7931417094712271e-06, | |
| "loss": 2.905482177734375, | |
| "step": 41750 | |
| }, | |
| { | |
| "epoch": 1.7611864835257438, | |
| "grad_norm": 0.8181668519973755, | |
| "learning_rate": 1.762371749916536e-06, | |
| "loss": 2.9306365966796877, | |
| "step": 41800 | |
| }, | |
| { | |
| "epoch": 1.7632931659223057, | |
| "grad_norm": 0.7998356223106384, | |
| "learning_rate": 1.7318584357430529e-06, | |
| "loss": 2.905430908203125, | |
| "step": 41850 | |
| }, | |
| { | |
| "epoch": 1.7653998483188675, | |
| "grad_norm": 0.7857435345649719, | |
| "learning_rate": 1.7016021039516689e-06, | |
| "loss": 2.9229693603515625, | |
| "step": 41900 | |
| }, | |
| { | |
| "epoch": 1.7675065307154294, | |
| "grad_norm": 0.8163794279098511, | |
| "learning_rate": 1.6716030887050372e-06, | |
| "loss": 2.935548095703125, | |
| "step": 41950 | |
| }, | |
| { | |
| "epoch": 1.7696132131119913, | |
| "grad_norm": 0.8158469796180725, | |
| "learning_rate": 1.6418617213239313e-06, | |
| "loss": 2.9157159423828123, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 1.7696132131119913, | |
| "eval_loss": 2.9121968746185303, | |
| "eval_runtime": 108.9624, | |
| "eval_samples_per_second": 140.81, | |
| "eval_steps_per_second": 4.405, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 1.7717198955085531, | |
| "grad_norm": 0.7961393594741821, | |
| "learning_rate": 1.61237833028354e-06, | |
| "loss": 2.9196566772460937, | |
| "step": 42050 | |
| }, | |
| { | |
| "epoch": 1.773826577905115, | |
| "grad_norm": 0.7808844447135925, | |
| "learning_rate": 1.5831532412098838e-06, | |
| "loss": 2.926371765136719, | |
| "step": 42100 | |
| }, | |
| { | |
| "epoch": 1.7759332603016769, | |
| "grad_norm": 0.7773975729942322, | |
| "learning_rate": 1.5541867768761736e-06, | |
| "loss": 2.932283935546875, | |
| "step": 42150 | |
| }, | |
| { | |
| "epoch": 1.7780399426982387, | |
| "grad_norm": 0.7964804768562317, | |
| "learning_rate": 1.525479257199286e-06, | |
| "loss": 2.9389453125, | |
| "step": 42200 | |
| }, | |
| { | |
| "epoch": 1.7801466250948006, | |
| "grad_norm": 0.8084786534309387, | |
| "learning_rate": 1.4970309992362014e-06, | |
| "loss": 2.91684326171875, | |
| "step": 42250 | |
| }, | |
| { | |
| "epoch": 1.7822533074913625, | |
| "grad_norm": 0.7834380269050598, | |
| "learning_rate": 1.4688423171805132e-06, | |
| "loss": 2.9197119140625, | |
| "step": 42300 | |
| }, | |
| { | |
| "epoch": 1.7843599898879245, | |
| "grad_norm": 0.8210534453392029, | |
| "learning_rate": 1.4409135223589694e-06, | |
| "loss": 2.932439270019531, | |
| "step": 42350 | |
| }, | |
| { | |
| "epoch": 1.7864666722844864, | |
| "grad_norm": 0.8303399682044983, | |
| "learning_rate": 1.4132449232280103e-06, | |
| "loss": 2.9323825073242187, | |
| "step": 42400 | |
| }, | |
| { | |
| "epoch": 1.7885733546810483, | |
| "grad_norm": 0.7951374053955078, | |
| "learning_rate": 1.3858368253703791e-06, | |
| "loss": 2.933565673828125, | |
| "step": 42450 | |
| }, | |
| { | |
| "epoch": 1.7906800370776101, | |
| "grad_norm": 0.7852106094360352, | |
| "learning_rate": 1.3586895314917414e-06, | |
| "loss": 2.9080514526367187, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 1.7927867194741722, | |
| "grad_norm": 0.7875521183013916, | |
| "learning_rate": 1.3318033414173425e-06, | |
| "loss": 2.9189569091796876, | |
| "step": 42550 | |
| }, | |
| { | |
| "epoch": 1.794893401870734, | |
| "grad_norm": 0.797749936580658, | |
| "learning_rate": 1.3051785520886983e-06, | |
| "loss": 2.929200439453125, | |
| "step": 42600 | |
| }, | |
| { | |
| "epoch": 1.797000084267296, | |
| "grad_norm": 0.8121727108955383, | |
| "learning_rate": 1.2788154575603062e-06, | |
| "loss": 2.9154443359375, | |
| "step": 42650 | |
| }, | |
| { | |
| "epoch": 1.7991067666638578, | |
| "grad_norm": 0.799733579158783, | |
| "learning_rate": 1.2532338018359263e-06, | |
| "loss": 2.931458740234375, | |
| "step": 42700 | |
| }, | |
| { | |
| "epoch": 1.8012134490604197, | |
| "grad_norm": 0.7826681137084961, | |
| "learning_rate": 1.227389719216271e-06, | |
| "loss": 2.9412982177734377, | |
| "step": 42750 | |
| }, | |
| { | |
| "epoch": 1.8033201314569816, | |
| "grad_norm": 0.7729964852333069, | |
| "learning_rate": 1.2018081905269152e-06, | |
| "loss": 2.9225433349609373, | |
| "step": 42800 | |
| }, | |
| { | |
| "epoch": 1.8054268138535434, | |
| "grad_norm": 0.8305932879447937, | |
| "learning_rate": 1.17648949830019e-06, | |
| "loss": 2.954083251953125, | |
| "step": 42850 | |
| }, | |
| { | |
| "epoch": 1.8075334962501053, | |
| "grad_norm": 0.8129289150238037, | |
| "learning_rate": 1.1514339221655512e-06, | |
| "loss": 2.933909606933594, | |
| "step": 42900 | |
| }, | |
| { | |
| "epoch": 1.8096401786466672, | |
| "grad_norm": 0.7922786474227905, | |
| "learning_rate": 1.1266417388465067e-06, | |
| "loss": 2.9360870361328124, | |
| "step": 42950 | |
| }, | |
| { | |
| "epoch": 1.811746861043229, | |
| "grad_norm": 0.8204851150512695, | |
| "learning_rate": 1.1021132221575387e-06, | |
| "loss": 2.9386032104492186, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 1.811746861043229, | |
| "eval_loss": 2.9120566844940186, | |
| "eval_runtime": 127.9465, | |
| "eval_samples_per_second": 119.917, | |
| "eval_steps_per_second": 3.752, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 1.8138535434397909, | |
| "grad_norm": 0.7862714529037476, | |
| "learning_rate": 1.0778486430011113e-06, | |
| "loss": 2.905498046875, | |
| "step": 43050 | |
| }, | |
| { | |
| "epoch": 1.8159602258363527, | |
| "grad_norm": 0.8294594883918762, | |
| "learning_rate": 1.0538482693646435e-06, | |
| "loss": 2.9325042724609376, | |
| "step": 43100 | |
| }, | |
| { | |
| "epoch": 1.8180669082329148, | |
| "grad_norm": 0.8204466104507446, | |
| "learning_rate": 1.0301123663175705e-06, | |
| "loss": 2.927686767578125, | |
| "step": 43150 | |
| }, | |
| { | |
| "epoch": 1.8201735906294767, | |
| "grad_norm": 0.8042780756950378, | |
| "learning_rate": 1.0066411960084205e-06, | |
| "loss": 2.9334552001953127, | |
| "step": 43200 | |
| }, | |
| { | |
| "epoch": 1.8222802730260386, | |
| "grad_norm": 0.8087170124053955, | |
| "learning_rate": 9.834350176618962e-07, | |
| "loss": 2.9226800537109376, | |
| "step": 43250 | |
| }, | |
| { | |
| "epoch": 1.8243869554226007, | |
| "grad_norm": 0.7973117828369141, | |
| "learning_rate": 9.604940875760415e-07, | |
| "loss": 2.9171844482421876, | |
| "step": 43300 | |
| }, | |
| { | |
| "epoch": 1.8264936378191625, | |
| "grad_norm": 0.8308363556861877, | |
| "learning_rate": 9.378186591193827e-07, | |
| "loss": 2.9369036865234377, | |
| "step": 43350 | |
| }, | |
| { | |
| "epoch": 1.8286003202157244, | |
| "grad_norm": 0.7975155115127563, | |
| "learning_rate": 9.154089827281498e-07, | |
| "loss": 2.933402099609375, | |
| "step": 43400 | |
| }, | |
| { | |
| "epoch": 1.8307070026122862, | |
| "grad_norm": 0.7863073945045471, | |
| "learning_rate": 8.932653059035013e-07, | |
| "loss": 2.9464892578125, | |
| "step": 43450 | |
| }, | |
| { | |
| "epoch": 1.8328136850088481, | |
| "grad_norm": 0.7683363556861877, | |
| "learning_rate": 8.71387873208801e-07, | |
| "loss": 2.9376263427734375, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 1.83492036740541, | |
| "grad_norm": 0.8093059062957764, | |
| "learning_rate": 8.497769262668959e-07, | |
| "loss": 2.9380569458007812, | |
| "step": 43550 | |
| }, | |
| { | |
| "epoch": 1.8370270498019718, | |
| "grad_norm": 0.8482558131217957, | |
| "learning_rate": 8.284327037574646e-07, | |
| "loss": 2.9273577880859376, | |
| "step": 43600 | |
| }, | |
| { | |
| "epoch": 1.8391337321985337, | |
| "grad_norm": 0.7909052968025208, | |
| "learning_rate": 8.073554414143897e-07, | |
| "loss": 2.918182067871094, | |
| "step": 43650 | |
| }, | |
| { | |
| "epoch": 1.8412404145950956, | |
| "grad_norm": 0.7663418054580688, | |
| "learning_rate": 7.865453720231286e-07, | |
| "loss": 2.919114685058594, | |
| "step": 43700 | |
| }, | |
| { | |
| "epoch": 1.8433470969916574, | |
| "grad_norm": 0.7878078818321228, | |
| "learning_rate": 7.660027254181634e-07, | |
| "loss": 2.90847900390625, | |
| "step": 43750 | |
| }, | |
| { | |
| "epoch": 1.8454537793882193, | |
| "grad_norm": 0.8218978047370911, | |
| "learning_rate": 7.45727728480447e-07, | |
| "loss": 2.92931396484375, | |
| "step": 43800 | |
| }, | |
| { | |
| "epoch": 1.8475604617847812, | |
| "grad_norm": 0.781768798828125, | |
| "learning_rate": 7.257206051349163e-07, | |
| "loss": 2.9267474365234376, | |
| "step": 43850 | |
| }, | |
| { | |
| "epoch": 1.8496671441813433, | |
| "grad_norm": 0.8082556128501892, | |
| "learning_rate": 7.059815763479999e-07, | |
| "loss": 2.901339111328125, | |
| "step": 43900 | |
| }, | |
| { | |
| "epoch": 1.8517738265779051, | |
| "grad_norm": 0.8154715299606323, | |
| "learning_rate": 6.865108601251918e-07, | |
| "loss": 2.9370611572265624, | |
| "step": 43950 | |
| }, | |
| { | |
| "epoch": 1.853880508974467, | |
| "grad_norm": 0.8003249168395996, | |
| "learning_rate": 6.673086715086374e-07, | |
| "loss": 2.9026998901367187, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 1.853880508974467, | |
| "eval_loss": 2.911797523498535, | |
| "eval_runtime": 128.0728, | |
| "eval_samples_per_second": 119.799, | |
| "eval_steps_per_second": 3.748, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 1.8559871913710289, | |
| "grad_norm": 0.7780888676643372, | |
| "learning_rate": 6.483752225747486e-07, | |
| "loss": 2.9130596923828125, | |
| "step": 44050 | |
| }, | |
| { | |
| "epoch": 1.858093873767591, | |
| "grad_norm": 0.7948802709579468, | |
| "learning_rate": 6.297107224318893e-07, | |
| "loss": 2.9231292724609377, | |
| "step": 44100 | |
| }, | |
| { | |
| "epoch": 1.8602005561641528, | |
| "grad_norm": 0.7986658215522766, | |
| "learning_rate": 6.113153772180358e-07, | |
| "loss": 2.9504754638671873, | |
| "step": 44150 | |
| }, | |
| { | |
| "epoch": 1.8623072385607147, | |
| "grad_norm": 0.8019037842750549, | |
| "learning_rate": 5.93189390098517e-07, | |
| "loss": 2.9416473388671873, | |
| "step": 44200 | |
| }, | |
| { | |
| "epoch": 1.8644139209572765, | |
| "grad_norm": 0.7774601578712463, | |
| "learning_rate": 5.753329612637664e-07, | |
| "loss": 2.936625671386719, | |
| "step": 44250 | |
| }, | |
| { | |
| "epoch": 1.8665206033538384, | |
| "grad_norm": 0.7865827679634094, | |
| "learning_rate": 5.577462879271162e-07, | |
| "loss": 2.9374444580078123, | |
| "step": 44300 | |
| }, | |
| { | |
| "epoch": 1.8686272857504003, | |
| "grad_norm": 0.7820749282836914, | |
| "learning_rate": 5.404295643226004e-07, | |
| "loss": 2.9128753662109377, | |
| "step": 44350 | |
| }, | |
| { | |
| "epoch": 1.8707339681469621, | |
| "grad_norm": 0.7874165177345276, | |
| "learning_rate": 5.233829817028413e-07, | |
| "loss": 2.908509521484375, | |
| "step": 44400 | |
| }, | |
| { | |
| "epoch": 1.872840650543524, | |
| "grad_norm": 0.792330801486969, | |
| "learning_rate": 5.06606728336903e-07, | |
| "loss": 2.923705749511719, | |
| "step": 44450 | |
| }, | |
| { | |
| "epoch": 1.8749473329400859, | |
| "grad_norm": 0.8028748631477356, | |
| "learning_rate": 4.901009895082382e-07, | |
| "loss": 2.9056796264648437, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 1.8770540153366477, | |
| "grad_norm": 0.7632665038108826, | |
| "learning_rate": 4.738659475126306e-07, | |
| "loss": 2.93132568359375, | |
| "step": 44550 | |
| }, | |
| { | |
| "epoch": 1.8791606977332096, | |
| "grad_norm": 0.7705265879631042, | |
| "learning_rate": 4.579017816561837e-07, | |
| "loss": 2.91453857421875, | |
| "step": 44600 | |
| }, | |
| { | |
| "epoch": 1.8812673801297717, | |
| "grad_norm": 0.781696081161499, | |
| "learning_rate": 4.4220866825333826e-07, | |
| "loss": 2.9119244384765626, | |
| "step": 44650 | |
| }, | |
| { | |
| "epoch": 1.8833740625263335, | |
| "grad_norm": 0.790869414806366, | |
| "learning_rate": 4.267867806249215e-07, | |
| "loss": 2.9221722412109377, | |
| "step": 44700 | |
| }, | |
| { | |
| "epoch": 1.8854807449228954, | |
| "grad_norm": 0.7911215424537659, | |
| "learning_rate": 4.116362890962511e-07, | |
| "loss": 2.935125732421875, | |
| "step": 44750 | |
| }, | |
| { | |
| "epoch": 1.8875874273194573, | |
| "grad_norm": 0.7805041670799255, | |
| "learning_rate": 3.967573609952258e-07, | |
| "loss": 2.9434979248046873, | |
| "step": 44800 | |
| }, | |
| { | |
| "epoch": 1.8896941097160194, | |
| "grad_norm": 0.774147629737854, | |
| "learning_rate": 3.8243964067696815e-07, | |
| "loss": 2.9308935546875, | |
| "step": 44850 | |
| }, | |
| { | |
| "epoch": 1.8918007921125812, | |
| "grad_norm": 0.8052771687507629, | |
| "learning_rate": 3.6809889007286825e-07, | |
| "loss": 2.93390380859375, | |
| "step": 44900 | |
| }, | |
| { | |
| "epoch": 1.893907474509143, | |
| "grad_norm": 0.7773950695991516, | |
| "learning_rate": 3.540301837403392e-07, | |
| "loss": 2.91735595703125, | |
| "step": 44950 | |
| }, | |
| { | |
| "epoch": 1.896014156905705, | |
| "grad_norm": 0.8081435561180115, | |
| "learning_rate": 3.4023367705963707e-07, | |
| "loss": 2.9174813842773437, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 1.896014156905705, | |
| "eval_loss": 2.911701202392578, | |
| "eval_runtime": 119.8001, | |
| "eval_samples_per_second": 128.072, | |
| "eval_steps_per_second": 4.007, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 1.8981208393022668, | |
| "grad_norm": 0.79304438829422, | |
| "learning_rate": 3.2670952240472505e-07, | |
| "loss": 2.922022705078125, | |
| "step": 45050 | |
| }, | |
| { | |
| "epoch": 1.9002275216988287, | |
| "grad_norm": 0.796622097492218, | |
| "learning_rate": 3.134578691416201e-07, | |
| "loss": 2.9227664184570314, | |
| "step": 45100 | |
| }, | |
| { | |
| "epoch": 1.9023342040953906, | |
| "grad_norm": 0.7990376353263855, | |
| "learning_rate": 3.004788636267242e-07, | |
| "loss": 2.917940673828125, | |
| "step": 45150 | |
| }, | |
| { | |
| "epoch": 1.9044408864919524, | |
| "grad_norm": 0.8256587386131287, | |
| "learning_rate": 2.8777264920520354e-07, | |
| "loss": 2.922840576171875, | |
| "step": 45200 | |
| }, | |
| { | |
| "epoch": 1.9065475688885143, | |
| "grad_norm": 0.8110597133636475, | |
| "learning_rate": 2.7533936620942334e-07, | |
| "loss": 2.9242098999023436, | |
| "step": 45250 | |
| }, | |
| { | |
| "epoch": 1.9086542512850762, | |
| "grad_norm": 0.7913885116577148, | |
| "learning_rate": 2.631791519573823e-07, | |
| "loss": 2.9055206298828127, | |
| "step": 45300 | |
| }, | |
| { | |
| "epoch": 1.910760933681638, | |
| "grad_norm": 0.8255484700202942, | |
| "learning_rate": 2.512921407512053e-07, | |
| "loss": 2.9211932373046876, | |
| "step": 45350 | |
| }, | |
| { | |
| "epoch": 1.9128676160782, | |
| "grad_norm": 0.815882682800293, | |
| "learning_rate": 2.396784638756505e-07, | |
| "loss": 2.9394482421875, | |
| "step": 45400 | |
| }, | |
| { | |
| "epoch": 1.914974298474762, | |
| "grad_norm": 0.7824210524559021, | |
| "learning_rate": 2.283382495966796e-07, | |
| "loss": 2.9364703369140623, | |
| "step": 45450 | |
| }, | |
| { | |
| "epoch": 1.9170809808713238, | |
| "grad_norm": 0.7879770994186401, | |
| "learning_rate": 2.1727162316001193e-07, | |
| "loss": 2.92824951171875, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 1.9191876632678857, | |
| "grad_norm": 0.8016529679298401, | |
| "learning_rate": 2.0647870678976444e-07, | |
| "loss": 2.911627197265625, | |
| "step": 45550 | |
| }, | |
| { | |
| "epoch": 1.9212943456644478, | |
| "grad_norm": 0.8015576004981995, | |
| "learning_rate": 1.9595961968709442e-07, | |
| "loss": 2.926610107421875, | |
| "step": 45600 | |
| }, | |
| { | |
| "epoch": 1.9234010280610097, | |
| "grad_norm": 0.7970701456069946, | |
| "learning_rate": 1.8571447802887832e-07, | |
| "loss": 2.9195391845703127, | |
| "step": 45650 | |
| }, | |
| { | |
| "epoch": 1.9255077104575715, | |
| "grad_norm": 0.7912192344665527, | |
| "learning_rate": 1.7574339496643777e-07, | |
| "loss": 2.9200799560546873, | |
| "step": 45700 | |
| }, | |
| { | |
| "epoch": 1.9276143928541334, | |
| "grad_norm": 0.8180844783782959, | |
| "learning_rate": 1.6604648062427675e-07, | |
| "loss": 2.9211965942382814, | |
| "step": 45750 | |
| }, | |
| { | |
| "epoch": 1.9297210752506953, | |
| "grad_norm": 0.806681215763092, | |
| "learning_rate": 1.5662384209888802e-07, | |
| "loss": 2.9474331665039064, | |
| "step": 45800 | |
| }, | |
| { | |
| "epoch": 1.9318277576472571, | |
| "grad_norm": 0.7923222184181213, | |
| "learning_rate": 1.4747558345753754e-07, | |
| "loss": 2.9127471923828123, | |
| "step": 45850 | |
| }, | |
| { | |
| "epoch": 1.933934440043819, | |
| "grad_norm": 0.7819007635116577, | |
| "learning_rate": 1.3860180573714855e-07, | |
| "loss": 2.9185443115234375, | |
| "step": 45900 | |
| }, | |
| { | |
| "epoch": 1.9360411224403808, | |
| "grad_norm": 0.7989100813865662, | |
| "learning_rate": 1.3000260694316367e-07, | |
| "loss": 2.90548095703125, | |
| "step": 45950 | |
| }, | |
| { | |
| "epoch": 1.9381478048369427, | |
| "grad_norm": 0.7897266745567322, | |
| "learning_rate": 1.216780820484681e-07, | |
| "loss": 2.924876403808594, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 1.9381478048369427, | |
| "eval_loss": 2.9114933013916016, | |
| "eval_runtime": 119.4231, | |
| "eval_samples_per_second": 128.476, | |
| "eval_steps_per_second": 4.019, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 1.9402544872335046, | |
| "grad_norm": 0.7921951413154602, | |
| "learning_rate": 1.1362832299234572e-07, | |
| "loss": 2.933022155761719, | |
| "step": 46050 | |
| }, | |
| { | |
| "epoch": 1.9423611696300664, | |
| "grad_norm": 0.7944632172584534, | |
| "learning_rate": 1.0585341867946075e-07, | |
| "loss": 2.9296893310546874, | |
| "step": 46100 | |
| }, | |
| { | |
| "epoch": 1.9444678520266283, | |
| "grad_norm": 0.7957251071929932, | |
| "learning_rate": 9.83534549788695e-08, | |
| "loss": 2.9290631103515623, | |
| "step": 46150 | |
| }, | |
| { | |
| "epoch": 1.9465745344231904, | |
| "grad_norm": 0.8100662231445312, | |
| "learning_rate": 9.112851472308225e-08, | |
| "loss": 2.921019287109375, | |
| "step": 46200 | |
| }, | |
| { | |
| "epoch": 1.9486812168197523, | |
| "grad_norm": 0.7996461391448975, | |
| "learning_rate": 8.417867770714182e-08, | |
| "loss": 2.9238427734375, | |
| "step": 46250 | |
| }, | |
| { | |
| "epoch": 1.9507878992163141, | |
| "grad_norm": 0.7716062664985657, | |
| "learning_rate": 7.750402068774088e-08, | |
| "loss": 2.917015380859375, | |
| "step": 46300 | |
| }, | |
| { | |
| "epoch": 1.952894581612876, | |
| "grad_norm": 0.7902867197990417, | |
| "learning_rate": 7.11046173823865e-08, | |
| "loss": 2.9637591552734377, | |
| "step": 46350 | |
| }, | |
| { | |
| "epoch": 1.955001264009438, | |
| "grad_norm": 0.7814793586730957, | |
| "learning_rate": 6.498053846856756e-08, | |
| "loss": 2.8917889404296875, | |
| "step": 46400 | |
| }, | |
| { | |
| "epoch": 1.957107946406, | |
| "grad_norm": 0.8013756275177002, | |
| "learning_rate": 5.9131851582988616e-08, | |
| "loss": 2.915605163574219, | |
| "step": 46450 | |
| }, | |
| { | |
| "epoch": 1.9592146288025618, | |
| "grad_norm": 0.796045184135437, | |
| "learning_rate": 5.355862132081502e-08, | |
| "loss": 2.9377127075195313, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 1.9613213111991237, | |
| "grad_norm": 0.8041338324546814, | |
| "learning_rate": 4.8260909234967845e-08, | |
| "loss": 2.9272357177734376, | |
| "step": 46550 | |
| }, | |
| { | |
| "epoch": 1.9634279935956855, | |
| "grad_norm": 0.7950087189674377, | |
| "learning_rate": 4.323877383543284e-08, | |
| "loss": 2.921788635253906, | |
| "step": 46600 | |
| }, | |
| { | |
| "epoch": 1.9655346759922474, | |
| "grad_norm": 0.808857262134552, | |
| "learning_rate": 3.849227058861926e-08, | |
| "loss": 2.9121929931640627, | |
| "step": 46650 | |
| }, | |
| { | |
| "epoch": 1.9676413583888093, | |
| "grad_norm": 0.7998260855674744, | |
| "learning_rate": 3.402145191675199e-08, | |
| "loss": 2.910013427734375, | |
| "step": 46700 | |
| }, | |
| { | |
| "epoch": 1.9697480407853711, | |
| "grad_norm": 0.8039520978927612, | |
| "learning_rate": 2.982636719728593e-08, | |
| "loss": 2.9208221435546875, | |
| "step": 46750 | |
| }, | |
| { | |
| "epoch": 1.971854723181933, | |
| "grad_norm": 0.7953173518180847, | |
| "learning_rate": 2.590706276236754e-08, | |
| "loss": 2.9036734008789065, | |
| "step": 46800 | |
| }, | |
| { | |
| "epoch": 1.9739614055784949, | |
| "grad_norm": 0.8002514839172363, | |
| "learning_rate": 2.2263581898310236e-08, | |
| "loss": 2.919159240722656, | |
| "step": 46850 | |
| }, | |
| { | |
| "epoch": 1.9760680879750567, | |
| "grad_norm": 0.8161515593528748, | |
| "learning_rate": 1.8960613475257904e-08, | |
| "loss": 2.9043402099609374, | |
| "step": 46900 | |
| }, | |
| { | |
| "epoch": 1.9781747703716188, | |
| "grad_norm": 0.7810516357421875, | |
| "learning_rate": 1.586337906136215e-08, | |
| "loss": 2.914085693359375, | |
| "step": 46950 | |
| }, | |
| { | |
| "epoch": 1.9802814527681807, | |
| "grad_norm": 0.7850042581558228, | |
| "learning_rate": 1.3042079144670594e-08, | |
| "loss": 2.9438748168945312, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 1.9802814527681807, | |
| "eval_loss": 2.9114818572998047, | |
| "eval_runtime": 119.604, | |
| "eval_samples_per_second": 128.282, | |
| "eval_steps_per_second": 4.013, | |
| "step": 47000 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 47466, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.85966432256e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |