diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,7103 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 3.9919316563834837, + "eval_steps": 500, + "global_step": 101064, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.003955070400253125, + "grad_norm": 0.3985116183757782, + "learning_rate": 1.5e-06, + "loss": 2.3445, + "step": 100 + }, + { + "epoch": 0.00791014080050625, + "grad_norm": 0.5253750681877136, + "learning_rate": 3e-06, + "loss": 2.2315, + "step": 200 + }, + { + "epoch": 0.011865211200759373, + "grad_norm": 0.7505590319633484, + "learning_rate": 4.5e-06, + "loss": 2.0777, + "step": 300 + }, + { + "epoch": 0.0158202816010125, + "grad_norm": 0.7067473530769348, + "learning_rate": 6e-06, + "loss": 2.0165, + "step": 400 + }, + { + "epoch": 0.019775352001265623, + "grad_norm": 0.8964686393737793, + "learning_rate": 7.5e-06, + "loss": 1.9872, + "step": 500 + }, + { + "epoch": 0.023730422401518746, + "grad_norm": 1.0233873128890991, + "learning_rate": 9e-06, + "loss": 1.9621, + "step": 600 + }, + { + "epoch": 0.02768549280177187, + "grad_norm": 1.0008875131607056, + "learning_rate": 1.05e-05, + "loss": 1.9519, + "step": 700 + }, + { + "epoch": 0.031640563202025, + "grad_norm": 0.9941542148590088, + "learning_rate": 1.2e-05, + "loss": 1.922, + "step": 800 + }, + { + "epoch": 0.03559563360227812, + "grad_norm": 1.0296485424041748, + "learning_rate": 1.3500000000000001e-05, + "loss": 1.9305, + "step": 900 + }, + { + "epoch": 0.039550704002531245, + "grad_norm": 1.056834101676941, + "learning_rate": 1.5e-05, + "loss": 1.9199, + "step": 1000 + }, + { + "epoch": 0.04350577440278437, + "grad_norm": 1.1616301536560059, + "learning_rate": 1.4999976437535872e-05, + "loss": 1.9058, + "step": 1100 + }, + { + "epoch": 0.04746084480303749, + "grad_norm": 1.0753331184387207, + "learning_rate": 1.4999905750291538e-05, + "loss": 1.9321, + "step": 1200 + }, + { + "epoch": 0.05141591520329062, + "grad_norm": 1.3542801141738892, + "learning_rate": 1.4999787938711148e-05, + "loss": 1.8942, + "step": 1300 + }, + { + "epoch": 0.05537098560354374, + "grad_norm": 1.2391093969345093, + "learning_rate": 1.499962300353495e-05, + "loss": 1.8677, + "step": 1400 + }, + { + "epoch": 0.05932605600379687, + "grad_norm": 1.0844732522964478, + "learning_rate": 1.4999410945799291e-05, + "loss": 1.8723, + "step": 1500 + }, + { + "epoch": 0.06328112640405, + "grad_norm": 1.0814001560211182, + "learning_rate": 1.499915176683659e-05, + "loss": 1.8715, + "step": 1600 + }, + { + "epoch": 0.06723619680430312, + "grad_norm": 1.2003209590911865, + "learning_rate": 1.4998845468275357e-05, + "loss": 1.8739, + "step": 1700 + }, + { + "epoch": 0.07119126720455624, + "grad_norm": 1.2625234127044678, + "learning_rate": 1.4998492052040163e-05, + "loss": 1.8475, + "step": 1800 + }, + { + "epoch": 0.07514633760480936, + "grad_norm": 1.144286870956421, + "learning_rate": 1.499809152035164e-05, + "loss": 1.8517, + "step": 1900 + }, + { + "epoch": 0.07910140800506249, + "grad_norm": 1.2154852151870728, + "learning_rate": 1.4997643875726454e-05, + "loss": 1.8262, + "step": 2000 + }, + { + "epoch": 0.08305647840531562, + "grad_norm": 1.389930009841919, + "learning_rate": 1.4997149120977304e-05, + "loss": 1.8419, + "step": 2100 + }, + { + "epoch": 0.08701154880556874, + "grad_norm": 1.265163779258728, + "learning_rate": 1.4996607259212892e-05, + "loss": 1.8344, + "step": 2200 + }, + { + "epoch": 0.09096661920582186, + "grad_norm": 1.1876202821731567, + "learning_rate": 1.4996018293837914e-05, + "loss": 1.8356, + "step": 2300 + }, + { + "epoch": 0.09492168960607499, + "grad_norm": 1.2677972316741943, + "learning_rate": 1.4995382228553028e-05, + "loss": 1.8463, + "step": 2400 + }, + { + "epoch": 0.09887676000632811, + "grad_norm": 1.206444501876831, + "learning_rate": 1.4994699067354838e-05, + "loss": 1.8256, + "step": 2500 + }, + { + "epoch": 0.10283183040658124, + "grad_norm": 1.2112213373184204, + "learning_rate": 1.4993968814535867e-05, + "loss": 1.8627, + "step": 2600 + }, + { + "epoch": 0.10678690080683437, + "grad_norm": 1.1587677001953125, + "learning_rate": 1.4993191474684532e-05, + "loss": 1.8458, + "step": 2700 + }, + { + "epoch": 0.11074197120708748, + "grad_norm": 1.2365622520446777, + "learning_rate": 1.4992367052685107e-05, + "loss": 1.8383, + "step": 2800 + }, + { + "epoch": 0.11469704160734061, + "grad_norm": 1.179803490638733, + "learning_rate": 1.4991495553717708e-05, + "loss": 1.8305, + "step": 2900 + }, + { + "epoch": 0.11865211200759374, + "grad_norm": 1.336897850036621, + "learning_rate": 1.499057698325824e-05, + "loss": 1.8381, + "step": 3000 + }, + { + "epoch": 0.12260718240784686, + "grad_norm": 1.346336007118225, + "learning_rate": 1.498961134707838e-05, + "loss": 1.8467, + "step": 3100 + }, + { + "epoch": 0.1265622528081, + "grad_norm": 1.2718663215637207, + "learning_rate": 1.4988598651245534e-05, + "loss": 1.8076, + "step": 3200 + }, + { + "epoch": 0.1305173232083531, + "grad_norm": 1.1807332038879395, + "learning_rate": 1.4987538902122799e-05, + "loss": 1.8212, + "step": 3300 + }, + { + "epoch": 0.13447239360860624, + "grad_norm": 1.4749420881271362, + "learning_rate": 1.4986432106368917e-05, + "loss": 1.8039, + "step": 3400 + }, + { + "epoch": 0.13842746400885936, + "grad_norm": 1.2670525312423706, + "learning_rate": 1.4985278270938247e-05, + "loss": 1.8137, + "step": 3500 + }, + { + "epoch": 0.14238253440911247, + "grad_norm": 1.5255069732666016, + "learning_rate": 1.4984077403080711e-05, + "loss": 1.8161, + "step": 3600 + }, + { + "epoch": 0.1463376048093656, + "grad_norm": 1.1995693445205688, + "learning_rate": 1.4982829510341751e-05, + "loss": 1.8104, + "step": 3700 + }, + { + "epoch": 0.15029267520961873, + "grad_norm": 1.3007076978683472, + "learning_rate": 1.4981534600562279e-05, + "loss": 1.7952, + "step": 3800 + }, + { + "epoch": 0.15424774560987187, + "grad_norm": 1.4348576068878174, + "learning_rate": 1.4980192681878635e-05, + "loss": 1.819, + "step": 3900 + }, + { + "epoch": 0.15820281601012498, + "grad_norm": 1.3245015144348145, + "learning_rate": 1.4978803762722526e-05, + "loss": 1.8043, + "step": 4000 + }, + { + "epoch": 0.1621578864103781, + "grad_norm": 1.3621593713760376, + "learning_rate": 1.4977367851820984e-05, + "loss": 1.7992, + "step": 4100 + }, + { + "epoch": 0.16611295681063123, + "grad_norm": 1.2361946105957031, + "learning_rate": 1.4975884958196297e-05, + "loss": 1.8179, + "step": 4200 + }, + { + "epoch": 0.17006802721088435, + "grad_norm": 1.5746525526046753, + "learning_rate": 1.4974355091165972e-05, + "loss": 1.8045, + "step": 4300 + }, + { + "epoch": 0.1740230976111375, + "grad_norm": 1.4326754808425903, + "learning_rate": 1.497277826034265e-05, + "loss": 1.8155, + "step": 4400 + }, + { + "epoch": 0.1779781680113906, + "grad_norm": 1.3772553205490112, + "learning_rate": 1.4971154475634081e-05, + "loss": 1.7838, + "step": 4500 + }, + { + "epoch": 0.18193323841164372, + "grad_norm": 1.4580802917480469, + "learning_rate": 1.4969483747243023e-05, + "loss": 1.7997, + "step": 4600 + }, + { + "epoch": 0.18588830881189686, + "grad_norm": 1.2635383605957031, + "learning_rate": 1.4967766085667204e-05, + "loss": 1.8091, + "step": 4700 + }, + { + "epoch": 0.18984337921214997, + "grad_norm": 1.523247241973877, + "learning_rate": 1.496600150169925e-05, + "loss": 1.8086, + "step": 4800 + }, + { + "epoch": 0.1937984496124031, + "grad_norm": 1.4835641384124756, + "learning_rate": 1.496419000642661e-05, + "loss": 1.8001, + "step": 4900 + }, + { + "epoch": 0.19775352001265623, + "grad_norm": 1.3758567571640015, + "learning_rate": 1.4962331611231496e-05, + "loss": 1.7773, + "step": 5000 + }, + { + "epoch": 0.20170859041290934, + "grad_norm": 1.343885898590088, + "learning_rate": 1.4960426327790808e-05, + "loss": 1.7884, + "step": 5100 + }, + { + "epoch": 0.20566366081316248, + "grad_norm": 1.4712055921554565, + "learning_rate": 1.4958474168076061e-05, + "loss": 1.7904, + "step": 5200 + }, + { + "epoch": 0.2096187312134156, + "grad_norm": 1.3729618787765503, + "learning_rate": 1.4956475144353305e-05, + "loss": 1.7883, + "step": 5300 + }, + { + "epoch": 0.21357380161366873, + "grad_norm": 1.4087861776351929, + "learning_rate": 1.4954429269183049e-05, + "loss": 1.7764, + "step": 5400 + }, + { + "epoch": 0.21752887201392185, + "grad_norm": 1.359567642211914, + "learning_rate": 1.4952336555420194e-05, + "loss": 1.7522, + "step": 5500 + }, + { + "epoch": 0.22148394241417496, + "grad_norm": 1.5321180820465088, + "learning_rate": 1.4950197016213935e-05, + "loss": 1.7858, + "step": 5600 + }, + { + "epoch": 0.2254390128144281, + "grad_norm": 1.4172133207321167, + "learning_rate": 1.4948010665007694e-05, + "loss": 1.7889, + "step": 5700 + }, + { + "epoch": 0.22939408321468122, + "grad_norm": 1.389819622039795, + "learning_rate": 1.4945777515539018e-05, + "loss": 1.7787, + "step": 5800 + }, + { + "epoch": 0.23334915361493436, + "grad_norm": 1.319344162940979, + "learning_rate": 1.4943497581839515e-05, + "loss": 1.7832, + "step": 5900 + }, + { + "epoch": 0.23730422401518747, + "grad_norm": 1.3472243547439575, + "learning_rate": 1.4941170878234739e-05, + "loss": 1.7708, + "step": 6000 + }, + { + "epoch": 0.24125929441544058, + "grad_norm": 1.319574236869812, + "learning_rate": 1.4938797419344127e-05, + "loss": 1.8013, + "step": 6100 + }, + { + "epoch": 0.24521436481569373, + "grad_norm": 1.3281052112579346, + "learning_rate": 1.4936377220080886e-05, + "loss": 1.7657, + "step": 6200 + }, + { + "epoch": 0.24916943521594684, + "grad_norm": 1.4613900184631348, + "learning_rate": 1.4933910295651914e-05, + "loss": 1.7955, + "step": 6300 + }, + { + "epoch": 0.2531245056162, + "grad_norm": 1.6248208284378052, + "learning_rate": 1.4931396661557699e-05, + "loss": 1.7775, + "step": 6400 + }, + { + "epoch": 0.2570795760164531, + "grad_norm": 1.4050931930541992, + "learning_rate": 1.492883633359221e-05, + "loss": 1.749, + "step": 6500 + }, + { + "epoch": 0.2610346464167062, + "grad_norm": 1.3959672451019287, + "learning_rate": 1.4926229327842822e-05, + "loss": 1.7735, + "step": 6600 + }, + { + "epoch": 0.2649897168169593, + "grad_norm": 1.3407930135726929, + "learning_rate": 1.4923575660690197e-05, + "loss": 1.7685, + "step": 6700 + }, + { + "epoch": 0.2689447872172125, + "grad_norm": 1.4229165315628052, + "learning_rate": 1.4920875348808181e-05, + "loss": 1.769, + "step": 6800 + }, + { + "epoch": 0.2728998576174656, + "grad_norm": 1.3997846841812134, + "learning_rate": 1.4918128409163712e-05, + "loss": 1.7804, + "step": 6900 + }, + { + "epoch": 0.2768549280177187, + "grad_norm": 1.7216688394546509, + "learning_rate": 1.4915334859016699e-05, + "loss": 1.7699, + "step": 7000 + }, + { + "epoch": 0.28080999841797183, + "grad_norm": 1.4955641031265259, + "learning_rate": 1.491249471591992e-05, + "loss": 1.7615, + "step": 7100 + }, + { + "epoch": 0.28476506881822494, + "grad_norm": 1.631832480430603, + "learning_rate": 1.4909607997718917e-05, + "loss": 1.7708, + "step": 7200 + }, + { + "epoch": 0.2887201392184781, + "grad_norm": 1.4372748136520386, + "learning_rate": 1.4906674722551872e-05, + "loss": 1.7618, + "step": 7300 + }, + { + "epoch": 0.2926752096187312, + "grad_norm": 1.3430101871490479, + "learning_rate": 1.4903694908849506e-05, + "loss": 1.7734, + "step": 7400 + }, + { + "epoch": 0.29663028001898434, + "grad_norm": 1.4826927185058594, + "learning_rate": 1.4900668575334953e-05, + "loss": 1.7679, + "step": 7500 + }, + { + "epoch": 0.30058535041923745, + "grad_norm": 1.4105191230773926, + "learning_rate": 1.4897595741023642e-05, + "loss": 1.7666, + "step": 7600 + }, + { + "epoch": 0.30454042081949056, + "grad_norm": 1.3381356000900269, + "learning_rate": 1.4894476425223191e-05, + "loss": 1.7697, + "step": 7700 + }, + { + "epoch": 0.30849549121974373, + "grad_norm": 1.3745373487472534, + "learning_rate": 1.4891310647533266e-05, + "loss": 1.7707, + "step": 7800 + }, + { + "epoch": 0.31245056161999685, + "grad_norm": 1.3524316549301147, + "learning_rate": 1.488809842784548e-05, + "loss": 1.7515, + "step": 7900 + }, + { + "epoch": 0.31640563202024996, + "grad_norm": 1.4299299716949463, + "learning_rate": 1.4884839786343242e-05, + "loss": 1.7799, + "step": 8000 + }, + { + "epoch": 0.3203607024205031, + "grad_norm": 1.4132308959960938, + "learning_rate": 1.4881534743501656e-05, + "loss": 1.7258, + "step": 8100 + }, + { + "epoch": 0.3243157728207562, + "grad_norm": 1.4797372817993164, + "learning_rate": 1.4878183320087377e-05, + "loss": 1.7657, + "step": 8200 + }, + { + "epoch": 0.32827084322100936, + "grad_norm": 1.3737105131149292, + "learning_rate": 1.4874785537158479e-05, + "loss": 1.7845, + "step": 8300 + }, + { + "epoch": 0.33222591362126247, + "grad_norm": 1.4676398038864136, + "learning_rate": 1.4871341416064337e-05, + "loss": 1.7652, + "step": 8400 + }, + { + "epoch": 0.3361809840215156, + "grad_norm": 1.3782434463500977, + "learning_rate": 1.4867850978445476e-05, + "loss": 1.7516, + "step": 8500 + }, + { + "epoch": 0.3401360544217687, + "grad_norm": 1.5370761156082153, + "learning_rate": 1.4864314246233448e-05, + "loss": 1.75, + "step": 8600 + }, + { + "epoch": 0.3440911248220218, + "grad_norm": 1.4677528142929077, + "learning_rate": 1.486073124165068e-05, + "loss": 1.7518, + "step": 8700 + }, + { + "epoch": 0.348046195222275, + "grad_norm": 1.4215251207351685, + "learning_rate": 1.4857101987210359e-05, + "loss": 1.7634, + "step": 8800 + }, + { + "epoch": 0.3520012656225281, + "grad_norm": 1.4959337711334229, + "learning_rate": 1.4853426505716261e-05, + "loss": 1.7491, + "step": 8900 + }, + { + "epoch": 0.3559563360227812, + "grad_norm": 1.4005351066589355, + "learning_rate": 1.4849704820262627e-05, + "loss": 1.7713, + "step": 9000 + }, + { + "epoch": 0.3599114064230343, + "grad_norm": 1.4689812660217285, + "learning_rate": 1.484593695423401e-05, + "loss": 1.7448, + "step": 9100 + }, + { + "epoch": 0.36386647682328743, + "grad_norm": 1.5371148586273193, + "learning_rate": 1.4842122931305133e-05, + "loss": 1.7452, + "step": 9200 + }, + { + "epoch": 0.3678215472235406, + "grad_norm": 1.4465723037719727, + "learning_rate": 1.4838262775440741e-05, + "loss": 1.7452, + "step": 9300 + }, + { + "epoch": 0.3717766176237937, + "grad_norm": 1.5890401601791382, + "learning_rate": 1.4834356510895436e-05, + "loss": 1.737, + "step": 9400 + }, + { + "epoch": 0.37573168802404683, + "grad_norm": 1.4862806797027588, + "learning_rate": 1.4830404162213549e-05, + "loss": 1.7426, + "step": 9500 + }, + { + "epoch": 0.37968675842429994, + "grad_norm": 1.5449295043945312, + "learning_rate": 1.4826405754228963e-05, + "loss": 1.7379, + "step": 9600 + }, + { + "epoch": 0.38364182882455306, + "grad_norm": 1.5151877403259277, + "learning_rate": 1.482236131206497e-05, + "loss": 1.7269, + "step": 9700 + }, + { + "epoch": 0.3875968992248062, + "grad_norm": 1.600046157836914, + "learning_rate": 1.4818270861134113e-05, + "loss": 1.7556, + "step": 9800 + }, + { + "epoch": 0.39155196962505934, + "grad_norm": 1.4293779134750366, + "learning_rate": 1.4814134427138015e-05, + "loss": 1.7368, + "step": 9900 + }, + { + "epoch": 0.39550704002531245, + "grad_norm": 1.378175973892212, + "learning_rate": 1.4809952036067231e-05, + "loss": 1.7405, + "step": 10000 + }, + { + "epoch": 0.39946211042556556, + "grad_norm": 1.417622447013855, + "learning_rate": 1.4805723714201079e-05, + "loss": 1.7484, + "step": 10100 + }, + { + "epoch": 0.4034171808258187, + "grad_norm": 1.5106312036514282, + "learning_rate": 1.4801449488107477e-05, + "loss": 1.7218, + "step": 10200 + }, + { + "epoch": 0.40737225122607185, + "grad_norm": 1.5248609781265259, + "learning_rate": 1.4797129384642768e-05, + "loss": 1.7328, + "step": 10300 + }, + { + "epoch": 0.41132732162632496, + "grad_norm": 1.4607023000717163, + "learning_rate": 1.4792763430951562e-05, + "loss": 1.7131, + "step": 10400 + }, + { + "epoch": 0.4152823920265781, + "grad_norm": 1.4600701332092285, + "learning_rate": 1.4788351654466556e-05, + "loss": 1.7418, + "step": 10500 + }, + { + "epoch": 0.4192374624268312, + "grad_norm": 1.3468823432922363, + "learning_rate": 1.4783894082908377e-05, + "loss": 1.7649, + "step": 10600 + }, + { + "epoch": 0.4231925328270843, + "grad_norm": 1.5118048191070557, + "learning_rate": 1.4779390744285386e-05, + "loss": 1.7233, + "step": 10700 + }, + { + "epoch": 0.42714760322733747, + "grad_norm": 1.5199166536331177, + "learning_rate": 1.4774841666893515e-05, + "loss": 1.7238, + "step": 10800 + }, + { + "epoch": 0.4311026736275906, + "grad_norm": 1.6537836790084839, + "learning_rate": 1.4770246879316097e-05, + "loss": 1.7216, + "step": 10900 + }, + { + "epoch": 0.4350577440278437, + "grad_norm": 1.37918221950531, + "learning_rate": 1.4765606410423666e-05, + "loss": 1.7481, + "step": 11000 + }, + { + "epoch": 0.4390128144280968, + "grad_norm": 1.526502013206482, + "learning_rate": 1.4760920289373791e-05, + "loss": 1.7141, + "step": 11100 + }, + { + "epoch": 0.4429678848283499, + "grad_norm": 1.3577282428741455, + "learning_rate": 1.4756188545610884e-05, + "loss": 1.7507, + "step": 11200 + }, + { + "epoch": 0.4469229552286031, + "grad_norm": 1.557986855506897, + "learning_rate": 1.475141120886603e-05, + "loss": 1.7103, + "step": 11300 + }, + { + "epoch": 0.4508780256288562, + "grad_norm": 1.638221025466919, + "learning_rate": 1.474658830915678e-05, + "loss": 1.7363, + "step": 11400 + }, + { + "epoch": 0.4548330960291093, + "grad_norm": 1.472142219543457, + "learning_rate": 1.474171987678697e-05, + "loss": 1.7331, + "step": 11500 + }, + { + "epoch": 0.45878816642936243, + "grad_norm": 1.4680249691009521, + "learning_rate": 1.4736805942346542e-05, + "loss": 1.7273, + "step": 11600 + }, + { + "epoch": 0.46274323682961555, + "grad_norm": 1.4165573120117188, + "learning_rate": 1.4731846536711337e-05, + "loss": 1.7159, + "step": 11700 + }, + { + "epoch": 0.4666983072298687, + "grad_norm": 2.1816458702087402, + "learning_rate": 1.4726841691042902e-05, + "loss": 1.7236, + "step": 11800 + }, + { + "epoch": 0.4706533776301218, + "grad_norm": 1.5376547574996948, + "learning_rate": 1.4721791436788307e-05, + "loss": 1.7227, + "step": 11900 + }, + { + "epoch": 0.47460844803037494, + "grad_norm": 1.6850054264068604, + "learning_rate": 1.4716695805679932e-05, + "loss": 1.7116, + "step": 12000 + }, + { + "epoch": 0.47856351843062805, + "grad_norm": 1.7338590621948242, + "learning_rate": 1.471155482973528e-05, + "loss": 1.7129, + "step": 12100 + }, + { + "epoch": 0.48251858883088117, + "grad_norm": 1.4183164834976196, + "learning_rate": 1.4706368541256762e-05, + "loss": 1.7267, + "step": 12200 + }, + { + "epoch": 0.48647365923113434, + "grad_norm": 1.7117156982421875, + "learning_rate": 1.4701136972831513e-05, + "loss": 1.7149, + "step": 12300 + }, + { + "epoch": 0.49042872963138745, + "grad_norm": 1.4747951030731201, + "learning_rate": 1.4695860157331169e-05, + "loss": 1.7218, + "step": 12400 + }, + { + "epoch": 0.49438380003164056, + "grad_norm": 1.6341221332550049, + "learning_rate": 1.4690538127911672e-05, + "loss": 1.7331, + "step": 12500 + }, + { + "epoch": 0.4983388704318937, + "grad_norm": 1.4981880187988281, + "learning_rate": 1.4685170918013054e-05, + "loss": 1.7182, + "step": 12600 + }, + { + "epoch": 0.5022939408321468, + "grad_norm": 1.5774872303009033, + "learning_rate": 1.4679758561359232e-05, + "loss": 1.7154, + "step": 12700 + }, + { + "epoch": 0.5062490112324, + "grad_norm": 1.5503437519073486, + "learning_rate": 1.4674301091957795e-05, + "loss": 1.716, + "step": 12800 + }, + { + "epoch": 0.5102040816326531, + "grad_norm": 1.5208927392959595, + "learning_rate": 1.4668798544099795e-05, + "loss": 1.7041, + "step": 12900 + }, + { + "epoch": 0.5141591520329062, + "grad_norm": 1.8089638948440552, + "learning_rate": 1.4663250952359516e-05, + "loss": 1.7276, + "step": 13000 + }, + { + "epoch": 0.5181142224331593, + "grad_norm": 1.5653834342956543, + "learning_rate": 1.4657658351594275e-05, + "loss": 1.7164, + "step": 13100 + }, + { + "epoch": 0.5220692928334124, + "grad_norm": 1.7017031908035278, + "learning_rate": 1.4652020776944194e-05, + "loss": 1.7053, + "step": 13200 + }, + { + "epoch": 0.5260243632336655, + "grad_norm": 1.6849620342254639, + "learning_rate": 1.4646338263831977e-05, + "loss": 1.7134, + "step": 13300 + }, + { + "epoch": 0.5299794336339186, + "grad_norm": 1.8098126649856567, + "learning_rate": 1.4640610847962699e-05, + "loss": 1.7158, + "step": 13400 + }, + { + "epoch": 0.5339345040341718, + "grad_norm": 1.7234479188919067, + "learning_rate": 1.4634838565323563e-05, + "loss": 1.7229, + "step": 13500 + }, + { + "epoch": 0.537889574434425, + "grad_norm": 1.35356867313385, + "learning_rate": 1.4629021452183695e-05, + "loss": 1.715, + "step": 13600 + }, + { + "epoch": 0.5418446448346781, + "grad_norm": 1.5286564826965332, + "learning_rate": 1.4623159545093895e-05, + "loss": 1.7011, + "step": 13700 + }, + { + "epoch": 0.5457997152349312, + "grad_norm": 1.5586360692977905, + "learning_rate": 1.4617252880886427e-05, + "loss": 1.6978, + "step": 13800 + }, + { + "epoch": 0.5497547856351843, + "grad_norm": 1.5301753282546997, + "learning_rate": 1.461130149667477e-05, + "loss": 1.6984, + "step": 13900 + }, + { + "epoch": 0.5537098560354374, + "grad_norm": 1.6551586389541626, + "learning_rate": 1.4605305429853402e-05, + "loss": 1.6935, + "step": 14000 + }, + { + "epoch": 0.5576649264356905, + "grad_norm": 1.522283911705017, + "learning_rate": 1.4599264718097552e-05, + "loss": 1.6795, + "step": 14100 + }, + { + "epoch": 0.5616199968359437, + "grad_norm": 1.519173502922058, + "learning_rate": 1.4593179399362967e-05, + "loss": 1.6948, + "step": 14200 + }, + { + "epoch": 0.5655750672361968, + "grad_norm": 1.582780122756958, + "learning_rate": 1.4587049511885675e-05, + "loss": 1.7168, + "step": 14300 + }, + { + "epoch": 0.5695301376364499, + "grad_norm": 1.5130764245986938, + "learning_rate": 1.458087509418174e-05, + "loss": 1.7049, + "step": 14400 + }, + { + "epoch": 0.573485208036703, + "grad_norm": 1.581992268562317, + "learning_rate": 1.4574656185047033e-05, + "loss": 1.695, + "step": 14500 + }, + { + "epoch": 0.5774402784369562, + "grad_norm": 1.4675225019454956, + "learning_rate": 1.456839282355697e-05, + "loss": 1.7015, + "step": 14600 + }, + { + "epoch": 0.5813953488372093, + "grad_norm": 1.5948406457901, + "learning_rate": 1.4562085049066282e-05, + "loss": 1.7129, + "step": 14700 + }, + { + "epoch": 0.5853504192374624, + "grad_norm": 1.8901729583740234, + "learning_rate": 1.4555732901208756e-05, + "loss": 1.7062, + "step": 14800 + }, + { + "epoch": 0.5893054896377156, + "grad_norm": 1.6940269470214844, + "learning_rate": 1.4549336419896993e-05, + "loss": 1.7025, + "step": 14900 + }, + { + "epoch": 0.5932605600379687, + "grad_norm": 1.5160539150238037, + "learning_rate": 1.454289564532216e-05, + "loss": 1.688, + "step": 15000 + }, + { + "epoch": 0.5972156304382218, + "grad_norm": 1.6424893140792847, + "learning_rate": 1.4536410617953726e-05, + "loss": 1.696, + "step": 15100 + }, + { + "epoch": 0.6011707008384749, + "grad_norm": 1.492990493774414, + "learning_rate": 1.4529881378539218e-05, + "loss": 1.6768, + "step": 15200 + }, + { + "epoch": 0.605125771238728, + "grad_norm": 1.7309181690216064, + "learning_rate": 1.452330796810396e-05, + "loss": 1.6972, + "step": 15300 + }, + { + "epoch": 0.6090808416389811, + "grad_norm": 1.684484601020813, + "learning_rate": 1.451669042795082e-05, + "loss": 1.6903, + "step": 15400 + }, + { + "epoch": 0.6130359120392342, + "grad_norm": 1.5465792417526245, + "learning_rate": 1.4510028799659944e-05, + "loss": 1.714, + "step": 15500 + }, + { + "epoch": 0.6169909824394875, + "grad_norm": 1.8257033824920654, + "learning_rate": 1.4503323125088501e-05, + "loss": 1.6894, + "step": 15600 + }, + { + "epoch": 0.6209460528397406, + "grad_norm": 1.5299944877624512, + "learning_rate": 1.4496573446370414e-05, + "loss": 1.6944, + "step": 15700 + }, + { + "epoch": 0.6249011232399937, + "grad_norm": 1.7090293169021606, + "learning_rate": 1.44897798059161e-05, + "loss": 1.6878, + "step": 15800 + }, + { + "epoch": 0.6288561936402468, + "grad_norm": 1.690470576286316, + "learning_rate": 1.4482942246412203e-05, + "loss": 1.6807, + "step": 15900 + }, + { + "epoch": 0.6328112640404999, + "grad_norm": 1.8261181116104126, + "learning_rate": 1.4476060810821319e-05, + "loss": 1.6887, + "step": 16000 + }, + { + "epoch": 0.636766334440753, + "grad_norm": 1.5878318548202515, + "learning_rate": 1.4469135542381741e-05, + "loss": 1.6618, + "step": 16100 + }, + { + "epoch": 0.6407214048410061, + "grad_norm": 1.5003888607025146, + "learning_rate": 1.4462166484607167e-05, + "loss": 1.6734, + "step": 16200 + }, + { + "epoch": 0.6446764752412593, + "grad_norm": 1.7296781539916992, + "learning_rate": 1.445515368128645e-05, + "loss": 1.6712, + "step": 16300 + }, + { + "epoch": 0.6486315456415124, + "grad_norm": 1.6283060312271118, + "learning_rate": 1.4448097176483299e-05, + "loss": 1.6963, + "step": 16400 + }, + { + "epoch": 0.6525866160417655, + "grad_norm": 1.5867258310317993, + "learning_rate": 1.444099701453602e-05, + "loss": 1.6834, + "step": 16500 + }, + { + "epoch": 0.6565416864420187, + "grad_norm": 1.8763879537582397, + "learning_rate": 1.4433853240057229e-05, + "loss": 1.6811, + "step": 16600 + }, + { + "epoch": 0.6604967568422718, + "grad_norm": 1.5323275327682495, + "learning_rate": 1.4426665897933574e-05, + "loss": 1.6778, + "step": 16700 + }, + { + "epoch": 0.6644518272425249, + "grad_norm": 1.581667184829712, + "learning_rate": 1.4419435033325455e-05, + "loss": 1.6926, + "step": 16800 + }, + { + "epoch": 0.668406897642778, + "grad_norm": 1.6673179864883423, + "learning_rate": 1.441216069166673e-05, + "loss": 1.6806, + "step": 16900 + }, + { + "epoch": 0.6723619680430312, + "grad_norm": 1.8026336431503296, + "learning_rate": 1.4404842918664446e-05, + "loss": 1.6829, + "step": 17000 + }, + { + "epoch": 0.6763170384432843, + "grad_norm": 1.6094428300857544, + "learning_rate": 1.4397481760298542e-05, + "loss": 1.6763, + "step": 17100 + }, + { + "epoch": 0.6802721088435374, + "grad_norm": 1.565843105316162, + "learning_rate": 1.4390077262821559e-05, + "loss": 1.659, + "step": 17200 + }, + { + "epoch": 0.6842271792437905, + "grad_norm": 1.7567963600158691, + "learning_rate": 1.4382629472758346e-05, + "loss": 1.666, + "step": 17300 + }, + { + "epoch": 0.6881822496440436, + "grad_norm": 1.591693639755249, + "learning_rate": 1.4375138436905786e-05, + "loss": 1.6666, + "step": 17400 + }, + { + "epoch": 0.6921373200442967, + "grad_norm": 1.638576865196228, + "learning_rate": 1.436760420233248e-05, + "loss": 1.6554, + "step": 17500 + }, + { + "epoch": 0.69609239044455, + "grad_norm": 1.7055751085281372, + "learning_rate": 1.4360026816378462e-05, + "loss": 1.671, + "step": 17600 + }, + { + "epoch": 0.7000474608448031, + "grad_norm": 1.6867974996566772, + "learning_rate": 1.4352406326654905e-05, + "loss": 1.6722, + "step": 17700 + }, + { + "epoch": 0.7040025312450562, + "grad_norm": 1.7862675189971924, + "learning_rate": 1.4344742781043809e-05, + "loss": 1.6965, + "step": 17800 + }, + { + "epoch": 0.7079576016453093, + "grad_norm": 1.7989298105239868, + "learning_rate": 1.4337036227697715e-05, + "loss": 1.6762, + "step": 17900 + }, + { + "epoch": 0.7119126720455624, + "grad_norm": 1.7017799615859985, + "learning_rate": 1.4329286715039392e-05, + "loss": 1.6614, + "step": 18000 + }, + { + "epoch": 0.7158677424458155, + "grad_norm": 1.6665624380111694, + "learning_rate": 1.4321494291761537e-05, + "loss": 1.662, + "step": 18100 + }, + { + "epoch": 0.7198228128460686, + "grad_norm": 1.9826371669769287, + "learning_rate": 1.4313659006826468e-05, + "loss": 1.6638, + "step": 18200 + }, + { + "epoch": 0.7237778832463218, + "grad_norm": 1.7711797952651978, + "learning_rate": 1.430578090946582e-05, + "loss": 1.6625, + "step": 18300 + }, + { + "epoch": 0.7277329536465749, + "grad_norm": 1.8430758714675903, + "learning_rate": 1.4297860049180223e-05, + "loss": 1.6824, + "step": 18400 + }, + { + "epoch": 0.731688024046828, + "grad_norm": 1.8522826433181763, + "learning_rate": 1.4289896475739012e-05, + "loss": 1.6599, + "step": 18500 + }, + { + "epoch": 0.7356430944470812, + "grad_norm": 1.6422381401062012, + "learning_rate": 1.4281890239179897e-05, + "loss": 1.6584, + "step": 18600 + }, + { + "epoch": 0.7395981648473343, + "grad_norm": 1.7232320308685303, + "learning_rate": 1.4273841389808653e-05, + "loss": 1.6604, + "step": 18700 + }, + { + "epoch": 0.7435532352475874, + "grad_norm": 1.6723573207855225, + "learning_rate": 1.4265749978198805e-05, + "loss": 1.6654, + "step": 18800 + }, + { + "epoch": 0.7475083056478405, + "grad_norm": 1.776376485824585, + "learning_rate": 1.4257616055191316e-05, + "loss": 1.672, + "step": 18900 + }, + { + "epoch": 0.7514633760480937, + "grad_norm": 1.9001519680023193, + "learning_rate": 1.4249439671894253e-05, + "loss": 1.6473, + "step": 19000 + }, + { + "epoch": 0.7554184464483468, + "grad_norm": 1.5602456331253052, + "learning_rate": 1.4241220879682484e-05, + "loss": 1.6558, + "step": 19100 + }, + { + "epoch": 0.7593735168485999, + "grad_norm": 1.6195124387741089, + "learning_rate": 1.423295973019734e-05, + "loss": 1.6704, + "step": 19200 + }, + { + "epoch": 0.763328587248853, + "grad_norm": 1.7354437112808228, + "learning_rate": 1.4224656275346295e-05, + "loss": 1.6697, + "step": 19300 + }, + { + "epoch": 0.7672836576491061, + "grad_norm": 1.6655138731002808, + "learning_rate": 1.4216310567302648e-05, + "loss": 1.6567, + "step": 19400 + }, + { + "epoch": 0.7712387280493592, + "grad_norm": 1.9168803691864014, + "learning_rate": 1.4207922658505184e-05, + "loss": 1.6398, + "step": 19500 + }, + { + "epoch": 0.7751937984496124, + "grad_norm": 1.7918130159378052, + "learning_rate": 1.4199492601657848e-05, + "loss": 1.654, + "step": 19600 + }, + { + "epoch": 0.7791488688498656, + "grad_norm": 1.7283716201782227, + "learning_rate": 1.4191020449729417e-05, + "loss": 1.6534, + "step": 19700 + }, + { + "epoch": 0.7831039392501187, + "grad_norm": 1.860144019126892, + "learning_rate": 1.4182506255953167e-05, + "loss": 1.6553, + "step": 19800 + }, + { + "epoch": 0.7870590096503718, + "grad_norm": 1.7320619821548462, + "learning_rate": 1.4173950073826531e-05, + "loss": 1.6586, + "step": 19900 + }, + { + "epoch": 0.7910140800506249, + "grad_norm": 1.7704521417617798, + "learning_rate": 1.4165351957110772e-05, + "loss": 1.6599, + "step": 20000 + }, + { + "epoch": 0.794969150450878, + "grad_norm": 2.051400661468506, + "learning_rate": 1.4156711959830644e-05, + "loss": 1.6695, + "step": 20100 + }, + { + "epoch": 0.7989242208511311, + "grad_norm": 1.7863457202911377, + "learning_rate": 1.4148030136274043e-05, + "loss": 1.6538, + "step": 20200 + }, + { + "epoch": 0.8028792912513842, + "grad_norm": 1.85243558883667, + "learning_rate": 1.413930654099168e-05, + "loss": 1.63, + "step": 20300 + }, + { + "epoch": 0.8068343616516374, + "grad_norm": 1.7953428030014038, + "learning_rate": 1.413054122879673e-05, + "loss": 1.6374, + "step": 20400 + }, + { + "epoch": 0.8107894320518905, + "grad_norm": 1.8957959413528442, + "learning_rate": 1.4121734254764482e-05, + "loss": 1.6445, + "step": 20500 + }, + { + "epoch": 0.8147445024521437, + "grad_norm": 1.7762993574142456, + "learning_rate": 1.4112885674232011e-05, + "loss": 1.6503, + "step": 20600 + }, + { + "epoch": 0.8186995728523968, + "grad_norm": 1.723813772201538, + "learning_rate": 1.410399554279781e-05, + "loss": 1.6416, + "step": 20700 + }, + { + "epoch": 0.8226546432526499, + "grad_norm": 1.9667476415634155, + "learning_rate": 1.4095063916321456e-05, + "loss": 1.6297, + "step": 20800 + }, + { + "epoch": 0.826609713652903, + "grad_norm": 1.8855000734329224, + "learning_rate": 1.4086090850923246e-05, + "loss": 1.6684, + "step": 20900 + }, + { + "epoch": 0.8305647840531561, + "grad_norm": 1.8669531345367432, + "learning_rate": 1.4077076402983857e-05, + "loss": 1.6344, + "step": 21000 + }, + { + "epoch": 0.8345198544534093, + "grad_norm": 2.0968127250671387, + "learning_rate": 1.4068020629143985e-05, + "loss": 1.6524, + "step": 21100 + }, + { + "epoch": 0.8384749248536624, + "grad_norm": 1.712428092956543, + "learning_rate": 1.4058923586303988e-05, + "loss": 1.6333, + "step": 21200 + }, + { + "epoch": 0.8424299952539155, + "grad_norm": 1.7560315132141113, + "learning_rate": 1.4049785331623534e-05, + "loss": 1.6337, + "step": 21300 + }, + { + "epoch": 0.8463850656541686, + "grad_norm": 1.875779390335083, + "learning_rate": 1.4040605922521231e-05, + "loss": 1.6328, + "step": 21400 + }, + { + "epoch": 0.8503401360544217, + "grad_norm": 1.9142519235610962, + "learning_rate": 1.403138541667428e-05, + "loss": 1.6333, + "step": 21500 + }, + { + "epoch": 0.8542952064546749, + "grad_norm": 2.263770580291748, + "learning_rate": 1.4022123872018107e-05, + "loss": 1.635, + "step": 21600 + }, + { + "epoch": 0.858250276854928, + "grad_norm": 1.838889479637146, + "learning_rate": 1.4012821346745995e-05, + "loss": 1.6328, + "step": 21700 + }, + { + "epoch": 0.8622053472551812, + "grad_norm": 1.9877907037734985, + "learning_rate": 1.400347789930872e-05, + "loss": 1.6427, + "step": 21800 + }, + { + "epoch": 0.8661604176554343, + "grad_norm": 1.9885168075561523, + "learning_rate": 1.399409358841419e-05, + "loss": 1.637, + "step": 21900 + }, + { + "epoch": 0.8701154880556874, + "grad_norm": 1.8428804874420166, + "learning_rate": 1.3984668473027065e-05, + "loss": 1.6377, + "step": 22000 + }, + { + "epoch": 0.8740705584559405, + "grad_norm": 1.9875133037567139, + "learning_rate": 1.39752026123684e-05, + "loss": 1.6248, + "step": 22100 + }, + { + "epoch": 0.8780256288561936, + "grad_norm": 2.0525448322296143, + "learning_rate": 1.3965696065915262e-05, + "loss": 1.6501, + "step": 22200 + }, + { + "epoch": 0.8819806992564467, + "grad_norm": 1.9695172309875488, + "learning_rate": 1.3956148893400357e-05, + "loss": 1.6308, + "step": 22300 + }, + { + "epoch": 0.8859357696566998, + "grad_norm": 2.084592580795288, + "learning_rate": 1.3946561154811664e-05, + "loss": 1.6335, + "step": 22400 + }, + { + "epoch": 0.889890840056953, + "grad_norm": 1.7602378129959106, + "learning_rate": 1.3936932910392048e-05, + "loss": 1.6195, + "step": 22500 + }, + { + "epoch": 0.8938459104572062, + "grad_norm": 2.0157277584075928, + "learning_rate": 1.3927264220638889e-05, + "loss": 1.6395, + "step": 22600 + }, + { + "epoch": 0.8978009808574593, + "grad_norm": 2.184307336807251, + "learning_rate": 1.391755514630369e-05, + "loss": 1.6448, + "step": 22700 + }, + { + "epoch": 0.9017560512577124, + "grad_norm": 1.9747377634048462, + "learning_rate": 1.390780574839171e-05, + "loss": 1.6302, + "step": 22800 + }, + { + "epoch": 0.9057111216579655, + "grad_norm": 2.1203644275665283, + "learning_rate": 1.3898016088161575e-05, + "loss": 1.6447, + "step": 22900 + }, + { + "epoch": 0.9096661920582186, + "grad_norm": 2.0279908180236816, + "learning_rate": 1.3888186227124885e-05, + "loss": 1.622, + "step": 23000 + }, + { + "epoch": 0.9136212624584718, + "grad_norm": 1.9809517860412598, + "learning_rate": 1.3878316227045846e-05, + "loss": 1.6189, + "step": 23100 + }, + { + "epoch": 0.9175763328587249, + "grad_norm": 2.1499814987182617, + "learning_rate": 1.386840614994086e-05, + "loss": 1.6192, + "step": 23200 + }, + { + "epoch": 0.921531403258978, + "grad_norm": 1.9230985641479492, + "learning_rate": 1.3858456058078148e-05, + "loss": 1.6251, + "step": 23300 + }, + { + "epoch": 0.9254864736592311, + "grad_norm": 2.1623005867004395, + "learning_rate": 1.3848466013977365e-05, + "loss": 1.6145, + "step": 23400 + }, + { + "epoch": 0.9294415440594842, + "grad_norm": 2.0579729080200195, + "learning_rate": 1.3838436080409188e-05, + "loss": 1.628, + "step": 23500 + }, + { + "epoch": 0.9333966144597374, + "grad_norm": 1.9566960334777832, + "learning_rate": 1.3828366320394937e-05, + "loss": 1.6317, + "step": 23600 + }, + { + "epoch": 0.9373516848599905, + "grad_norm": 1.9498804807662964, + "learning_rate": 1.3818256797206177e-05, + "loss": 1.6107, + "step": 23700 + }, + { + "epoch": 0.9413067552602437, + "grad_norm": 1.8530848026275635, + "learning_rate": 1.3808107574364312e-05, + "loss": 1.6186, + "step": 23800 + }, + { + "epoch": 0.9452618256604968, + "grad_norm": 2.0309464931488037, + "learning_rate": 1.3797918715640197e-05, + "loss": 1.6208, + "step": 23900 + }, + { + "epoch": 0.9492168960607499, + "grad_norm": 2.0447559356689453, + "learning_rate": 1.3787690285053732e-05, + "loss": 1.6226, + "step": 24000 + }, + { + "epoch": 0.953171966461003, + "grad_norm": 2.198214292526245, + "learning_rate": 1.3777422346873453e-05, + "loss": 1.5954, + "step": 24100 + }, + { + "epoch": 0.9571270368612561, + "grad_norm": 2.0797901153564453, + "learning_rate": 1.3767114965616143e-05, + "loss": 1.6378, + "step": 24200 + }, + { + "epoch": 0.9610821072615092, + "grad_norm": 2.361177444458008, + "learning_rate": 1.3756768206046418e-05, + "loss": 1.6207, + "step": 24300 + }, + { + "epoch": 0.9650371776617623, + "grad_norm": 2.194758653640747, + "learning_rate": 1.3746382133176314e-05, + "loss": 1.6147, + "step": 24400 + }, + { + "epoch": 0.9689922480620154, + "grad_norm": 2.1259610652923584, + "learning_rate": 1.3735956812264893e-05, + "loss": 1.583, + "step": 24500 + }, + { + "epoch": 0.9729473184622687, + "grad_norm": 2.084428548812866, + "learning_rate": 1.372549230881782e-05, + "loss": 1.6257, + "step": 24600 + }, + { + "epoch": 0.9769023888625218, + "grad_norm": 2.0704309940338135, + "learning_rate": 1.3714988688586958e-05, + "loss": 1.6062, + "step": 24700 + }, + { + "epoch": 0.9808574592627749, + "grad_norm": 1.9661308526992798, + "learning_rate": 1.3704446017569953e-05, + "loss": 1.6164, + "step": 24800 + }, + { + "epoch": 0.984812529663028, + "grad_norm": 2.1788337230682373, + "learning_rate": 1.3693864362009821e-05, + "loss": 1.6188, + "step": 24900 + }, + { + "epoch": 0.9887676000632811, + "grad_norm": 1.9492045640945435, + "learning_rate": 1.3683243788394534e-05, + "loss": 1.609, + "step": 25000 + }, + { + "epoch": 0.9927226704635342, + "grad_norm": 2.2324581146240234, + "learning_rate": 1.3672584363456587e-05, + "loss": 1.6058, + "step": 25100 + }, + { + "epoch": 0.9966777408637874, + "grad_norm": 2.14666485786438, + "learning_rate": 1.3661886154172602e-05, + "loss": 1.6059, + "step": 25200 + }, + { + "epoch": 1.0006328112640406, + "grad_norm": 2.1909172534942627, + "learning_rate": 1.3651149227762893e-05, + "loss": 1.6006, + "step": 25300 + }, + { + "epoch": 1.0025726272461015, + "grad_norm": 2.143050193786621, + "learning_rate": 1.3640373651691044e-05, + "loss": 1.5379, + "step": 25400 + }, + { + "epoch": 1.0065305153170268, + "grad_norm": 2.6964924335479736, + "learning_rate": 1.3629559493663487e-05, + "loss": 1.507, + "step": 25500 + }, + { + "epoch": 1.0104884033879522, + "grad_norm": 2.556349992752075, + "learning_rate": 1.361870682162908e-05, + "loss": 1.5266, + "step": 25600 + }, + { + "epoch": 1.0144462914588774, + "grad_norm": 2.2218785285949707, + "learning_rate": 1.3607815703778673e-05, + "loss": 1.4867, + "step": 25700 + }, + { + "epoch": 1.0184041795298029, + "grad_norm": 2.5236029624938965, + "learning_rate": 1.3596886208544687e-05, + "loss": 1.4879, + "step": 25800 + }, + { + "epoch": 1.0223620676007283, + "grad_norm": 2.5998125076293945, + "learning_rate": 1.3585918404600679e-05, + "loss": 1.524, + "step": 25900 + }, + { + "epoch": 1.0263199556716536, + "grad_norm": 2.6749346256256104, + "learning_rate": 1.3574912360860912e-05, + "loss": 1.5148, + "step": 26000 + }, + { + "epoch": 1.030277843742579, + "grad_norm": 2.4651994705200195, + "learning_rate": 1.3563868146479921e-05, + "loss": 1.4905, + "step": 26100 + }, + { + "epoch": 1.0342357318135043, + "grad_norm": 2.4166133403778076, + "learning_rate": 1.3552785830852084e-05, + "loss": 1.5155, + "step": 26200 + }, + { + "epoch": 1.0381936198844297, + "grad_norm": 2.2493224143981934, + "learning_rate": 1.3541665483611175e-05, + "loss": 1.515, + "step": 26300 + }, + { + "epoch": 1.042151507955355, + "grad_norm": 2.7555856704711914, + "learning_rate": 1.3530507174629938e-05, + "loss": 1.5136, + "step": 26400 + }, + { + "epoch": 1.0461093960262804, + "grad_norm": 2.7666208744049072, + "learning_rate": 1.3519310974019639e-05, + "loss": 1.4917, + "step": 26500 + }, + { + "epoch": 1.0500672840972056, + "grad_norm": 2.7659711837768555, + "learning_rate": 1.3508076952129634e-05, + "loss": 1.5076, + "step": 26600 + }, + { + "epoch": 1.054025172168131, + "grad_norm": 2.5507092475891113, + "learning_rate": 1.3496805179546919e-05, + "loss": 1.5052, + "step": 26700 + }, + { + "epoch": 1.0579830602390565, + "grad_norm": 2.302542209625244, + "learning_rate": 1.3485495727095687e-05, + "loss": 1.5034, + "step": 26800 + }, + { + "epoch": 1.0619409483099818, + "grad_norm": 2.578275203704834, + "learning_rate": 1.3474148665836894e-05, + "loss": 1.4886, + "step": 26900 + }, + { + "epoch": 1.0658988363809072, + "grad_norm": 2.354796886444092, + "learning_rate": 1.3462764067067799e-05, + "loss": 1.506, + "step": 27000 + }, + { + "epoch": 1.0698567244518324, + "grad_norm": 2.5606160163879395, + "learning_rate": 1.345134200232152e-05, + "loss": 1.4931, + "step": 27100 + }, + { + "epoch": 1.073814612522758, + "grad_norm": 2.46881365776062, + "learning_rate": 1.343988254336659e-05, + "loss": 1.503, + "step": 27200 + }, + { + "epoch": 1.0777725005936831, + "grad_norm": 2.657731771469116, + "learning_rate": 1.3428385762206498e-05, + "loss": 1.5064, + "step": 27300 + }, + { + "epoch": 1.0817303886646086, + "grad_norm": 2.3708932399749756, + "learning_rate": 1.3416851731079244e-05, + "loss": 1.4943, + "step": 27400 + }, + { + "epoch": 1.0856882767355338, + "grad_norm": 2.6182353496551514, + "learning_rate": 1.340528052245688e-05, + "loss": 1.5001, + "step": 27500 + }, + { + "epoch": 1.0896461648064593, + "grad_norm": 2.7265477180480957, + "learning_rate": 1.3393672209045055e-05, + "loss": 1.4794, + "step": 27600 + }, + { + "epoch": 1.0936040528773847, + "grad_norm": 2.6186697483062744, + "learning_rate": 1.3382026863782559e-05, + "loss": 1.4878, + "step": 27700 + }, + { + "epoch": 1.09756194094831, + "grad_norm": 3.3754959106445312, + "learning_rate": 1.3370344559840868e-05, + "loss": 1.4769, + "step": 27800 + }, + { + "epoch": 1.1015198290192354, + "grad_norm": 3.0805869102478027, + "learning_rate": 1.3358625370623684e-05, + "loss": 1.5098, + "step": 27900 + }, + { + "epoch": 1.1054777170901606, + "grad_norm": 2.626561403274536, + "learning_rate": 1.334686936976646e-05, + "loss": 1.4965, + "step": 28000 + }, + { + "epoch": 1.109435605161086, + "grad_norm": 2.6148223876953125, + "learning_rate": 1.333507663113596e-05, + "loss": 1.4973, + "step": 28100 + }, + { + "epoch": 1.1133934932320113, + "grad_norm": 2.9437952041625977, + "learning_rate": 1.3323247228829781e-05, + "loss": 1.4861, + "step": 28200 + }, + { + "epoch": 1.1173513813029368, + "grad_norm": 2.7327873706817627, + "learning_rate": 1.3311381237175882e-05, + "loss": 1.4725, + "step": 28300 + }, + { + "epoch": 1.121309269373862, + "grad_norm": 2.8548924922943115, + "learning_rate": 1.3299478730732134e-05, + "loss": 1.4825, + "step": 28400 + }, + { + "epoch": 1.1252671574447874, + "grad_norm": 2.638568878173828, + "learning_rate": 1.3287539784285839e-05, + "loss": 1.4715, + "step": 28500 + }, + { + "epoch": 1.129225045515713, + "grad_norm": 3.1752021312713623, + "learning_rate": 1.327556447285326e-05, + "loss": 1.4749, + "step": 28600 + }, + { + "epoch": 1.1331829335866381, + "grad_norm": 2.8398923873901367, + "learning_rate": 1.3263552871679156e-05, + "loss": 1.4901, + "step": 28700 + }, + { + "epoch": 1.1371408216575636, + "grad_norm": 2.708963632583618, + "learning_rate": 1.3251505056236312e-05, + "loss": 1.4805, + "step": 28800 + }, + { + "epoch": 1.1410987097284888, + "grad_norm": 2.9168691635131836, + "learning_rate": 1.3239421102225049e-05, + "loss": 1.4653, + "step": 28900 + }, + { + "epoch": 1.1450565977994143, + "grad_norm": 2.523481845855713, + "learning_rate": 1.322730108557276e-05, + "loss": 1.4822, + "step": 29000 + }, + { + "epoch": 1.1490144858703395, + "grad_norm": 3.2788479328155518, + "learning_rate": 1.3215145082433436e-05, + "loss": 1.4685, + "step": 29100 + }, + { + "epoch": 1.152972373941265, + "grad_norm": 2.765491485595703, + "learning_rate": 1.3202953169187181e-05, + "loss": 1.4878, + "step": 29200 + }, + { + "epoch": 1.1569302620121902, + "grad_norm": 2.9480185508728027, + "learning_rate": 1.3190725422439734e-05, + "loss": 1.4607, + "step": 29300 + }, + { + "epoch": 1.1608881500831156, + "grad_norm": 2.9779725074768066, + "learning_rate": 1.3178461919021984e-05, + "loss": 1.4724, + "step": 29400 + }, + { + "epoch": 1.164846038154041, + "grad_norm": 2.768763780593872, + "learning_rate": 1.3166162735989497e-05, + "loss": 1.4989, + "step": 29500 + }, + { + "epoch": 1.1688039262249663, + "grad_norm": 3.2208807468414307, + "learning_rate": 1.3153827950622019e-05, + "loss": 1.4695, + "step": 29600 + }, + { + "epoch": 1.1727618142958918, + "grad_norm": 2.7532846927642822, + "learning_rate": 1.3141457640423002e-05, + "loss": 1.4841, + "step": 29700 + }, + { + "epoch": 1.176719702366817, + "grad_norm": 3.399897575378418, + "learning_rate": 1.3129051883119107e-05, + "loss": 1.4627, + "step": 29800 + }, + { + "epoch": 1.1806775904377425, + "grad_norm": 2.892542600631714, + "learning_rate": 1.311661075665973e-05, + "loss": 1.4754, + "step": 29900 + }, + { + "epoch": 1.1846354785086677, + "grad_norm": 2.6261606216430664, + "learning_rate": 1.310413433921649e-05, + "loss": 1.4847, + "step": 30000 + }, + { + "epoch": 1.1885933665795931, + "grad_norm": 2.6923210620880127, + "learning_rate": 1.3091622709182762e-05, + "loss": 1.4722, + "step": 30100 + }, + { + "epoch": 1.1925512546505184, + "grad_norm": 3.0266880989074707, + "learning_rate": 1.3079075945173164e-05, + "loss": 1.4922, + "step": 30200 + }, + { + "epoch": 1.1965091427214438, + "grad_norm": 2.9252758026123047, + "learning_rate": 1.306649412602308e-05, + "loss": 1.4692, + "step": 30300 + }, + { + "epoch": 1.2004670307923693, + "grad_norm": 2.757887125015259, + "learning_rate": 1.305387733078815e-05, + "loss": 1.465, + "step": 30400 + }, + { + "epoch": 1.2044249188632945, + "grad_norm": 3.536576271057129, + "learning_rate": 1.304122563874379e-05, + "loss": 1.4625, + "step": 30500 + }, + { + "epoch": 1.20838280693422, + "grad_norm": 2.765883684158325, + "learning_rate": 1.3028539129384668e-05, + "loss": 1.4509, + "step": 30600 + }, + { + "epoch": 1.2123406950051452, + "grad_norm": 2.8738183975219727, + "learning_rate": 1.3015817882424235e-05, + "loss": 1.4421, + "step": 30700 + }, + { + "epoch": 1.2162985830760706, + "grad_norm": 2.8780033588409424, + "learning_rate": 1.3003061977794207e-05, + "loss": 1.4749, + "step": 30800 + }, + { + "epoch": 1.2202564711469959, + "grad_norm": 3.499163866043091, + "learning_rate": 1.2990271495644059e-05, + "loss": 1.4564, + "step": 30900 + }, + { + "epoch": 1.2242143592179213, + "grad_norm": 2.900714159011841, + "learning_rate": 1.297744651634053e-05, + "loss": 1.4642, + "step": 31000 + }, + { + "epoch": 1.2281722472888466, + "grad_norm": 2.939680337905884, + "learning_rate": 1.2964587120467122e-05, + "loss": 1.4512, + "step": 31100 + }, + { + "epoch": 1.232130135359772, + "grad_norm": 3.2539045810699463, + "learning_rate": 1.2951693388823577e-05, + "loss": 1.4478, + "step": 31200 + }, + { + "epoch": 1.2360880234306975, + "grad_norm": 3.0612869262695312, + "learning_rate": 1.293876540242539e-05, + "loss": 1.4674, + "step": 31300 + }, + { + "epoch": 1.2400459115016227, + "grad_norm": 2.9142935276031494, + "learning_rate": 1.2925803242503287e-05, + "loss": 1.4704, + "step": 31400 + }, + { + "epoch": 1.2440037995725481, + "grad_norm": 3.0163750648498535, + "learning_rate": 1.291280699050271e-05, + "loss": 1.4433, + "step": 31500 + }, + { + "epoch": 1.2479616876434734, + "grad_norm": 2.712174654006958, + "learning_rate": 1.289977672808332e-05, + "loss": 1.4637, + "step": 31600 + }, + { + "epoch": 1.2519195757143988, + "grad_norm": 2.7295100688934326, + "learning_rate": 1.2886712537118475e-05, + "loss": 1.4599, + "step": 31700 + }, + { + "epoch": 1.2558774637853243, + "grad_norm": 3.3285765647888184, + "learning_rate": 1.2873614499694717e-05, + "loss": 1.4432, + "step": 31800 + }, + { + "epoch": 1.2598353518562495, + "grad_norm": 3.4205710887908936, + "learning_rate": 1.2860482698111254e-05, + "loss": 1.4571, + "step": 31900 + }, + { + "epoch": 1.2637932399271747, + "grad_norm": 2.865621328353882, + "learning_rate": 1.2847317214879451e-05, + "loss": 1.4434, + "step": 32000 + }, + { + "epoch": 1.2677511279981002, + "grad_norm": 2.7961373329162598, + "learning_rate": 1.2834118132722296e-05, + "loss": 1.4305, + "step": 32100 + }, + { + "epoch": 1.2717090160690256, + "grad_norm": 2.86441707611084, + "learning_rate": 1.2820885534573903e-05, + "loss": 1.4592, + "step": 32200 + }, + { + "epoch": 1.2756669041399509, + "grad_norm": 3.3199241161346436, + "learning_rate": 1.2807619503578964e-05, + "loss": 1.451, + "step": 32300 + }, + { + "epoch": 1.2796247922108763, + "grad_norm": 2.9922525882720947, + "learning_rate": 1.2794320123092248e-05, + "loss": 1.4447, + "step": 32400 + }, + { + "epoch": 1.2835826802818016, + "grad_norm": 2.9650540351867676, + "learning_rate": 1.2780987476678072e-05, + "loss": 1.4492, + "step": 32500 + }, + { + "epoch": 1.287540568352727, + "grad_norm": 3.2158126831054688, + "learning_rate": 1.2767621648109765e-05, + "loss": 1.428, + "step": 32600 + }, + { + "epoch": 1.2914984564236525, + "grad_norm": 3.462463617324829, + "learning_rate": 1.275422272136916e-05, + "loss": 1.4355, + "step": 32700 + }, + { + "epoch": 1.2954563444945777, + "grad_norm": 2.9805209636688232, + "learning_rate": 1.2740790780646048e-05, + "loss": 1.4153, + "step": 32800 + }, + { + "epoch": 1.299414232565503, + "grad_norm": 3.3206562995910645, + "learning_rate": 1.2727325910337665e-05, + "loss": 1.425, + "step": 32900 + }, + { + "epoch": 1.3033721206364284, + "grad_norm": 3.306009531021118, + "learning_rate": 1.2713828195048149e-05, + "loss": 1.4398, + "step": 33000 + }, + { + "epoch": 1.3073300087073538, + "grad_norm": 3.6652069091796875, + "learning_rate": 1.2700297719588015e-05, + "loss": 1.4126, + "step": 33100 + }, + { + "epoch": 1.311287896778279, + "grad_norm": 3.067331314086914, + "learning_rate": 1.268673456897362e-05, + "loss": 1.4453, + "step": 33200 + }, + { + "epoch": 1.3152457848492045, + "grad_norm": 3.4072649478912354, + "learning_rate": 1.2673138828426633e-05, + "loss": 1.4195, + "step": 33300 + }, + { + "epoch": 1.3192036729201297, + "grad_norm": 3.721276044845581, + "learning_rate": 1.2659510583373492e-05, + "loss": 1.4308, + "step": 33400 + }, + { + "epoch": 1.3231615609910552, + "grad_norm": 3.381657361984253, + "learning_rate": 1.2645849919444875e-05, + "loss": 1.4102, + "step": 33500 + }, + { + "epoch": 1.3271194490619806, + "grad_norm": 3.2021045684814453, + "learning_rate": 1.2632156922475153e-05, + "loss": 1.4248, + "step": 33600 + }, + { + "epoch": 1.3310773371329059, + "grad_norm": 3.0373260974884033, + "learning_rate": 1.2618431678501862e-05, + "loss": 1.4156, + "step": 33700 + }, + { + "epoch": 1.335035225203831, + "grad_norm": 2.7702269554138184, + "learning_rate": 1.2604674273765154e-05, + "loss": 1.4511, + "step": 33800 + }, + { + "epoch": 1.3389931132747566, + "grad_norm": 3.3153131008148193, + "learning_rate": 1.2590884794707254e-05, + "loss": 1.4523, + "step": 33900 + }, + { + "epoch": 1.342951001345682, + "grad_norm": 3.249516248703003, + "learning_rate": 1.2577063327971927e-05, + "loss": 1.4225, + "step": 34000 + }, + { + "epoch": 1.3469088894166072, + "grad_norm": 3.879835367202759, + "learning_rate": 1.2563209960403921e-05, + "loss": 1.4248, + "step": 34100 + }, + { + "epoch": 1.3508667774875327, + "grad_norm": 3.548116445541382, + "learning_rate": 1.2549324779048432e-05, + "loss": 1.4248, + "step": 34200 + }, + { + "epoch": 1.354824665558458, + "grad_norm": 3.109065294265747, + "learning_rate": 1.253540787115055e-05, + "loss": 1.4269, + "step": 34300 + }, + { + "epoch": 1.3587825536293834, + "grad_norm": 3.3330225944519043, + "learning_rate": 1.2521459324154708e-05, + "loss": 1.4354, + "step": 34400 + }, + { + "epoch": 1.3627404417003088, + "grad_norm": 3.0380284786224365, + "learning_rate": 1.2507479225704149e-05, + "loss": 1.3966, + "step": 34500 + }, + { + "epoch": 1.366698329771234, + "grad_norm": 3.3855364322662354, + "learning_rate": 1.2493467663640356e-05, + "loss": 1.402, + "step": 34600 + }, + { + "epoch": 1.3706562178421593, + "grad_norm": 3.3429582118988037, + "learning_rate": 1.247942472600251e-05, + "loss": 1.4315, + "step": 34700 + }, + { + "epoch": 1.3746141059130847, + "grad_norm": 3.0505242347717285, + "learning_rate": 1.2465350501026931e-05, + "loss": 1.425, + "step": 34800 + }, + { + "epoch": 1.3785719939840102, + "grad_norm": 3.4241063594818115, + "learning_rate": 1.245124507714654e-05, + "loss": 1.4179, + "step": 34900 + }, + { + "epoch": 1.3825298820549354, + "grad_norm": 3.458108901977539, + "learning_rate": 1.2437108542990274e-05, + "loss": 1.4133, + "step": 35000 + }, + { + "epoch": 1.3864877701258609, + "grad_norm": 3.9022340774536133, + "learning_rate": 1.2422940987382556e-05, + "loss": 1.4112, + "step": 35100 + }, + { + "epoch": 1.3904456581967861, + "grad_norm": 3.5141968727111816, + "learning_rate": 1.240874249934273e-05, + "loss": 1.4453, + "step": 35200 + }, + { + "epoch": 1.3944035462677116, + "grad_norm": 3.4254074096679688, + "learning_rate": 1.2394513168084485e-05, + "loss": 1.4096, + "step": 35300 + }, + { + "epoch": 1.398361434338637, + "grad_norm": 3.460205316543579, + "learning_rate": 1.2380253083015321e-05, + "loss": 1.4145, + "step": 35400 + }, + { + "epoch": 1.4023193224095623, + "grad_norm": 3.7515103816986084, + "learning_rate": 1.236596233373597e-05, + "loss": 1.4132, + "step": 35500 + }, + { + "epoch": 1.4062772104804875, + "grad_norm": 3.2565503120422363, + "learning_rate": 1.2351641010039833e-05, + "loss": 1.3945, + "step": 35600 + }, + { + "epoch": 1.410235098551413, + "grad_norm": 3.754737138748169, + "learning_rate": 1.2337289201912429e-05, + "loss": 1.381, + "step": 35700 + }, + { + "epoch": 1.4141929866223384, + "grad_norm": 3.7933449745178223, + "learning_rate": 1.2322906999530811e-05, + "loss": 1.3943, + "step": 35800 + }, + { + "epoch": 1.4181508746932636, + "grad_norm": 3.108177900314331, + "learning_rate": 1.2308494493263014e-05, + "loss": 1.4127, + "step": 35900 + }, + { + "epoch": 1.422108762764189, + "grad_norm": 3.393486499786377, + "learning_rate": 1.2294051773667482e-05, + "loss": 1.3921, + "step": 36000 + }, + { + "epoch": 1.4260666508351143, + "grad_norm": 3.9485793113708496, + "learning_rate": 1.22795789314925e-05, + "loss": 1.42, + "step": 36100 + }, + { + "epoch": 1.4300245389060398, + "grad_norm": 3.353940725326538, + "learning_rate": 1.2265076057675615e-05, + "loss": 1.412, + "step": 36200 + }, + { + "epoch": 1.4339824269769652, + "grad_norm": 3.861928939819336, + "learning_rate": 1.2250543243343082e-05, + "loss": 1.3952, + "step": 36300 + }, + { + "epoch": 1.4379403150478904, + "grad_norm": 2.9782791137695312, + "learning_rate": 1.2235980579809283e-05, + "loss": 1.3872, + "step": 36400 + }, + { + "epoch": 1.4418982031188157, + "grad_norm": 3.552558660507202, + "learning_rate": 1.2221388158576142e-05, + "loss": 1.3855, + "step": 36500 + }, + { + "epoch": 1.4458560911897411, + "grad_norm": 3.034158229827881, + "learning_rate": 1.2206766071332568e-05, + "loss": 1.4028, + "step": 36600 + }, + { + "epoch": 1.4498139792606666, + "grad_norm": 3.669677495956421, + "learning_rate": 1.219211440995387e-05, + "loss": 1.3865, + "step": 36700 + }, + { + "epoch": 1.4537718673315918, + "grad_norm": 3.4838759899139404, + "learning_rate": 1.2177433266501182e-05, + "loss": 1.3987, + "step": 36800 + }, + { + "epoch": 1.4577297554025173, + "grad_norm": 3.4691314697265625, + "learning_rate": 1.2162722733220877e-05, + "loss": 1.3939, + "step": 36900 + }, + { + "epoch": 1.4616876434734425, + "grad_norm": 3.5910284519195557, + "learning_rate": 1.2147982902544004e-05, + "loss": 1.3972, + "step": 37000 + }, + { + "epoch": 1.465645531544368, + "grad_norm": 3.2121059894561768, + "learning_rate": 1.2133213867085686e-05, + "loss": 1.3762, + "step": 37100 + }, + { + "epoch": 1.4696034196152934, + "grad_norm": 3.8289687633514404, + "learning_rate": 1.2118415719644557e-05, + "loss": 1.4032, + "step": 37200 + }, + { + "epoch": 1.4735613076862186, + "grad_norm": 3.597191095352173, + "learning_rate": 1.2103588553202167e-05, + "loss": 1.3925, + "step": 37300 + }, + { + "epoch": 1.4775191957571439, + "grad_norm": 3.4253151416778564, + "learning_rate": 1.2088732460922407e-05, + "loss": 1.3715, + "step": 37400 + }, + { + "epoch": 1.4814770838280693, + "grad_norm": 3.624340772628784, + "learning_rate": 1.2073847536150912e-05, + "loss": 1.387, + "step": 37500 + }, + { + "epoch": 1.4854349718989948, + "grad_norm": 3.2783761024475098, + "learning_rate": 1.2058933872414484e-05, + "loss": 1.3837, + "step": 37600 + }, + { + "epoch": 1.48939285996992, + "grad_norm": 3.377274990081787, + "learning_rate": 1.2043991563420501e-05, + "loss": 1.3515, + "step": 37700 + }, + { + "epoch": 1.4933507480408454, + "grad_norm": 3.676497459411621, + "learning_rate": 1.2029020703056327e-05, + "loss": 1.3647, + "step": 37800 + }, + { + "epoch": 1.4973086361117707, + "grad_norm": 4.441483020782471, + "learning_rate": 1.2014021385388727e-05, + "loss": 1.3594, + "step": 37900 + }, + { + "epoch": 1.5012665241826961, + "grad_norm": 4.011296272277832, + "learning_rate": 1.1998993704663267e-05, + "loss": 1.3855, + "step": 38000 + }, + { + "epoch": 1.5052244122536216, + "grad_norm": 3.3986339569091797, + "learning_rate": 1.1983937755303735e-05, + "loss": 1.3549, + "step": 38100 + }, + { + "epoch": 1.5091823003245468, + "grad_norm": 3.2860589027404785, + "learning_rate": 1.1968853631911532e-05, + "loss": 1.3481, + "step": 38200 + }, + { + "epoch": 1.513140188395472, + "grad_norm": 4.623264789581299, + "learning_rate": 1.1953741429265089e-05, + "loss": 1.3609, + "step": 38300 + }, + { + "epoch": 1.5170980764663975, + "grad_norm": 3.3357603549957275, + "learning_rate": 1.1938601242319269e-05, + "loss": 1.373, + "step": 38400 + }, + { + "epoch": 1.521055964537323, + "grad_norm": 3.6516709327697754, + "learning_rate": 1.1923433166204768e-05, + "loss": 1.3568, + "step": 38500 + }, + { + "epoch": 1.5250138526082484, + "grad_norm": 4.045721530914307, + "learning_rate": 1.1908237296227522e-05, + "loss": 1.3419, + "step": 38600 + }, + { + "epoch": 1.5289717406791736, + "grad_norm": 3.331871271133423, + "learning_rate": 1.1893013727868098e-05, + "loss": 1.3575, + "step": 38700 + }, + { + "epoch": 1.5329296287500989, + "grad_norm": 3.959519624710083, + "learning_rate": 1.1877762556781109e-05, + "loss": 1.3464, + "step": 38800 + }, + { + "epoch": 1.5368875168210243, + "grad_norm": 4.424190998077393, + "learning_rate": 1.1862483878794596e-05, + "loss": 1.3593, + "step": 38900 + }, + { + "epoch": 1.5408454048919498, + "grad_norm": 4.052654266357422, + "learning_rate": 1.1847177789909441e-05, + "loss": 1.3474, + "step": 39000 + }, + { + "epoch": 1.544803292962875, + "grad_norm": 3.552598237991333, + "learning_rate": 1.1831844386298758e-05, + "loss": 1.346, + "step": 39100 + }, + { + "epoch": 1.5487611810338002, + "grad_norm": 3.979213237762451, + "learning_rate": 1.1816483764307286e-05, + "loss": 1.3557, + "step": 39200 + }, + { + "epoch": 1.5527190691047257, + "grad_norm": 4.073390960693359, + "learning_rate": 1.1801096020450786e-05, + "loss": 1.3658, + "step": 39300 + }, + { + "epoch": 1.5566769571756511, + "grad_norm": 4.211179256439209, + "learning_rate": 1.1785681251415431e-05, + "loss": 1.346, + "step": 39400 + }, + { + "epoch": 1.5606348452465766, + "grad_norm": 3.6185340881347656, + "learning_rate": 1.177023955405721e-05, + "loss": 1.3686, + "step": 39500 + }, + { + "epoch": 1.5645927333175018, + "grad_norm": 3.7389111518859863, + "learning_rate": 1.1754771025401307e-05, + "loss": 1.3536, + "step": 39600 + }, + { + "epoch": 1.568550621388427, + "grad_norm": 4.2574357986450195, + "learning_rate": 1.1739275762641494e-05, + "loss": 1.352, + "step": 39700 + }, + { + "epoch": 1.5725085094593525, + "grad_norm": 3.516805410385132, + "learning_rate": 1.1723753863139529e-05, + "loss": 1.3411, + "step": 39800 + }, + { + "epoch": 1.576466397530278, + "grad_norm": 3.5958383083343506, + "learning_rate": 1.1708205424424521e-05, + "loss": 1.3433, + "step": 39900 + }, + { + "epoch": 1.5804242856012032, + "grad_norm": 3.995814323425293, + "learning_rate": 1.1692630544192354e-05, + "loss": 1.3529, + "step": 40000 + }, + { + "epoch": 1.5843821736721284, + "grad_norm": 3.817218780517578, + "learning_rate": 1.1677029320305041e-05, + "loss": 1.3469, + "step": 40100 + }, + { + "epoch": 1.5883400617430539, + "grad_norm": 4.439276695251465, + "learning_rate": 1.1661401850790119e-05, + "loss": 1.3466, + "step": 40200 + }, + { + "epoch": 1.5922979498139793, + "grad_norm": 3.7181553840637207, + "learning_rate": 1.1645748233840044e-05, + "loss": 1.3476, + "step": 40300 + }, + { + "epoch": 1.5962558378849048, + "grad_norm": 10.218334197998047, + "learning_rate": 1.1630068567811557e-05, + "loss": 1.3602, + "step": 40400 + }, + { + "epoch": 1.60021372595583, + "grad_norm": 4.133950710296631, + "learning_rate": 1.1614362951225075e-05, + "loss": 1.3485, + "step": 40500 + }, + { + "epoch": 1.6041716140267552, + "grad_norm": 4.007839202880859, + "learning_rate": 1.1598631482764074e-05, + "loss": 1.3594, + "step": 40600 + }, + { + "epoch": 1.6081295020976807, + "grad_norm": 4.194820404052734, + "learning_rate": 1.1582874261274463e-05, + "loss": 1.3383, + "step": 40700 + }, + { + "epoch": 1.6120873901686061, + "grad_norm": 4.193638801574707, + "learning_rate": 1.1567091385763965e-05, + "loss": 1.3715, + "step": 40800 + }, + { + "epoch": 1.6160452782395314, + "grad_norm": 3.5800745487213135, + "learning_rate": 1.15512829554015e-05, + "loss": 1.3494, + "step": 40900 + }, + { + "epoch": 1.6200031663104566, + "grad_norm": 4.548177719116211, + "learning_rate": 1.1535449069516552e-05, + "loss": 1.3442, + "step": 41000 + }, + { + "epoch": 1.623961054381382, + "grad_norm": 3.7496066093444824, + "learning_rate": 1.1519589827598553e-05, + "loss": 1.329, + "step": 41100 + }, + { + "epoch": 1.6279189424523075, + "grad_norm": 4.815052032470703, + "learning_rate": 1.1503705329296252e-05, + "loss": 1.32, + "step": 41200 + }, + { + "epoch": 1.631876830523233, + "grad_norm": 4.1927103996276855, + "learning_rate": 1.14877956744171e-05, + "loss": 1.3172, + "step": 41300 + }, + { + "epoch": 1.6358347185941582, + "grad_norm": 3.9431440830230713, + "learning_rate": 1.1471860962926604e-05, + "loss": 1.3271, + "step": 41400 + }, + { + "epoch": 1.6397926066650834, + "grad_norm": 4.615567684173584, + "learning_rate": 1.1455901294947722e-05, + "loss": 1.348, + "step": 41500 + }, + { + "epoch": 1.6437504947360089, + "grad_norm": 4.151221752166748, + "learning_rate": 1.143991677076021e-05, + "loss": 1.3336, + "step": 41600 + }, + { + "epoch": 1.6477083828069343, + "grad_norm": 4.409358978271484, + "learning_rate": 1.142390749080001e-05, + "loss": 1.3254, + "step": 41700 + }, + { + "epoch": 1.6516662708778596, + "grad_norm": 4.490970134735107, + "learning_rate": 1.140787355565861e-05, + "loss": 1.3303, + "step": 41800 + }, + { + "epoch": 1.6556241589487848, + "grad_norm": 4.116312026977539, + "learning_rate": 1.1391815066082418e-05, + "loss": 1.3315, + "step": 41900 + }, + { + "epoch": 1.6595820470197102, + "grad_norm": 4.251399517059326, + "learning_rate": 1.1375732122972124e-05, + "loss": 1.3243, + "step": 42000 + }, + { + "epoch": 1.6635399350906357, + "grad_norm": 4.184506416320801, + "learning_rate": 1.1359624827382062e-05, + "loss": 1.3226, + "step": 42100 + }, + { + "epoch": 1.6674978231615611, + "grad_norm": 4.628664970397949, + "learning_rate": 1.134349328051959e-05, + "loss": 1.3399, + "step": 42200 + }, + { + "epoch": 1.6714557112324864, + "grad_norm": 4.529860496520996, + "learning_rate": 1.132733758374444e-05, + "loss": 1.2967, + "step": 42300 + }, + { + "epoch": 1.6754135993034116, + "grad_norm": 4.5048699378967285, + "learning_rate": 1.1311157838568083e-05, + "loss": 1.3255, + "step": 42400 + }, + { + "epoch": 1.679371487374337, + "grad_norm": 4.321528911590576, + "learning_rate": 1.1294954146653094e-05, + "loss": 1.311, + "step": 42500 + }, + { + "epoch": 1.6833293754452625, + "grad_norm": 4.919022083282471, + "learning_rate": 1.1278726609812523e-05, + "loss": 1.3219, + "step": 42600 + }, + { + "epoch": 1.6872872635161877, + "grad_norm": 4.146111965179443, + "learning_rate": 1.126247533000923e-05, + "loss": 1.298, + "step": 42700 + }, + { + "epoch": 1.691245151587113, + "grad_norm": 4.071747779846191, + "learning_rate": 1.1246200409355271e-05, + "loss": 1.313, + "step": 42800 + }, + { + "epoch": 1.6952030396580384, + "grad_norm": 3.8871426582336426, + "learning_rate": 1.1229901950111245e-05, + "loss": 1.3176, + "step": 42900 + }, + { + "epoch": 1.6991609277289639, + "grad_norm": 3.9479401111602783, + "learning_rate": 1.1213580054685644e-05, + "loss": 1.3112, + "step": 43000 + }, + { + "epoch": 1.7031188157998893, + "grad_norm": 4.039346694946289, + "learning_rate": 1.1197234825634222e-05, + "loss": 1.3109, + "step": 43100 + }, + { + "epoch": 1.7070767038708146, + "grad_norm": 4.356393814086914, + "learning_rate": 1.1180866365659346e-05, + "loss": 1.3202, + "step": 43200 + }, + { + "epoch": 1.7110345919417398, + "grad_norm": 4.11832857131958, + "learning_rate": 1.1164474777609351e-05, + "loss": 1.313, + "step": 43300 + }, + { + "epoch": 1.7149924800126652, + "grad_norm": 4.101129531860352, + "learning_rate": 1.1148060164477887e-05, + "loss": 1.2968, + "step": 43400 + }, + { + "epoch": 1.7189503680835907, + "grad_norm": 3.728778600692749, + "learning_rate": 1.1131622629403289e-05, + "loss": 1.2869, + "step": 43500 + }, + { + "epoch": 1.722908256154516, + "grad_norm": 3.846654176712036, + "learning_rate": 1.1115162275667909e-05, + "loss": 1.3054, + "step": 43600 + }, + { + "epoch": 1.7268661442254412, + "grad_norm": 4.967803478240967, + "learning_rate": 1.1098679206697474e-05, + "loss": 1.3165, + "step": 43700 + }, + { + "epoch": 1.7308240322963666, + "grad_norm": 4.610755443572998, + "learning_rate": 1.1082173526060454e-05, + "loss": 1.3234, + "step": 43800 + }, + { + "epoch": 1.734781920367292, + "grad_norm": 4.377742290496826, + "learning_rate": 1.1065645337467375e-05, + "loss": 1.3261, + "step": 43900 + }, + { + "epoch": 1.7387398084382175, + "grad_norm": 5.010995864868164, + "learning_rate": 1.1049094744770201e-05, + "loss": 1.3026, + "step": 44000 + }, + { + "epoch": 1.7426976965091427, + "grad_norm": 4.31913423538208, + "learning_rate": 1.1032521851961665e-05, + "loss": 1.2697, + "step": 44100 + }, + { + "epoch": 1.746655584580068, + "grad_norm": 4.2657060623168945, + "learning_rate": 1.1015926763174617e-05, + "loss": 1.3101, + "step": 44200 + }, + { + "epoch": 1.7506134726509934, + "grad_norm": 3.801684617996216, + "learning_rate": 1.0999309582681372e-05, + "loss": 1.268, + "step": 44300 + }, + { + "epoch": 1.7545713607219189, + "grad_norm": 4.505929946899414, + "learning_rate": 1.0982670414893057e-05, + "loss": 1.2957, + "step": 44400 + }, + { + "epoch": 1.7585292487928441, + "grad_norm": 3.837562084197998, + "learning_rate": 1.0966009364358948e-05, + "loss": 1.3078, + "step": 44500 + }, + { + "epoch": 1.7624871368637693, + "grad_norm": 4.854923248291016, + "learning_rate": 1.0949326535765823e-05, + "loss": 1.3119, + "step": 44600 + }, + { + "epoch": 1.7664450249346948, + "grad_norm": 5.621912002563477, + "learning_rate": 1.0932622033937294e-05, + "loss": 1.2847, + "step": 44700 + }, + { + "epoch": 1.7704029130056202, + "grad_norm": 4.009350776672363, + "learning_rate": 1.0915895963833152e-05, + "loss": 1.3006, + "step": 44800 + }, + { + "epoch": 1.7743608010765457, + "grad_norm": 3.6888113021850586, + "learning_rate": 1.0899148430548716e-05, + "loss": 1.3032, + "step": 44900 + }, + { + "epoch": 1.778318689147471, + "grad_norm": 4.511534690856934, + "learning_rate": 1.0882379539314155e-05, + "loss": 1.2615, + "step": 45000 + }, + { + "epoch": 1.7822765772183962, + "grad_norm": 5.086874485015869, + "learning_rate": 1.0865589395493845e-05, + "loss": 1.2634, + "step": 45100 + }, + { + "epoch": 1.7862344652893216, + "grad_norm": 4.865400791168213, + "learning_rate": 1.0848778104585692e-05, + "loss": 1.2876, + "step": 45200 + }, + { + "epoch": 1.790192353360247, + "grad_norm": 4.537430763244629, + "learning_rate": 1.0831945772220487e-05, + "loss": 1.2764, + "step": 45300 + }, + { + "epoch": 1.7941502414311723, + "grad_norm": 4.448334693908691, + "learning_rate": 1.0815092504161214e-05, + "loss": 1.2814, + "step": 45400 + }, + { + "epoch": 1.7981081295020975, + "grad_norm": 4.650451183319092, + "learning_rate": 1.0798218406302422e-05, + "loss": 1.2819, + "step": 45500 + }, + { + "epoch": 1.802066017573023, + "grad_norm": 4.031219005584717, + "learning_rate": 1.0781323584669524e-05, + "loss": 1.2729, + "step": 45600 + }, + { + "epoch": 1.8060239056439484, + "grad_norm": 4.477336883544922, + "learning_rate": 1.0764408145418157e-05, + "loss": 1.2586, + "step": 45700 + }, + { + "epoch": 1.8099817937148739, + "grad_norm": 4.118893146514893, + "learning_rate": 1.0747472194833506e-05, + "loss": 1.2591, + "step": 45800 + }, + { + "epoch": 1.8139396817857991, + "grad_norm": 4.766265392303467, + "learning_rate": 1.073051583932963e-05, + "loss": 1.2693, + "step": 45900 + }, + { + "epoch": 1.8178975698567243, + "grad_norm": 5.545733451843262, + "learning_rate": 1.0713539185448795e-05, + "loss": 1.2691, + "step": 46000 + }, + { + "epoch": 1.8218554579276498, + "grad_norm": 4.723430633544922, + "learning_rate": 1.069654233986082e-05, + "loss": 1.2582, + "step": 46100 + }, + { + "epoch": 1.8258133459985753, + "grad_norm": 4.899106025695801, + "learning_rate": 1.0679525409362387e-05, + "loss": 1.2802, + "step": 46200 + }, + { + "epoch": 1.8297712340695005, + "grad_norm": 4.531938552856445, + "learning_rate": 1.066248850087638e-05, + "loss": 1.2683, + "step": 46300 + }, + { + "epoch": 1.8337291221404257, + "grad_norm": 4.174386501312256, + "learning_rate": 1.0645431721451212e-05, + "loss": 1.2538, + "step": 46400 + }, + { + "epoch": 1.8376870102113512, + "grad_norm": 4.827451229095459, + "learning_rate": 1.0628355178260147e-05, + "loss": 1.2571, + "step": 46500 + }, + { + "epoch": 1.8416448982822766, + "grad_norm": 4.749929904937744, + "learning_rate": 1.0611258978600638e-05, + "loss": 1.2759, + "step": 46600 + }, + { + "epoch": 1.845602786353202, + "grad_norm": 4.659051418304443, + "learning_rate": 1.0594143229893643e-05, + "loss": 1.2698, + "step": 46700 + }, + { + "epoch": 1.8495606744241273, + "grad_norm": 4.492525100708008, + "learning_rate": 1.057700803968295e-05, + "loss": 1.2583, + "step": 46800 + }, + { + "epoch": 1.8535185624950525, + "grad_norm": 4.04518461227417, + "learning_rate": 1.0559853515634509e-05, + "loss": 1.268, + "step": 46900 + }, + { + "epoch": 1.857476450565978, + "grad_norm": 5.029372215270996, + "learning_rate": 1.054267976553575e-05, + "loss": 1.238, + "step": 47000 + }, + { + "epoch": 1.8614343386369034, + "grad_norm": 4.29434871673584, + "learning_rate": 1.05254868972949e-05, + "loss": 1.2431, + "step": 47100 + }, + { + "epoch": 1.8653922267078287, + "grad_norm": 4.874353408813477, + "learning_rate": 1.050827501894032e-05, + "loss": 1.2339, + "step": 47200 + }, + { + "epoch": 1.869350114778754, + "grad_norm": 4.865941047668457, + "learning_rate": 1.0491044238619817e-05, + "loss": 1.2427, + "step": 47300 + }, + { + "epoch": 1.8733080028496794, + "grad_norm": 4.548977375030518, + "learning_rate": 1.0473794664599957e-05, + "loss": 1.2586, + "step": 47400 + }, + { + "epoch": 1.8772658909206048, + "grad_norm": 5.584561824798584, + "learning_rate": 1.0456526405265402e-05, + "loss": 1.2388, + "step": 47500 + }, + { + "epoch": 1.8812237789915303, + "grad_norm": 4.570620536804199, + "learning_rate": 1.0439239569118215e-05, + "loss": 1.2499, + "step": 47600 + }, + { + "epoch": 1.8851816670624555, + "grad_norm": 4.9521379470825195, + "learning_rate": 1.0421934264777186e-05, + "loss": 1.2486, + "step": 47700 + }, + { + "epoch": 1.8891395551333807, + "grad_norm": 4.744478702545166, + "learning_rate": 1.0404610600977141e-05, + "loss": 1.2428, + "step": 47800 + }, + { + "epoch": 1.8930974432043062, + "grad_norm": 4.681623935699463, + "learning_rate": 1.0387268686568275e-05, + "loss": 1.2577, + "step": 47900 + }, + { + "epoch": 1.8970553312752316, + "grad_norm": 4.406890392303467, + "learning_rate": 1.0369908630515445e-05, + "loss": 1.2367, + "step": 48000 + }, + { + "epoch": 1.9010132193461569, + "grad_norm": 4.650542259216309, + "learning_rate": 1.0352530541897507e-05, + "loss": 1.243, + "step": 48100 + }, + { + "epoch": 1.904971107417082, + "grad_norm": 5.0188164710998535, + "learning_rate": 1.0335134529906619e-05, + "loss": 1.2222, + "step": 48200 + }, + { + "epoch": 1.9089289954880075, + "grad_norm": 4.498706340789795, + "learning_rate": 1.0317720703847554e-05, + "loss": 1.2508, + "step": 48300 + }, + { + "epoch": 1.912886883558933, + "grad_norm": 4.909203052520752, + "learning_rate": 1.0300289173137021e-05, + "loss": 1.2241, + "step": 48400 + }, + { + "epoch": 1.9168447716298584, + "grad_norm": 4.85788631439209, + "learning_rate": 1.0282840047302967e-05, + "loss": 1.2268, + "step": 48500 + }, + { + "epoch": 1.9208026597007837, + "grad_norm": 4.537557601928711, + "learning_rate": 1.0265373435983907e-05, + "loss": 1.2515, + "step": 48600 + }, + { + "epoch": 1.924760547771709, + "grad_norm": 4.660990238189697, + "learning_rate": 1.0247889448928208e-05, + "loss": 1.2531, + "step": 48700 + }, + { + "epoch": 1.9287184358426344, + "grad_norm": 4.9931511878967285, + "learning_rate": 1.0230388195993424e-05, + "loss": 1.2336, + "step": 48800 + }, + { + "epoch": 1.9326763239135598, + "grad_norm": 4.516580581665039, + "learning_rate": 1.0212869787145594e-05, + "loss": 1.2292, + "step": 48900 + }, + { + "epoch": 1.936634211984485, + "grad_norm": 4.5982255935668945, + "learning_rate": 1.0195334332458552e-05, + "loss": 1.2417, + "step": 49000 + }, + { + "epoch": 1.9405921000554103, + "grad_norm": 5.071137428283691, + "learning_rate": 1.0177781942113238e-05, + "loss": 1.2194, + "step": 49100 + }, + { + "epoch": 1.9445499881263357, + "grad_norm": 4.76341438293457, + "learning_rate": 1.0160212726397001e-05, + "loss": 1.2207, + "step": 49200 + }, + { + "epoch": 1.9485078761972612, + "grad_norm": 4.945827484130859, + "learning_rate": 1.0142626795702916e-05, + "loss": 1.208, + "step": 49300 + }, + { + "epoch": 1.9524657642681866, + "grad_norm": 5.068126678466797, + "learning_rate": 1.0125024260529075e-05, + "loss": 1.2372, + "step": 49400 + }, + { + "epoch": 1.9564236523391119, + "grad_norm": 5.238717079162598, + "learning_rate": 1.010740523147791e-05, + "loss": 1.2287, + "step": 49500 + }, + { + "epoch": 1.960381540410037, + "grad_norm": 4.586404800415039, + "learning_rate": 1.008976981925548e-05, + "loss": 1.2023, + "step": 49600 + }, + { + "epoch": 1.9643394284809625, + "grad_norm": 5.284154415130615, + "learning_rate": 1.0072118134670792e-05, + "loss": 1.2254, + "step": 49700 + }, + { + "epoch": 1.968297316551888, + "grad_norm": 4.639484882354736, + "learning_rate": 1.0054450288635098e-05, + "loss": 1.2298, + "step": 49800 + }, + { + "epoch": 1.9722552046228132, + "grad_norm": 5.642242908477783, + "learning_rate": 1.003676639216119e-05, + "loss": 1.2196, + "step": 49900 + }, + { + "epoch": 1.9762130926937385, + "grad_norm": 5.445943355560303, + "learning_rate": 1.0019066556362718e-05, + "loss": 1.2035, + "step": 50000 + }, + { + "epoch": 1.980170980764664, + "grad_norm": 5.127743244171143, + "learning_rate": 1.000135089245348e-05, + "loss": 1.2281, + "step": 50100 + }, + { + "epoch": 1.9841288688355894, + "grad_norm": 4.978102684020996, + "learning_rate": 9.98361951174673e-06, + "loss": 1.22, + "step": 50200 + }, + { + "epoch": 1.9880867569065148, + "grad_norm": 4.681839466094971, + "learning_rate": 9.965872525654468e-06, + "loss": 1.2011, + "step": 50300 + }, + { + "epoch": 1.99204464497744, + "grad_norm": 5.249551296234131, + "learning_rate": 9.948110045686763e-06, + "loss": 1.1931, + "step": 50400 + }, + { + "epoch": 1.9960025330483653, + "grad_norm": 4.80012845993042, + "learning_rate": 9.930332183451022e-06, + "loss": 1.2282, + "step": 50500 + }, + { + "epoch": 1.9999604211192907, + "grad_norm": 5.271570205688477, + "learning_rate": 9.912539050651309e-06, + "loss": 1.1956, + "step": 50600 + }, + { + "epoch": 2.0018986590720305, + "grad_norm": 7.154332637786865, + "learning_rate": 9.894730759087638e-06, + "loss": 1.1201, + "step": 50700 + }, + { + "epoch": 2.005854198805427, + "grad_norm": 5.223172187805176, + "learning_rate": 9.876907420655272e-06, + "loss": 1.0108, + "step": 50800 + }, + { + "epoch": 2.0098097385388236, + "grad_norm": 5.335000038146973, + "learning_rate": 9.859069147344016e-06, + "loss": 1.0126, + "step": 50900 + }, + { + "epoch": 2.01376527827222, + "grad_norm": 5.62663459777832, + "learning_rate": 9.841216051237518e-06, + "loss": 1.0281, + "step": 51000 + }, + { + "epoch": 2.0177208180056168, + "grad_norm": 6.090157508850098, + "learning_rate": 9.82334824451256e-06, + "loss": 1.017, + "step": 51100 + }, + { + "epoch": 2.0216763577390133, + "grad_norm": 5.461913108825684, + "learning_rate": 9.805465839438357e-06, + "loss": 1.0429, + "step": 51200 + }, + { + "epoch": 2.02563189747241, + "grad_norm": 4.846314907073975, + "learning_rate": 9.78756894837585e-06, + "loss": 0.9998, + "step": 51300 + }, + { + "epoch": 2.029587437205807, + "grad_norm": 6.22315788269043, + "learning_rate": 9.769657683776997e-06, + "loss": 1.0199, + "step": 51400 + }, + { + "epoch": 2.0335429769392035, + "grad_norm": 5.477888107299805, + "learning_rate": 9.75173215818408e-06, + "loss": 1.0033, + "step": 51500 + }, + { + "epoch": 2.0374985166726, + "grad_norm": 6.461225509643555, + "learning_rate": 9.733792484228973e-06, + "loss": 1.0282, + "step": 51600 + }, + { + "epoch": 2.0414540564059966, + "grad_norm": 6.070066928863525, + "learning_rate": 9.715838774632455e-06, + "loss": 1.0251, + "step": 51700 + }, + { + "epoch": 2.045409596139393, + "grad_norm": 6.459588050842285, + "learning_rate": 9.697871142203504e-06, + "loss": 0.9998, + "step": 51800 + }, + { + "epoch": 2.04936513587279, + "grad_norm": 6.001095771789551, + "learning_rate": 9.679889699838565e-06, + "loss": 0.9961, + "step": 51900 + }, + { + "epoch": 2.0533206756061864, + "grad_norm": 5.671863555908203, + "learning_rate": 9.661894560520868e-06, + "loss": 1.0103, + "step": 52000 + }, + { + "epoch": 2.057276215339583, + "grad_norm": 6.713314056396484, + "learning_rate": 9.643885837319692e-06, + "loss": 1.0084, + "step": 52100 + }, + { + "epoch": 2.0612317550729795, + "grad_norm": 5.7298903465271, + "learning_rate": 9.62586364338968e-06, + "loss": 1.0145, + "step": 52200 + }, + { + "epoch": 2.0651872948063765, + "grad_norm": 6.283379554748535, + "learning_rate": 9.607828091970112e-06, + "loss": 1.0144, + "step": 52300 + }, + { + "epoch": 2.069142834539773, + "grad_norm": 5.741368293762207, + "learning_rate": 9.589779296384196e-06, + "loss": 0.979, + "step": 52400 + }, + { + "epoch": 2.0730983742731697, + "grad_norm": 5.731276035308838, + "learning_rate": 9.571717370038356e-06, + "loss": 0.9804, + "step": 52500 + }, + { + "epoch": 2.0770539140065662, + "grad_norm": 6.783426761627197, + "learning_rate": 9.553642426421525e-06, + "loss": 1.0146, + "step": 52600 + }, + { + "epoch": 2.081009453739963, + "grad_norm": 5.622731685638428, + "learning_rate": 9.535554579104428e-06, + "loss": 0.9786, + "step": 52700 + }, + { + "epoch": 2.0849649934733594, + "grad_norm": 5.445565700531006, + "learning_rate": 9.517453941738864e-06, + "loss": 1.0033, + "step": 52800 + }, + { + "epoch": 2.088920533206756, + "grad_norm": 6.861485958099365, + "learning_rate": 9.499340628056999e-06, + "loss": 1.0041, + "step": 52900 + }, + { + "epoch": 2.0928760729401525, + "grad_norm": 6.059487342834473, + "learning_rate": 9.481214751870645e-06, + "loss": 1.0032, + "step": 53000 + }, + { + "epoch": 2.096831612673549, + "grad_norm": 6.398611068725586, + "learning_rate": 9.463076427070551e-06, + "loss": 1.0016, + "step": 53100 + }, + { + "epoch": 2.100787152406946, + "grad_norm": 6.830140590667725, + "learning_rate": 9.444925767625685e-06, + "loss": 0.9879, + "step": 53200 + }, + { + "epoch": 2.1047426921403427, + "grad_norm": 6.739964008331299, + "learning_rate": 9.42676288758252e-06, + "loss": 0.9824, + "step": 53300 + }, + { + "epoch": 2.1086982318737393, + "grad_norm": 6.853861331939697, + "learning_rate": 9.408587901064308e-06, + "loss": 0.9888, + "step": 53400 + }, + { + "epoch": 2.112653771607136, + "grad_norm": 5.317558765411377, + "learning_rate": 9.390400922270373e-06, + "loss": 0.9797, + "step": 53500 + }, + { + "epoch": 2.1166093113405324, + "grad_norm": 6.174235820770264, + "learning_rate": 9.372202065475395e-06, + "loss": 0.9894, + "step": 53600 + }, + { + "epoch": 2.120564851073929, + "grad_norm": 5.993900299072266, + "learning_rate": 9.353991445028679e-06, + "loss": 0.9936, + "step": 53700 + }, + { + "epoch": 2.1245203908073256, + "grad_norm": 6.061595916748047, + "learning_rate": 9.335769175353455e-06, + "loss": 0.974, + "step": 53800 + }, + { + "epoch": 2.128475930540722, + "grad_norm": 6.07938289642334, + "learning_rate": 9.317535370946135e-06, + "loss": 0.9915, + "step": 53900 + }, + { + "epoch": 2.1324314702741187, + "grad_norm": 5.920003890991211, + "learning_rate": 9.299290146375627e-06, + "loss": 0.9919, + "step": 54000 + }, + { + "epoch": 2.1363870100075157, + "grad_norm": 5.969104290008545, + "learning_rate": 9.281033616282572e-06, + "loss": 0.9942, + "step": 54100 + }, + { + "epoch": 2.1403425497409123, + "grad_norm": 6.8970947265625, + "learning_rate": 9.262765895378667e-06, + "loss": 0.9804, + "step": 54200 + }, + { + "epoch": 2.144298089474309, + "grad_norm": 5.78402042388916, + "learning_rate": 9.244487098445915e-06, + "loss": 0.9709, + "step": 54300 + }, + { + "epoch": 2.1482536292077055, + "grad_norm": 6.274438381195068, + "learning_rate": 9.226197340335915e-06, + "loss": 0.9823, + "step": 54400 + }, + { + "epoch": 2.152209168941102, + "grad_norm": 6.251999378204346, + "learning_rate": 9.20789673596914e-06, + "loss": 0.988, + "step": 54500 + }, + { + "epoch": 2.1561647086744986, + "grad_norm": 6.099167346954346, + "learning_rate": 9.189585400334208e-06, + "loss": 0.9768, + "step": 54600 + }, + { + "epoch": 2.160120248407895, + "grad_norm": 7.657092094421387, + "learning_rate": 9.171263448487175e-06, + "loss": 0.9743, + "step": 54700 + }, + { + "epoch": 2.1640757881412918, + "grad_norm": 6.438418388366699, + "learning_rate": 9.152930995550795e-06, + "loss": 0.9871, + "step": 54800 + }, + { + "epoch": 2.1680313278746883, + "grad_norm": 6.303830623626709, + "learning_rate": 9.134588156713804e-06, + "loss": 0.9851, + "step": 54900 + }, + { + "epoch": 2.1719868676080853, + "grad_norm": 6.700922966003418, + "learning_rate": 9.116235047230198e-06, + "loss": 0.9758, + "step": 55000 + }, + { + "epoch": 2.175942407341482, + "grad_norm": 7.142539978027344, + "learning_rate": 9.097871782418505e-06, + "loss": 0.9637, + "step": 55100 + }, + { + "epoch": 2.1798979470748785, + "grad_norm": 4.764535903930664, + "learning_rate": 9.079498477661064e-06, + "loss": 1.0001, + "step": 55200 + }, + { + "epoch": 2.183853486808275, + "grad_norm": 7.237622261047363, + "learning_rate": 9.0611152484033e-06, + "loss": 0.9806, + "step": 55300 + }, + { + "epoch": 2.1878090265416716, + "grad_norm": 5.451415061950684, + "learning_rate": 9.042722210152988e-06, + "loss": 0.9846, + "step": 55400 + }, + { + "epoch": 2.191764566275068, + "grad_norm": 7.0478291511535645, + "learning_rate": 9.02431947847955e-06, + "loss": 0.9526, + "step": 55500 + }, + { + "epoch": 2.195720106008465, + "grad_norm": 6.550572872161865, + "learning_rate": 9.005907169013302e-06, + "loss": 0.957, + "step": 55600 + }, + { + "epoch": 2.1996756457418614, + "grad_norm": 6.440634727478027, + "learning_rate": 8.987485397444744e-06, + "loss": 0.9597, + "step": 55700 + }, + { + "epoch": 2.203631185475258, + "grad_norm": 6.600515842437744, + "learning_rate": 8.969054279523838e-06, + "loss": 0.9765, + "step": 55800 + }, + { + "epoch": 2.207586725208655, + "grad_norm": 6.594021797180176, + "learning_rate": 8.950613931059262e-06, + "loss": 0.9706, + "step": 55900 + }, + { + "epoch": 2.2115422649420515, + "grad_norm": 6.733046531677246, + "learning_rate": 8.932164467917696e-06, + "loss": 0.9315, + "step": 56000 + }, + { + "epoch": 2.215497804675448, + "grad_norm": 6.6295928955078125, + "learning_rate": 8.913706006023086e-06, + "loss": 0.9543, + "step": 56100 + }, + { + "epoch": 2.2194533444088447, + "grad_norm": 6.274661064147949, + "learning_rate": 8.89523866135593e-06, + "loss": 0.9809, + "step": 56200 + }, + { + "epoch": 2.2234088841422412, + "grad_norm": 6.869567394256592, + "learning_rate": 8.876762549952532e-06, + "loss": 0.9645, + "step": 56300 + }, + { + "epoch": 2.227364423875638, + "grad_norm": 6.911426067352295, + "learning_rate": 8.858277787904283e-06, + "loss": 0.9601, + "step": 56400 + }, + { + "epoch": 2.2313199636090344, + "grad_norm": 5.583986759185791, + "learning_rate": 8.839784491356926e-06, + "loss": 0.9339, + "step": 56500 + }, + { + "epoch": 2.235275503342431, + "grad_norm": 6.394748210906982, + "learning_rate": 8.821282776509828e-06, + "loss": 0.941, + "step": 56600 + }, + { + "epoch": 2.2392310430758275, + "grad_norm": 5.726268291473389, + "learning_rate": 8.802772759615259e-06, + "loss": 0.9332, + "step": 56700 + }, + { + "epoch": 2.243186582809224, + "grad_norm": 7.450072288513184, + "learning_rate": 8.784254556977645e-06, + "loss": 0.9577, + "step": 56800 + }, + { + "epoch": 2.247142122542621, + "grad_norm": 6.079563140869141, + "learning_rate": 8.76572828495285e-06, + "loss": 0.9683, + "step": 56900 + }, + { + "epoch": 2.2510976622760177, + "grad_norm": 7.1676836013793945, + "learning_rate": 8.747194059947439e-06, + "loss": 0.9258, + "step": 57000 + }, + { + "epoch": 2.2550532020094143, + "grad_norm": 6.986599445343018, + "learning_rate": 8.728651998417948e-06, + "loss": 0.9305, + "step": 57100 + }, + { + "epoch": 2.259008741742811, + "grad_norm": 5.925799369812012, + "learning_rate": 8.710102216870154e-06, + "loss": 0.9482, + "step": 57200 + }, + { + "epoch": 2.2629642814762074, + "grad_norm": 6.0704731941223145, + "learning_rate": 8.69154483185834e-06, + "loss": 0.9256, + "step": 57300 + }, + { + "epoch": 2.266919821209604, + "grad_norm": 7.043192386627197, + "learning_rate": 8.672979959984563e-06, + "loss": 0.9564, + "step": 57400 + }, + { + "epoch": 2.2708753609430006, + "grad_norm": 5.858555793762207, + "learning_rate": 8.654407717897925e-06, + "loss": 0.9329, + "step": 57500 + }, + { + "epoch": 2.274830900676397, + "grad_norm": 6.873895168304443, + "learning_rate": 8.635828222293834e-06, + "loss": 0.9305, + "step": 57600 + }, + { + "epoch": 2.2787864404097937, + "grad_norm": 7.290801048278809, + "learning_rate": 8.617241589913279e-06, + "loss": 0.9471, + "step": 57700 + }, + { + "epoch": 2.2827419801431903, + "grad_norm": 6.579754829406738, + "learning_rate": 8.598647937542088e-06, + "loss": 0.925, + "step": 57800 + }, + { + "epoch": 2.2866975198765873, + "grad_norm": 6.567941665649414, + "learning_rate": 8.5800473820102e-06, + "loss": 0.9461, + "step": 57900 + }, + { + "epoch": 2.290653059609984, + "grad_norm": 6.696491718292236, + "learning_rate": 8.561440040190929e-06, + "loss": 0.9557, + "step": 58000 + }, + { + "epoch": 2.2946085993433805, + "grad_norm": 6.369635581970215, + "learning_rate": 8.542826029000221e-06, + "loss": 0.9515, + "step": 58100 + }, + { + "epoch": 2.298564139076777, + "grad_norm": 6.151893138885498, + "learning_rate": 8.524205465395942e-06, + "loss": 0.9202, + "step": 58200 + }, + { + "epoch": 2.3025196788101736, + "grad_norm": 7.51953649520874, + "learning_rate": 8.505578466377119e-06, + "loss": 0.9508, + "step": 58300 + }, + { + "epoch": 2.30647521854357, + "grad_norm": 6.321157932281494, + "learning_rate": 8.486945148983215e-06, + "loss": 0.9225, + "step": 58400 + }, + { + "epoch": 2.3104307582769668, + "grad_norm": 7.385085582733154, + "learning_rate": 8.468305630293403e-06, + "loss": 0.9446, + "step": 58500 + }, + { + "epoch": 2.3143862980103633, + "grad_norm": 7.018087387084961, + "learning_rate": 8.449660027425807e-06, + "loss": 0.9214, + "step": 58600 + }, + { + "epoch": 2.31834183774376, + "grad_norm": 8.585993766784668, + "learning_rate": 8.431008457536794e-06, + "loss": 0.9326, + "step": 58700 + }, + { + "epoch": 2.322297377477157, + "grad_norm": 7.048767566680908, + "learning_rate": 8.412351037820208e-06, + "loss": 0.9418, + "step": 58800 + }, + { + "epoch": 2.3262529172105535, + "grad_norm": 6.425539016723633, + "learning_rate": 8.393687885506664e-06, + "loss": 0.9113, + "step": 58900 + }, + { + "epoch": 2.33020845694395, + "grad_norm": 6.738077163696289, + "learning_rate": 8.375019117862787e-06, + "loss": 0.9182, + "step": 59000 + }, + { + "epoch": 2.3341639966773466, + "grad_norm": 6.531277656555176, + "learning_rate": 8.356344852190489e-06, + "loss": 0.9346, + "step": 59100 + }, + { + "epoch": 2.338119536410743, + "grad_norm": 7.382998943328857, + "learning_rate": 8.337665205826229e-06, + "loss": 0.9214, + "step": 59200 + }, + { + "epoch": 2.34207507614414, + "grad_norm": 8.067983627319336, + "learning_rate": 8.31898029614027e-06, + "loss": 0.9384, + "step": 59300 + }, + { + "epoch": 2.3460306158775364, + "grad_norm": 7.663227558135986, + "learning_rate": 8.300290240535956e-06, + "loss": 0.9214, + "step": 59400 + }, + { + "epoch": 2.349986155610933, + "grad_norm": 7.118049144744873, + "learning_rate": 8.281595156448947e-06, + "loss": 0.9168, + "step": 59500 + }, + { + "epoch": 2.3539416953443295, + "grad_norm": 6.668679714202881, + "learning_rate": 8.262895161346518e-06, + "loss": 0.8854, + "step": 59600 + }, + { + "epoch": 2.3578972350777265, + "grad_norm": 5.939303874969482, + "learning_rate": 8.24419037272679e-06, + "loss": 0.9283, + "step": 59700 + }, + { + "epoch": 2.361852774811123, + "grad_norm": 7.534061431884766, + "learning_rate": 8.225480908118005e-06, + "loss": 0.9209, + "step": 59800 + }, + { + "epoch": 2.3658083145445197, + "grad_norm": 6.419159889221191, + "learning_rate": 8.206766885077787e-06, + "loss": 0.9132, + "step": 59900 + }, + { + "epoch": 2.3697638542779162, + "grad_norm": 6.306091785430908, + "learning_rate": 8.188048421192401e-06, + "loss": 0.9236, + "step": 60000 + }, + { + "epoch": 2.373719394011313, + "grad_norm": 7.05325984954834, + "learning_rate": 8.169325634076016e-06, + "loss": 0.9022, + "step": 60100 + }, + { + "epoch": 2.3776749337447094, + "grad_norm": 7.843636989593506, + "learning_rate": 8.15059864136997e-06, + "loss": 0.8979, + "step": 60200 + }, + { + "epoch": 2.381630473478106, + "grad_norm": 6.075581073760986, + "learning_rate": 8.131867560742015e-06, + "loss": 0.9238, + "step": 60300 + }, + { + "epoch": 2.3855860132115025, + "grad_norm": 8.050814628601074, + "learning_rate": 8.113132509885597e-06, + "loss": 0.8741, + "step": 60400 + }, + { + "epoch": 2.389541552944899, + "grad_norm": 6.963539123535156, + "learning_rate": 8.09439360651911e-06, + "loss": 0.9003, + "step": 60500 + }, + { + "epoch": 2.393497092678296, + "grad_norm": 6.0645365715026855, + "learning_rate": 8.075650968385143e-06, + "loss": 0.8924, + "step": 60600 + }, + { + "epoch": 2.3974526324116927, + "grad_norm": 6.974462032318115, + "learning_rate": 8.056904713249768e-06, + "loss": 0.9095, + "step": 60700 + }, + { + "epoch": 2.4014081721450893, + "grad_norm": 6.542774200439453, + "learning_rate": 8.03815495890177e-06, + "loss": 0.8817, + "step": 60800 + }, + { + "epoch": 2.405363711878486, + "grad_norm": 6.8144426345825195, + "learning_rate": 8.019401823151932e-06, + "loss": 0.8771, + "step": 60900 + }, + { + "epoch": 2.4093192516118824, + "grad_norm": 7.659158229827881, + "learning_rate": 8.00064542383227e-06, + "loss": 0.9049, + "step": 61000 + }, + { + "epoch": 2.413274791345279, + "grad_norm": 8.229734420776367, + "learning_rate": 7.981885878795319e-06, + "loss": 0.8774, + "step": 61100 + }, + { + "epoch": 2.4172303310786756, + "grad_norm": 6.553606033325195, + "learning_rate": 7.963123305913372e-06, + "loss": 0.8937, + "step": 61200 + }, + { + "epoch": 2.421185870812072, + "grad_norm": 6.9676737785339355, + "learning_rate": 7.94435782307775e-06, + "loss": 0.8998, + "step": 61300 + }, + { + "epoch": 2.4251414105454687, + "grad_norm": 7.392206192016602, + "learning_rate": 7.925589548198057e-06, + "loss": 0.8958, + "step": 61400 + }, + { + "epoch": 2.4290969502788657, + "grad_norm": 6.842816352844238, + "learning_rate": 7.906818599201438e-06, + "loss": 0.9044, + "step": 61500 + }, + { + "epoch": 2.4330524900122623, + "grad_norm": 7.204736709594727, + "learning_rate": 7.88804509403185e-06, + "loss": 0.884, + "step": 61600 + }, + { + "epoch": 2.437008029745659, + "grad_norm": 9.083603858947754, + "learning_rate": 7.869269150649295e-06, + "loss": 0.8925, + "step": 61700 + }, + { + "epoch": 2.4409635694790555, + "grad_norm": 6.365993499755859, + "learning_rate": 7.850490887029114e-06, + "loss": 0.8889, + "step": 61800 + }, + { + "epoch": 2.444919109212452, + "grad_norm": 7.185299873352051, + "learning_rate": 7.83171042116121e-06, + "loss": 0.8878, + "step": 61900 + }, + { + "epoch": 2.4488746489458486, + "grad_norm": 6.3879313468933105, + "learning_rate": 7.812927871049331e-06, + "loss": 0.8851, + "step": 62000 + }, + { + "epoch": 2.452830188679245, + "grad_norm": 6.968358993530273, + "learning_rate": 7.794143354710324e-06, + "loss": 0.9103, + "step": 62100 + }, + { + "epoch": 2.4567857284126418, + "grad_norm": 6.729606628417969, + "learning_rate": 7.775356990173385e-06, + "loss": 0.9025, + "step": 62200 + }, + { + "epoch": 2.4607412681460383, + "grad_norm": 7.090347766876221, + "learning_rate": 7.756568895479326e-06, + "loss": 0.8676, + "step": 62300 + }, + { + "epoch": 2.4646968078794353, + "grad_norm": 6.751077651977539, + "learning_rate": 7.737779188679826e-06, + "loss": 0.9007, + "step": 62400 + }, + { + "epoch": 2.468652347612832, + "grad_norm": 6.40106725692749, + "learning_rate": 7.7189879878367e-06, + "loss": 0.8638, + "step": 62500 + }, + { + "epoch": 2.4726078873462285, + "grad_norm": 6.530547142028809, + "learning_rate": 7.700195411021144e-06, + "loss": 0.8462, + "step": 62600 + }, + { + "epoch": 2.476563427079625, + "grad_norm": 7.113738536834717, + "learning_rate": 7.681401576313006e-06, + "loss": 0.8792, + "step": 62700 + }, + { + "epoch": 2.4805189668130216, + "grad_norm": 6.407541751861572, + "learning_rate": 7.66260660180003e-06, + "loss": 0.8497, + "step": 62800 + }, + { + "epoch": 2.484474506546418, + "grad_norm": 6.465631484985352, + "learning_rate": 7.643810605577129e-06, + "loss": 0.8836, + "step": 62900 + }, + { + "epoch": 2.488430046279815, + "grad_norm": 6.723154544830322, + "learning_rate": 7.625013705745631e-06, + "loss": 0.861, + "step": 63000 + }, + { + "epoch": 2.4923855860132114, + "grad_norm": 7.3425374031066895, + "learning_rate": 7.606216020412547e-06, + "loss": 0.8444, + "step": 63100 + }, + { + "epoch": 2.496341125746608, + "grad_norm": 7.729851722717285, + "learning_rate": 7.5874176676898145e-06, + "loss": 0.8566, + "step": 63200 + }, + { + "epoch": 2.500296665480005, + "grad_norm": 7.338516712188721, + "learning_rate": 7.568618765693576e-06, + "loss": 0.8695, + "step": 63300 + }, + { + "epoch": 2.5042522052134015, + "grad_norm": 6.624641418457031, + "learning_rate": 7.549819432543413e-06, + "loss": 0.8575, + "step": 63400 + }, + { + "epoch": 2.508207744946798, + "grad_norm": 7.888009071350098, + "learning_rate": 7.531019786361626e-06, + "loss": 0.8407, + "step": 63500 + }, + { + "epoch": 2.5121632846801947, + "grad_norm": 6.383202075958252, + "learning_rate": 7.512219945272479e-06, + "loss": 0.8558, + "step": 63600 + }, + { + "epoch": 2.5161188244135912, + "grad_norm": 7.635325908660889, + "learning_rate": 7.493420027401461e-06, + "loss": 0.8549, + "step": 63700 + }, + { + "epoch": 2.520074364146988, + "grad_norm": 7.805452346801758, + "learning_rate": 7.474620150874539e-06, + "loss": 0.8518, + "step": 63800 + }, + { + "epoch": 2.5240299038803844, + "grad_norm": 7.43981409072876, + "learning_rate": 7.455820433817427e-06, + "loss": 0.8319, + "step": 63900 + }, + { + "epoch": 2.527985443613781, + "grad_norm": 7.546152114868164, + "learning_rate": 7.437020994354832e-06, + "loss": 0.8696, + "step": 64000 + }, + { + "epoch": 2.5319409833471775, + "grad_norm": 7.283714771270752, + "learning_rate": 7.418221950609721e-06, + "loss": 0.8252, + "step": 64100 + }, + { + "epoch": 2.5358965230805746, + "grad_norm": 6.995747089385986, + "learning_rate": 7.399423420702571e-06, + "loss": 0.8304, + "step": 64200 + }, + { + "epoch": 2.5398520628139707, + "grad_norm": 7.93922233581543, + "learning_rate": 7.380625522750634e-06, + "loss": 0.8077, + "step": 64300 + }, + { + "epoch": 2.5438076025473677, + "grad_norm": 7.082808494567871, + "learning_rate": 7.361828374867182e-06, + "loss": 0.8093, + "step": 64400 + }, + { + "epoch": 2.5477631422807643, + "grad_norm": 7.408179759979248, + "learning_rate": 7.343032095160793e-06, + "loss": 0.8371, + "step": 64500 + }, + { + "epoch": 2.551718682014161, + "grad_norm": 6.96567964553833, + "learning_rate": 7.324236801734566e-06, + "loss": 0.8307, + "step": 64600 + }, + { + "epoch": 2.5556742217475574, + "grad_norm": 6.544528007507324, + "learning_rate": 7.305442612685424e-06, + "loss": 0.8196, + "step": 64700 + }, + { + "epoch": 2.559629761480954, + "grad_norm": 7.655238151550293, + "learning_rate": 7.2866496461033335e-06, + "loss": 0.8378, + "step": 64800 + }, + { + "epoch": 2.5635853012143506, + "grad_norm": 7.466569423675537, + "learning_rate": 7.267858020070598e-06, + "loss": 0.8538, + "step": 64900 + }, + { + "epoch": 2.567540840947747, + "grad_norm": 6.754441261291504, + "learning_rate": 7.249067852661077e-06, + "loss": 0.848, + "step": 65000 + }, + { + "epoch": 2.571496380681144, + "grad_norm": 7.796608924865723, + "learning_rate": 7.230279261939484e-06, + "loss": 0.8364, + "step": 65100 + }, + { + "epoch": 2.5754519204145403, + "grad_norm": 7.588559150695801, + "learning_rate": 7.211492365960615e-06, + "loss": 0.87, + "step": 65200 + }, + { + "epoch": 2.5794074601479373, + "grad_norm": 6.7228312492370605, + "learning_rate": 7.19270728276862e-06, + "loss": 0.8605, + "step": 65300 + }, + { + "epoch": 2.583362999881334, + "grad_norm": 7.157680511474609, + "learning_rate": 7.173924130396258e-06, + "loss": 0.8147, + "step": 65400 + }, + { + "epoch": 2.5873185396147305, + "grad_norm": 7.738386154174805, + "learning_rate": 7.155143026864159e-06, + "loss": 0.8195, + "step": 65500 + }, + { + "epoch": 2.591274079348127, + "grad_norm": 7.59410285949707, + "learning_rate": 7.136364090180074e-06, + "loss": 0.8337, + "step": 65600 + }, + { + "epoch": 2.5952296190815236, + "grad_norm": 6.822688102722168, + "learning_rate": 7.1175874383381455e-06, + "loss": 0.83, + "step": 65700 + }, + { + "epoch": 2.59918515881492, + "grad_norm": 8.452685356140137, + "learning_rate": 7.098813189318153e-06, + "loss": 0.8319, + "step": 65800 + }, + { + "epoch": 2.6031406985483168, + "grad_norm": 6.633552551269531, + "learning_rate": 7.080041461084785e-06, + "loss": 0.8334, + "step": 65900 + }, + { + "epoch": 2.6070962382817138, + "grad_norm": 7.842954635620117, + "learning_rate": 7.061272371586885e-06, + "loss": 0.8062, + "step": 66000 + }, + { + "epoch": 2.61105177801511, + "grad_norm": 6.577658653259277, + "learning_rate": 7.0425060387567215e-06, + "loss": 0.8413, + "step": 66100 + }, + { + "epoch": 2.615007317748507, + "grad_norm": 7.939908027648926, + "learning_rate": 7.023742580509236e-06, + "loss": 0.8127, + "step": 66200 + }, + { + "epoch": 2.6189628574819035, + "grad_norm": 7.662463188171387, + "learning_rate": 7.004982114741316e-06, + "loss": 0.7996, + "step": 66300 + }, + { + "epoch": 2.6229183972153, + "grad_norm": 6.9583635330200195, + "learning_rate": 6.986224759331036e-06, + "loss": 0.8232, + "step": 66400 + }, + { + "epoch": 2.6268739369486966, + "grad_norm": 8.075411796569824, + "learning_rate": 6.967470632136942e-06, + "loss": 0.7982, + "step": 66500 + }, + { + "epoch": 2.630829476682093, + "grad_norm": 7.972960472106934, + "learning_rate": 6.948719850997278e-06, + "loss": 0.836, + "step": 66600 + }, + { + "epoch": 2.63478501641549, + "grad_norm": 7.087239742279053, + "learning_rate": 6.92997253372928e-06, + "loss": 0.8047, + "step": 66700 + }, + { + "epoch": 2.6387405561488864, + "grad_norm": 7.775416374206543, + "learning_rate": 6.911228798128405e-06, + "loss": 0.8307, + "step": 66800 + }, + { + "epoch": 2.6426960958822834, + "grad_norm": 7.944192409515381, + "learning_rate": 6.892488761967621e-06, + "loss": 0.8103, + "step": 66900 + }, + { + "epoch": 2.6466516356156795, + "grad_norm": 7.953451156616211, + "learning_rate": 6.8737525429966335e-06, + "loss": 0.8285, + "step": 67000 + }, + { + "epoch": 2.6506071753490765, + "grad_norm": 9.518111228942871, + "learning_rate": 6.855020258941179e-06, + "loss": 0.807, + "step": 67100 + }, + { + "epoch": 2.654562715082473, + "grad_norm": 7.476140975952148, + "learning_rate": 6.836292027502261e-06, + "loss": 0.7977, + "step": 67200 + }, + { + "epoch": 2.6585182548158697, + "grad_norm": 7.218750953674316, + "learning_rate": 6.8175679663554224e-06, + "loss": 0.8084, + "step": 67300 + }, + { + "epoch": 2.6624737945492662, + "grad_norm": 8.890192031860352, + "learning_rate": 6.798848193149998e-06, + "loss": 0.7993, + "step": 67400 + }, + { + "epoch": 2.666429334282663, + "grad_norm": 8.358793258666992, + "learning_rate": 6.78013282550839e-06, + "loss": 0.8348, + "step": 67500 + }, + { + "epoch": 2.6703848740160594, + "grad_norm": 8.094076156616211, + "learning_rate": 6.761421981025306e-06, + "loss": 0.8185, + "step": 67600 + }, + { + "epoch": 2.674340413749456, + "grad_norm": 8.059944152832031, + "learning_rate": 6.742715777267045e-06, + "loss": 0.8074, + "step": 67700 + }, + { + "epoch": 2.678295953482853, + "grad_norm": 7.6637282371521, + "learning_rate": 6.724014331770736e-06, + "loss": 0.7986, + "step": 67800 + }, + { + "epoch": 2.682251493216249, + "grad_norm": 8.657379150390625, + "learning_rate": 6.705317762043622e-06, + "loss": 0.8139, + "step": 67900 + }, + { + "epoch": 2.686207032949646, + "grad_norm": 7.7766337394714355, + "learning_rate": 6.6866261855623e-06, + "loss": 0.8059, + "step": 68000 + }, + { + "epoch": 2.6901625726830427, + "grad_norm": 7.386256694793701, + "learning_rate": 6.667939719772001e-06, + "loss": 0.7984, + "step": 68100 + }, + { + "epoch": 2.6941181124164393, + "grad_norm": 8.123387336730957, + "learning_rate": 6.649258482085834e-06, + "loss": 0.7747, + "step": 68200 + }, + { + "epoch": 2.698073652149836, + "grad_norm": 7.112607479095459, + "learning_rate": 6.6305825898840686e-06, + "loss": 0.7937, + "step": 68300 + }, + { + "epoch": 2.7020291918832324, + "grad_norm": 8.249123573303223, + "learning_rate": 6.611912160513378e-06, + "loss": 0.7727, + "step": 68400 + }, + { + "epoch": 2.705984731616629, + "grad_norm": 7.522756576538086, + "learning_rate": 6.5932473112861184e-06, + "loss": 0.7639, + "step": 68500 + }, + { + "epoch": 2.7099402713500256, + "grad_norm": 9.555477142333984, + "learning_rate": 6.574588159479576e-06, + "loss": 0.8042, + "step": 68600 + }, + { + "epoch": 2.7138958110834226, + "grad_norm": 6.161262035369873, + "learning_rate": 6.555934822335251e-06, + "loss": 0.8121, + "step": 68700 + }, + { + "epoch": 2.7178513508168187, + "grad_norm": 8.5349702835083, + "learning_rate": 6.53728741705809e-06, + "loss": 0.7859, + "step": 68800 + }, + { + "epoch": 2.7218068905502157, + "grad_norm": 9.312299728393555, + "learning_rate": 6.518646060815787e-06, + "loss": 0.7879, + "step": 68900 + }, + { + "epoch": 2.7257624302836123, + "grad_norm": 6.545658111572266, + "learning_rate": 6.50001087073801e-06, + "loss": 0.7874, + "step": 69000 + }, + { + "epoch": 2.729717970017009, + "grad_norm": 8.577535629272461, + "learning_rate": 6.4813819639157e-06, + "loss": 0.7922, + "step": 69100 + }, + { + "epoch": 2.7336735097504055, + "grad_norm": 8.140532493591309, + "learning_rate": 6.462759457400304e-06, + "loss": 0.7909, + "step": 69200 + }, + { + "epoch": 2.737629049483802, + "grad_norm": 8.126466751098633, + "learning_rate": 6.444143468203064e-06, + "loss": 0.7711, + "step": 69300 + }, + { + "epoch": 2.7415845892171986, + "grad_norm": 8.574898719787598, + "learning_rate": 6.425534113294266e-06, + "loss": 0.7845, + "step": 69400 + }, + { + "epoch": 2.745540128950595, + "grad_norm": 8.131915092468262, + "learning_rate": 6.4069315096025075e-06, + "loss": 0.7602, + "step": 69500 + }, + { + "epoch": 2.749495668683992, + "grad_norm": 7.594697952270508, + "learning_rate": 6.388335774013979e-06, + "loss": 0.7758, + "step": 69600 + }, + { + "epoch": 2.7534512084173883, + "grad_norm": 7.075917720794678, + "learning_rate": 6.369747023371698e-06, + "loss": 0.775, + "step": 69700 + }, + { + "epoch": 2.7574067481507853, + "grad_norm": 6.975893974304199, + "learning_rate": 6.351165374474812e-06, + "loss": 0.7949, + "step": 69800 + }, + { + "epoch": 2.761362287884182, + "grad_norm": 8.883713722229004, + "learning_rate": 6.332590944077832e-06, + "loss": 0.7902, + "step": 69900 + }, + { + "epoch": 2.7653178276175785, + "grad_norm": 7.061893939971924, + "learning_rate": 6.31402384888992e-06, + "loss": 0.7679, + "step": 70000 + }, + { + "epoch": 2.769273367350975, + "grad_norm": 7.214616775512695, + "learning_rate": 6.295464205574145e-06, + "loss": 0.7628, + "step": 70100 + }, + { + "epoch": 2.7732289070843716, + "grad_norm": 8.735739707946777, + "learning_rate": 6.276912130746758e-06, + "loss": 0.7693, + "step": 70200 + }, + { + "epoch": 2.777184446817768, + "grad_norm": 7.975681781768799, + "learning_rate": 6.258367740976447e-06, + "loss": 0.7804, + "step": 70300 + }, + { + "epoch": 2.781139986551165, + "grad_norm": 7.6719255447387695, + "learning_rate": 6.239831152783622e-06, + "loss": 0.7647, + "step": 70400 + }, + { + "epoch": 2.7850955262845614, + "grad_norm": 7.42200231552124, + "learning_rate": 6.221302482639665e-06, + "loss": 0.7633, + "step": 70500 + }, + { + "epoch": 2.789051066017958, + "grad_norm": 6.804473400115967, + "learning_rate": 6.202781846966211e-06, + "loss": 0.7689, + "step": 70600 + }, + { + "epoch": 2.793006605751355, + "grad_norm": 9.550134658813477, + "learning_rate": 6.184269362134409e-06, + "loss": 0.7762, + "step": 70700 + }, + { + "epoch": 2.7969621454847515, + "grad_norm": 8.442462921142578, + "learning_rate": 6.165765144464196e-06, + "loss": 0.752, + "step": 70800 + }, + { + "epoch": 2.800917685218148, + "grad_norm": 7.215641975402832, + "learning_rate": 6.147269310223561e-06, + "loss": 0.7751, + "step": 70900 + }, + { + "epoch": 2.8048732249515447, + "grad_norm": 7.342108249664307, + "learning_rate": 6.128781975627822e-06, + "loss": 0.7838, + "step": 71000 + }, + { + "epoch": 2.8088287646849412, + "grad_norm": 7.887964248657227, + "learning_rate": 6.110303256838884e-06, + "loss": 0.755, + "step": 71100 + }, + { + "epoch": 2.812784304418338, + "grad_norm": 7.857334613800049, + "learning_rate": 6.091833269964524e-06, + "loss": 0.7606, + "step": 71200 + }, + { + "epoch": 2.8167398441517344, + "grad_norm": 6.394157886505127, + "learning_rate": 6.073372131057644e-06, + "loss": 0.7729, + "step": 71300 + }, + { + "epoch": 2.820695383885131, + "grad_norm": 7.473947048187256, + "learning_rate": 6.054919956115565e-06, + "loss": 0.7462, + "step": 71400 + }, + { + "epoch": 2.8246509236185275, + "grad_norm": 7.744419574737549, + "learning_rate": 6.036476861079267e-06, + "loss": 0.7493, + "step": 71500 + }, + { + "epoch": 2.8286064633519246, + "grad_norm": 7.414424419403076, + "learning_rate": 6.018042961832697e-06, + "loss": 0.7574, + "step": 71600 + }, + { + "epoch": 2.832562003085321, + "grad_norm": 9.066329956054688, + "learning_rate": 5.999618374202004e-06, + "loss": 0.7342, + "step": 71700 + }, + { + "epoch": 2.8365175428187177, + "grad_norm": 8.319313049316406, + "learning_rate": 5.981203213954842e-06, + "loss": 0.7503, + "step": 71800 + }, + { + "epoch": 2.8404730825521143, + "grad_norm": 7.850893020629883, + "learning_rate": 5.962797596799623e-06, + "loss": 0.7589, + "step": 71900 + }, + { + "epoch": 2.844428622285511, + "grad_norm": 7.3950300216674805, + "learning_rate": 5.9444016383848016e-06, + "loss": 0.7339, + "step": 72000 + }, + { + "epoch": 2.8483841620189074, + "grad_norm": 7.9359450340271, + "learning_rate": 5.926015454298135e-06, + "loss": 0.7339, + "step": 72100 + }, + { + "epoch": 2.852339701752304, + "grad_norm": 7.991437911987305, + "learning_rate": 5.907639160065978e-06, + "loss": 0.7539, + "step": 72200 + }, + { + "epoch": 2.8562952414857006, + "grad_norm": 8.999034881591797, + "learning_rate": 5.88927287115253e-06, + "loss": 0.7305, + "step": 72300 + }, + { + "epoch": 2.860250781219097, + "grad_norm": 9.365317344665527, + "learning_rate": 5.870916702959137e-06, + "loss": 0.7632, + "step": 72400 + }, + { + "epoch": 2.864206320952494, + "grad_norm": 7.4115824699401855, + "learning_rate": 5.852570770823541e-06, + "loss": 0.7425, + "step": 72500 + }, + { + "epoch": 2.8681618606858907, + "grad_norm": 8.937634468078613, + "learning_rate": 5.834235190019177e-06, + "loss": 0.7133, + "step": 72600 + }, + { + "epoch": 2.8721174004192873, + "grad_norm": 7.646848201751709, + "learning_rate": 5.8159100757544335e-06, + "loss": 0.7018, + "step": 72700 + }, + { + "epoch": 2.876072940152684, + "grad_norm": 10.104166984558105, + "learning_rate": 5.797595543171939e-06, + "loss": 0.7148, + "step": 72800 + }, + { + "epoch": 2.8800284798860805, + "grad_norm": 8.80380630493164, + "learning_rate": 5.779291707347829e-06, + "loss": 0.7148, + "step": 72900 + }, + { + "epoch": 2.883984019619477, + "grad_norm": 9.618269920349121, + "learning_rate": 5.760998683291034e-06, + "loss": 0.7501, + "step": 73000 + }, + { + "epoch": 2.8879395593528736, + "grad_norm": 8.776264190673828, + "learning_rate": 5.742716585942541e-06, + "loss": 0.7338, + "step": 73100 + }, + { + "epoch": 2.89189509908627, + "grad_norm": 8.551417350769043, + "learning_rate": 5.724445530174691e-06, + "loss": 0.7408, + "step": 73200 + }, + { + "epoch": 2.8958506388196668, + "grad_norm": 7.699435710906982, + "learning_rate": 5.706185630790439e-06, + "loss": 0.7334, + "step": 73300 + }, + { + "epoch": 2.8998061785530638, + "grad_norm": 7.05002498626709, + "learning_rate": 5.687937002522653e-06, + "loss": 0.7347, + "step": 73400 + }, + { + "epoch": 2.9037617182864603, + "grad_norm": 7.351695537567139, + "learning_rate": 5.669699760033361e-06, + "loss": 0.7302, + "step": 73500 + }, + { + "epoch": 2.907717258019857, + "grad_norm": 8.033913612365723, + "learning_rate": 5.651474017913075e-06, + "loss": 0.7161, + "step": 73600 + }, + { + "epoch": 2.9116727977532535, + "grad_norm": 8.220582962036133, + "learning_rate": 5.633259890680022e-06, + "loss": 0.6886, + "step": 73700 + }, + { + "epoch": 2.91562833748665, + "grad_norm": 7.48123025894165, + "learning_rate": 5.615057492779469e-06, + "loss": 0.7375, + "step": 73800 + }, + { + "epoch": 2.9195838772200466, + "grad_norm": 9.332476615905762, + "learning_rate": 5.59686693858297e-06, + "loss": 0.7198, + "step": 73900 + }, + { + "epoch": 2.923539416953443, + "grad_norm": 7.856235027313232, + "learning_rate": 5.57868834238767e-06, + "loss": 0.7097, + "step": 74000 + }, + { + "epoch": 2.92749495668684, + "grad_norm": 8.680730819702148, + "learning_rate": 5.560521818415573e-06, + "loss": 0.711, + "step": 74100 + }, + { + "epoch": 2.9314504964202364, + "grad_norm": 8.170439720153809, + "learning_rate": 5.542367480812831e-06, + "loss": 0.717, + "step": 74200 + }, + { + "epoch": 2.9354060361536334, + "grad_norm": 9.730953216552734, + "learning_rate": 5.524225443649023e-06, + "loss": 0.698, + "step": 74300 + }, + { + "epoch": 2.9393615758870295, + "grad_norm": 9.266618728637695, + "learning_rate": 5.5060958209164475e-06, + "loss": 0.7083, + "step": 74400 + }, + { + "epoch": 2.9433171156204265, + "grad_norm": 8.5849027633667, + "learning_rate": 5.487978726529389e-06, + "loss": 0.7114, + "step": 74500 + }, + { + "epoch": 2.947272655353823, + "grad_norm": 8.211541175842285, + "learning_rate": 5.46987427432342e-06, + "loss": 0.7097, + "step": 74600 + }, + { + "epoch": 2.9512281950872197, + "grad_norm": 7.857696533203125, + "learning_rate": 5.451782578054672e-06, + "loss": 0.7058, + "step": 74700 + }, + { + "epoch": 2.9551837348206162, + "grad_norm": 9.360848426818848, + "learning_rate": 5.433703751399134e-06, + "loss": 0.7026, + "step": 74800 + }, + { + "epoch": 2.959139274554013, + "grad_norm": 7.426985740661621, + "learning_rate": 5.415637907951924e-06, + "loss": 0.7183, + "step": 74900 + }, + { + "epoch": 2.9630948142874094, + "grad_norm": 8.63637638092041, + "learning_rate": 5.397585161226589e-06, + "loss": 0.7005, + "step": 75000 + }, + { + "epoch": 2.967050354020806, + "grad_norm": 8.272201538085938, + "learning_rate": 5.379545624654378e-06, + "loss": 0.6982, + "step": 75100 + }, + { + "epoch": 2.971005893754203, + "grad_norm": 7.2338433265686035, + "learning_rate": 5.3615194115835435e-06, + "loss": 0.7259, + "step": 75200 + }, + { + "epoch": 2.974961433487599, + "grad_norm": 8.305835723876953, + "learning_rate": 5.343506635278613e-06, + "loss": 0.7132, + "step": 75300 + }, + { + "epoch": 2.978916973220996, + "grad_norm": 8.030599594116211, + "learning_rate": 5.325507408919701e-06, + "loss": 0.6689, + "step": 75400 + }, + { + "epoch": 2.9828725129543927, + "grad_norm": 7.884495735168457, + "learning_rate": 5.3075218456017635e-06, + "loss": 0.7073, + "step": 75500 + }, + { + "epoch": 2.9868280526877893, + "grad_norm": 7.318825721740723, + "learning_rate": 5.289550058333928e-06, + "loss": 0.7116, + "step": 75600 + }, + { + "epoch": 2.990783592421186, + "grad_norm": 7.598905563354492, + "learning_rate": 5.271592160038745e-06, + "loss": 0.6911, + "step": 75700 + }, + { + "epoch": 2.9947391321545824, + "grad_norm": 8.233808517456055, + "learning_rate": 5.2536482635515085e-06, + "loss": 0.7179, + "step": 75800 + }, + { + "epoch": 2.998694671887979, + "grad_norm": 10.12309455871582, + "learning_rate": 5.235718481619525e-06, + "loss": 0.6761, + "step": 75900 + }, + { + "epoch": 3.0006328112640404, + "grad_norm": 8.129044532775879, + "learning_rate": 5.217802926901425e-06, + "loss": 0.6679, + "step": 76000 + }, + { + "epoch": 3.0045878816642935, + "grad_norm": 7.123943328857422, + "learning_rate": 5.199901711966435e-06, + "loss": 0.547, + "step": 76100 + }, + { + "epoch": 3.0085429520645466, + "grad_norm": 7.461557388305664, + "learning_rate": 5.182014949293688e-06, + "loss": 0.5426, + "step": 76200 + }, + { + "epoch": 3.0124980224647997, + "grad_norm": 10.335936546325684, + "learning_rate": 5.164142751271501e-06, + "loss": 0.5431, + "step": 76300 + }, + { + "epoch": 3.016453092865053, + "grad_norm": 10.355412483215332, + "learning_rate": 5.146285230196684e-06, + "loss": 0.5642, + "step": 76400 + }, + { + "epoch": 3.020408163265306, + "grad_norm": 7.330338001251221, + "learning_rate": 5.12844249827382e-06, + "loss": 0.5149, + "step": 76500 + }, + { + "epoch": 3.024363233665559, + "grad_norm": 8.793408393859863, + "learning_rate": 5.1106146676145755e-06, + "loss": 0.5409, + "step": 76600 + }, + { + "epoch": 3.028318304065812, + "grad_norm": 10.123501777648926, + "learning_rate": 5.092801850236977e-06, + "loss": 0.553, + "step": 76700 + }, + { + "epoch": 3.0322733744660657, + "grad_norm": 7.840044021606445, + "learning_rate": 5.075004158064728e-06, + "loss": 0.5339, + "step": 76800 + }, + { + "epoch": 3.036228444866319, + "grad_norm": 7.7024736404418945, + "learning_rate": 5.057221702926488e-06, + "loss": 0.5595, + "step": 76900 + }, + { + "epoch": 3.040183515266572, + "grad_norm": 8.08385944366455, + "learning_rate": 5.039454596555181e-06, + "loss": 0.5555, + "step": 77000 + }, + { + "epoch": 3.044138585666825, + "grad_norm": 9.490301132202148, + "learning_rate": 5.021702950587289e-06, + "loss": 0.5487, + "step": 77100 + }, + { + "epoch": 3.048093656067078, + "grad_norm": 8.39875316619873, + "learning_rate": 5.00396687656215e-06, + "loss": 0.5362, + "step": 77200 + }, + { + "epoch": 3.0520487264673313, + "grad_norm": 10.81480884552002, + "learning_rate": 4.986246485921261e-06, + "loss": 0.5416, + "step": 77300 + }, + { + "epoch": 3.0560037968675844, + "grad_norm": 7.355183124542236, + "learning_rate": 4.968541890007572e-06, + "loss": 0.5241, + "step": 77400 + }, + { + "epoch": 3.0599588672678375, + "grad_norm": 7.786324977874756, + "learning_rate": 4.950853200064792e-06, + "loss": 0.5428, + "step": 77500 + }, + { + "epoch": 3.0639139376680906, + "grad_norm": 8.595235824584961, + "learning_rate": 4.933180527236687e-06, + "loss": 0.5429, + "step": 77600 + }, + { + "epoch": 3.0678690080683437, + "grad_norm": 7.885761260986328, + "learning_rate": 4.915523982566383e-06, + "loss": 0.553, + "step": 77700 + }, + { + "epoch": 3.071824078468597, + "grad_norm": 9.978317260742188, + "learning_rate": 4.897883676995664e-06, + "loss": 0.5847, + "step": 77800 + }, + { + "epoch": 3.07577914886885, + "grad_norm": 9.08287239074707, + "learning_rate": 4.8802597213642856e-06, + "loss": 0.5586, + "step": 77900 + }, + { + "epoch": 3.079734219269103, + "grad_norm": 9.484689712524414, + "learning_rate": 4.862652226409261e-06, + "loss": 0.5278, + "step": 78000 + }, + { + "epoch": 3.083689289669356, + "grad_norm": 9.68605899810791, + "learning_rate": 4.845061302764191e-06, + "loss": 0.5358, + "step": 78100 + }, + { + "epoch": 3.0876443600696093, + "grad_norm": 10.46710205078125, + "learning_rate": 4.827487060958533e-06, + "loss": 0.507, + "step": 78200 + }, + { + "epoch": 3.0915994304698624, + "grad_norm": 9.789128303527832, + "learning_rate": 4.80992961141695e-06, + "loss": 0.5263, + "step": 78300 + }, + { + "epoch": 3.0955545008701155, + "grad_norm": 8.19245719909668, + "learning_rate": 4.792389064458569e-06, + "loss": 0.5488, + "step": 78400 + }, + { + "epoch": 3.0995095712703686, + "grad_norm": 8.210286140441895, + "learning_rate": 4.774865530296338e-06, + "loss": 0.5138, + "step": 78500 + }, + { + "epoch": 3.1034646416706217, + "grad_norm": 11.341084480285645, + "learning_rate": 4.757359119036288e-06, + "loss": 0.5524, + "step": 78600 + }, + { + "epoch": 3.107419712070875, + "grad_norm": 8.606145858764648, + "learning_rate": 4.739869940676872e-06, + "loss": 0.5401, + "step": 78700 + }, + { + "epoch": 3.111374782471128, + "grad_norm": 8.024489402770996, + "learning_rate": 4.722398105108257e-06, + "loss": 0.541, + "step": 78800 + }, + { + "epoch": 3.115329852871381, + "grad_norm": 6.819253444671631, + "learning_rate": 4.704943722111649e-06, + "loss": 0.5267, + "step": 78900 + }, + { + "epoch": 3.119284923271634, + "grad_norm": 11.054219245910645, + "learning_rate": 4.6875069013585795e-06, + "loss": 0.5078, + "step": 79000 + }, + { + "epoch": 3.1232399936718873, + "grad_norm": 8.629249572753906, + "learning_rate": 4.670087752410245e-06, + "loss": 0.55, + "step": 79100 + }, + { + "epoch": 3.1271950640721404, + "grad_norm": 10.402203559875488, + "learning_rate": 4.652686384716795e-06, + "loss": 0.5301, + "step": 79200 + }, + { + "epoch": 3.1311501344723935, + "grad_norm": 9.260662078857422, + "learning_rate": 4.635302907616659e-06, + "loss": 0.5051, + "step": 79300 + }, + { + "epoch": 3.1351052048726467, + "grad_norm": 9.835843086242676, + "learning_rate": 4.617937430335848e-06, + "loss": 0.517, + "step": 79400 + }, + { + "epoch": 3.1390602752728998, + "grad_norm": 8.685378074645996, + "learning_rate": 4.600590061987283e-06, + "loss": 0.5537, + "step": 79500 + }, + { + "epoch": 3.143015345673153, + "grad_norm": 4.874628067016602, + "learning_rate": 4.583260911570091e-06, + "loss": 0.5297, + "step": 79600 + }, + { + "epoch": 3.146970416073406, + "grad_norm": 10.263830184936523, + "learning_rate": 4.5659500879689395e-06, + "loss": 0.5084, + "step": 79700 + }, + { + "epoch": 3.150925486473659, + "grad_norm": 8.627128601074219, + "learning_rate": 4.548657699953333e-06, + "loss": 0.5186, + "step": 79800 + }, + { + "epoch": 3.1548805568739122, + "grad_norm": 12.04641056060791, + "learning_rate": 4.531383856176949e-06, + "loss": 0.511, + "step": 79900 + }, + { + "epoch": 3.1588356272741653, + "grad_norm": 9.611827850341797, + "learning_rate": 4.514128665176935e-06, + "loss": 0.5326, + "step": 80000 + }, + { + "epoch": 3.1627906976744184, + "grad_norm": 8.870594024658203, + "learning_rate": 4.49689223537325e-06, + "loss": 0.5208, + "step": 80100 + }, + { + "epoch": 3.1667457680746716, + "grad_norm": 9.362154006958008, + "learning_rate": 4.479674675067954e-06, + "loss": 0.5293, + "step": 80200 + }, + { + "epoch": 3.1707008384749247, + "grad_norm": 8.815319061279297, + "learning_rate": 4.462476092444563e-06, + "loss": 0.5224, + "step": 80300 + }, + { + "epoch": 3.174655908875178, + "grad_norm": 8.474082946777344, + "learning_rate": 4.44529659556733e-06, + "loss": 0.5282, + "step": 80400 + }, + { + "epoch": 3.178610979275431, + "grad_norm": 8.675308227539062, + "learning_rate": 4.428136292380605e-06, + "loss": 0.5292, + "step": 80500 + }, + { + "epoch": 3.1825660496756845, + "grad_norm": 8.218753814697266, + "learning_rate": 4.410995290708121e-06, + "loss": 0.5235, + "step": 80600 + }, + { + "epoch": 3.1865211200759376, + "grad_norm": 9.475656509399414, + "learning_rate": 4.3938736982523485e-06, + "loss": 0.5016, + "step": 80700 + }, + { + "epoch": 3.1904761904761907, + "grad_norm": 6.521944522857666, + "learning_rate": 4.3767716225937895e-06, + "loss": 0.5392, + "step": 80800 + }, + { + "epoch": 3.194431260876444, + "grad_norm": 11.293815612792969, + "learning_rate": 4.359689171190329e-06, + "loss": 0.5172, + "step": 80900 + }, + { + "epoch": 3.198386331276697, + "grad_norm": 8.898575782775879, + "learning_rate": 4.342626451376533e-06, + "loss": 0.4907, + "step": 81000 + }, + { + "epoch": 3.20234140167695, + "grad_norm": 9.884123802185059, + "learning_rate": 4.325583570363003e-06, + "loss": 0.494, + "step": 81100 + }, + { + "epoch": 3.206296472077203, + "grad_norm": 9.265763282775879, + "learning_rate": 4.308560635235669e-06, + "loss": 0.4984, + "step": 81200 + }, + { + "epoch": 3.2102515424774563, + "grad_norm": 9.925707817077637, + "learning_rate": 4.2915577529551505e-06, + "loss": 0.4789, + "step": 81300 + }, + { + "epoch": 3.2142066128777094, + "grad_norm": 8.641393661499023, + "learning_rate": 4.274575030356057e-06, + "loss": 0.5038, + "step": 81400 + }, + { + "epoch": 3.2181616832779625, + "grad_norm": 8.077740669250488, + "learning_rate": 4.257612574146341e-06, + "loss": 0.5143, + "step": 81500 + }, + { + "epoch": 3.2221167536782156, + "grad_norm": 7.956730365753174, + "learning_rate": 4.240670490906596e-06, + "loss": 0.5002, + "step": 81600 + }, + { + "epoch": 3.2260718240784687, + "grad_norm": 9.40575122833252, + "learning_rate": 4.2237488870894245e-06, + "loss": 0.5153, + "step": 81700 + }, + { + "epoch": 3.230026894478722, + "grad_norm": 11.024624824523926, + "learning_rate": 4.2068478690187355e-06, + "loss": 0.5146, + "step": 81800 + }, + { + "epoch": 3.233981964878975, + "grad_norm": 10.030312538146973, + "learning_rate": 4.189967542889106e-06, + "loss": 0.4978, + "step": 81900 + }, + { + "epoch": 3.237937035279228, + "grad_norm": 8.483134269714355, + "learning_rate": 4.173108014765084e-06, + "loss": 0.5021, + "step": 82000 + }, + { + "epoch": 3.241892105679481, + "grad_norm": 9.623600959777832, + "learning_rate": 4.156269390580546e-06, + "loss": 0.5053, + "step": 82100 + }, + { + "epoch": 3.2458471760797343, + "grad_norm": 9.081612586975098, + "learning_rate": 4.139451776138015e-06, + "loss": 0.5038, + "step": 82200 + }, + { + "epoch": 3.2498022464799874, + "grad_norm": 7.760807991027832, + "learning_rate": 4.122655277108014e-06, + "loss": 0.4946, + "step": 82300 + }, + { + "epoch": 3.2537573168802405, + "grad_norm": 9.095376014709473, + "learning_rate": 4.1058799990283814e-06, + "loss": 0.4955, + "step": 82400 + }, + { + "epoch": 3.2577123872804936, + "grad_norm": 7.644027233123779, + "learning_rate": 4.08912604730362e-06, + "loss": 0.4959, + "step": 82500 + }, + { + "epoch": 3.2616674576807467, + "grad_norm": 10.785470962524414, + "learning_rate": 4.072393527204232e-06, + "loss": 0.5038, + "step": 82600 + }, + { + "epoch": 3.265622528081, + "grad_norm": 10.49952220916748, + "learning_rate": 4.055682543866063e-06, + "loss": 0.5116, + "step": 82700 + }, + { + "epoch": 3.269577598481253, + "grad_norm": 11.771068572998047, + "learning_rate": 4.038993202289629e-06, + "loss": 0.4978, + "step": 82800 + }, + { + "epoch": 3.273532668881506, + "grad_norm": 8.821813583374023, + "learning_rate": 4.022325607339466e-06, + "loss": 0.4768, + "step": 82900 + }, + { + "epoch": 3.277487739281759, + "grad_norm": 13.049178123474121, + "learning_rate": 4.005679863743468e-06, + "loss": 0.5041, + "step": 83000 + }, + { + "epoch": 3.2814428096820123, + "grad_norm": 9.644794464111328, + "learning_rate": 3.989056076092235e-06, + "loss": 0.5031, + "step": 83100 + }, + { + "epoch": 3.2853978800822654, + "grad_norm": 7.407175540924072, + "learning_rate": 3.972454348838405e-06, + "loss": 0.4883, + "step": 83200 + }, + { + "epoch": 3.2893529504825185, + "grad_norm": 13.028958320617676, + "learning_rate": 3.95587478629601e-06, + "loss": 0.4805, + "step": 83300 + }, + { + "epoch": 3.2933080208827716, + "grad_norm": 7.644301414489746, + "learning_rate": 3.9393174926398005e-06, + "loss": 0.5093, + "step": 83400 + }, + { + "epoch": 3.2972630912830247, + "grad_norm": 8.546706199645996, + "learning_rate": 3.922782571904621e-06, + "loss": 0.4667, + "step": 83500 + }, + { + "epoch": 3.301218161683278, + "grad_norm": 10.010725975036621, + "learning_rate": 3.906270127984725e-06, + "loss": 0.5127, + "step": 83600 + }, + { + "epoch": 3.305173232083531, + "grad_norm": 12.253199577331543, + "learning_rate": 3.889780264633153e-06, + "loss": 0.4845, + "step": 83700 + }, + { + "epoch": 3.309128302483784, + "grad_norm": 7.043243408203125, + "learning_rate": 3.873313085461048e-06, + "loss": 0.5092, + "step": 83800 + }, + { + "epoch": 3.313083372884037, + "grad_norm": 9.297669410705566, + "learning_rate": 3.856868693937032e-06, + "loss": 0.4827, + "step": 83900 + }, + { + "epoch": 3.3170384432842903, + "grad_norm": 8.122245788574219, + "learning_rate": 3.840447193386537e-06, + "loss": 0.4863, + "step": 84000 + }, + { + "epoch": 3.3209935136845434, + "grad_norm": 8.875086784362793, + "learning_rate": 3.8240486869911725e-06, + "loss": 0.5004, + "step": 84100 + }, + { + "epoch": 3.3249485840847965, + "grad_norm": 8.587276458740234, + "learning_rate": 3.807673277788062e-06, + "loss": 0.4888, + "step": 84200 + }, + { + "epoch": 3.3289036544850497, + "grad_norm": 10.221160888671875, + "learning_rate": 3.791321068669202e-06, + "loss": 0.4717, + "step": 84300 + }, + { + "epoch": 3.3328587248853028, + "grad_norm": 8.61902141571045, + "learning_rate": 3.7749921623808154e-06, + "loss": 0.5057, + "step": 84400 + }, + { + "epoch": 3.336813795285556, + "grad_norm": 9.123618125915527, + "learning_rate": 3.758686661522711e-06, + "loss": 0.4895, + "step": 84500 + }, + { + "epoch": 3.340768865685809, + "grad_norm": 10.211472511291504, + "learning_rate": 3.7424046685476288e-06, + "loss": 0.5143, + "step": 84600 + }, + { + "epoch": 3.344723936086062, + "grad_norm": 8.340943336486816, + "learning_rate": 3.726146285760601e-06, + "loss": 0.4589, + "step": 84700 + }, + { + "epoch": 3.3486790064863152, + "grad_norm": 8.512166976928711, + "learning_rate": 3.7099116153183117e-06, + "loss": 0.4511, + "step": 84800 + }, + { + "epoch": 3.3526340768865683, + "grad_norm": 7.626883506774902, + "learning_rate": 3.6937007592284475e-06, + "loss": 0.5123, + "step": 84900 + }, + { + "epoch": 3.356589147286822, + "grad_norm": 7.530759334564209, + "learning_rate": 3.677513819349073e-06, + "loss": 0.4814, + "step": 85000 + }, + { + "epoch": 3.360544217687075, + "grad_norm": 11.007448196411133, + "learning_rate": 3.6613508973879676e-06, + "loss": 0.4716, + "step": 85100 + }, + { + "epoch": 3.364499288087328, + "grad_norm": 10.204156875610352, + "learning_rate": 3.6452120949020044e-06, + "loss": 0.4846, + "step": 85200 + }, + { + "epoch": 3.3684543584875812, + "grad_norm": 9.174517631530762, + "learning_rate": 3.629097513296501e-06, + "loss": 0.4722, + "step": 85300 + }, + { + "epoch": 3.3724094288878343, + "grad_norm": 7.779313564300537, + "learning_rate": 3.613007253824597e-06, + "loss": 0.4753, + "step": 85400 + }, + { + "epoch": 3.3763644992880875, + "grad_norm": 9.513938903808594, + "learning_rate": 3.596941417586599e-06, + "loss": 0.4838, + "step": 85500 + }, + { + "epoch": 3.3803195696883406, + "grad_norm": 7.560364723205566, + "learning_rate": 3.5809001055293585e-06, + "loss": 0.4951, + "step": 85600 + }, + { + "epoch": 3.3842746400885937, + "grad_norm": 9.338923454284668, + "learning_rate": 3.5648834184456293e-06, + "loss": 0.4985, + "step": 85700 + }, + { + "epoch": 3.388229710488847, + "grad_norm": 8.588788032531738, + "learning_rate": 3.548891456973448e-06, + "loss": 0.4826, + "step": 85800 + }, + { + "epoch": 3.3921847808891, + "grad_norm": 7.1329545974731445, + "learning_rate": 3.53292432159548e-06, + "loss": 0.4604, + "step": 85900 + }, + { + "epoch": 3.396139851289353, + "grad_norm": 10.372248649597168, + "learning_rate": 3.5169821126384154e-06, + "loss": 0.5027, + "step": 86000 + }, + { + "epoch": 3.400094921689606, + "grad_norm": 8.757039070129395, + "learning_rate": 3.501064930272304e-06, + "loss": 0.5149, + "step": 86100 + }, + { + "epoch": 3.4040499920898593, + "grad_norm": 10.216232299804688, + "learning_rate": 3.485172874509963e-06, + "loss": 0.4668, + "step": 86200 + }, + { + "epoch": 3.4080050624901124, + "grad_norm": 8.485111236572266, + "learning_rate": 3.46930604520632e-06, + "loss": 0.4549, + "step": 86300 + }, + { + "epoch": 3.4119601328903655, + "grad_norm": 8.599973678588867, + "learning_rate": 3.4534645420578053e-06, + "loss": 0.4729, + "step": 86400 + }, + { + "epoch": 3.4159152032906186, + "grad_norm": 10.100400924682617, + "learning_rate": 3.43764846460171e-06, + "loss": 0.4587, + "step": 86500 + }, + { + "epoch": 3.4198702736908717, + "grad_norm": 11.832128524780273, + "learning_rate": 3.4218579122155695e-06, + "loss": 0.4667, + "step": 86600 + }, + { + "epoch": 3.423825344091125, + "grad_norm": 9.493983268737793, + "learning_rate": 3.406092984116533e-06, + "loss": 0.4581, + "step": 86700 + }, + { + "epoch": 3.427780414491378, + "grad_norm": 9.855833053588867, + "learning_rate": 3.3903537793607546e-06, + "loss": 0.4718, + "step": 86800 + }, + { + "epoch": 3.431735484891631, + "grad_norm": 12.606931686401367, + "learning_rate": 3.3746403968427506e-06, + "loss": 0.471, + "step": 86900 + }, + { + "epoch": 3.435690555291884, + "grad_norm": 7.937397480010986, + "learning_rate": 3.3589529352947908e-06, + "loss": 0.4752, + "step": 87000 + }, + { + "epoch": 3.4396456256921373, + "grad_norm": 9.963780403137207, + "learning_rate": 3.3432914932862728e-06, + "loss": 0.4724, + "step": 87100 + }, + { + "epoch": 3.4436006960923904, + "grad_norm": 8.761956214904785, + "learning_rate": 3.3276561692231115e-06, + "loss": 0.4714, + "step": 87200 + }, + { + "epoch": 3.4475557664926435, + "grad_norm": 9.65771198272705, + "learning_rate": 3.312047061347109e-06, + "loss": 0.4505, + "step": 87300 + }, + { + "epoch": 3.4515108368928966, + "grad_norm": 7.428150177001953, + "learning_rate": 3.296464267735344e-06, + "loss": 0.4636, + "step": 87400 + }, + { + "epoch": 3.4554659072931497, + "grad_norm": 8.332456588745117, + "learning_rate": 3.2809078862995496e-06, + "loss": 0.4822, + "step": 87500 + }, + { + "epoch": 3.459420977693403, + "grad_norm": 14.15750789642334, + "learning_rate": 3.265378014785513e-06, + "loss": 0.4496, + "step": 87600 + }, + { + "epoch": 3.463376048093656, + "grad_norm": 8.370489120483398, + "learning_rate": 3.2498747507724433e-06, + "loss": 0.4668, + "step": 87700 + }, + { + "epoch": 3.467331118493909, + "grad_norm": 11.614065170288086, + "learning_rate": 3.234398191672367e-06, + "loss": 0.4625, + "step": 87800 + }, + { + "epoch": 3.471286188894162, + "grad_norm": 7.348938941955566, + "learning_rate": 3.2189484347295134e-06, + "loss": 0.4417, + "step": 87900 + }, + { + "epoch": 3.4752412592944153, + "grad_norm": 11.13962459564209, + "learning_rate": 3.2035255770197124e-06, + "loss": 0.4827, + "step": 88000 + }, + { + "epoch": 3.4791963296946684, + "grad_norm": 6.6790056228637695, + "learning_rate": 3.1881297154497676e-06, + "loss": 0.481, + "step": 88100 + }, + { + "epoch": 3.4831514000949215, + "grad_norm": 8.494084358215332, + "learning_rate": 3.1727609467568688e-06, + "loss": 0.4458, + "step": 88200 + }, + { + "epoch": 3.4871064704951746, + "grad_norm": 8.11501693725586, + "learning_rate": 3.1574193675079564e-06, + "loss": 0.4547, + "step": 88300 + }, + { + "epoch": 3.4910615408954278, + "grad_norm": 9.092681884765625, + "learning_rate": 3.142105074099147e-06, + "loss": 0.476, + "step": 88400 + }, + { + "epoch": 3.4950166112956813, + "grad_norm": 10.274823188781738, + "learning_rate": 3.1268181627550997e-06, + "loss": 0.454, + "step": 88500 + }, + { + "epoch": 3.4989716816959344, + "grad_norm": 11.079930305480957, + "learning_rate": 3.1115587295284326e-06, + "loss": 0.4498, + "step": 88600 + }, + { + "epoch": 3.5029267520961875, + "grad_norm": 8.612648010253906, + "learning_rate": 3.0963268702991023e-06, + "loss": 0.4183, + "step": 88700 + }, + { + "epoch": 3.5068818224964406, + "grad_norm": 9.691105842590332, + "learning_rate": 3.0811226807738126e-06, + "loss": 0.4367, + "step": 88800 + }, + { + "epoch": 3.5108368928966938, + "grad_norm": 10.901847839355469, + "learning_rate": 3.065946256485406e-06, + "loss": 0.4764, + "step": 88900 + }, + { + "epoch": 3.514791963296947, + "grad_norm": 9.77769660949707, + "learning_rate": 3.050797692792275e-06, + "loss": 0.468, + "step": 89000 + }, + { + "epoch": 3.5187470336972, + "grad_norm": 10.783480644226074, + "learning_rate": 3.035677084877748e-06, + "loss": 0.465, + "step": 89100 + }, + { + "epoch": 3.522702104097453, + "grad_norm": 9.860946655273438, + "learning_rate": 3.0205845277494995e-06, + "loss": 0.4447, + "step": 89200 + }, + { + "epoch": 3.526657174497706, + "grad_norm": 9.348694801330566, + "learning_rate": 3.00552011623895e-06, + "loss": 0.459, + "step": 89300 + }, + { + "epoch": 3.5306122448979593, + "grad_norm": 7.3600358963012695, + "learning_rate": 2.990483945000679e-06, + "loss": 0.4423, + "step": 89400 + }, + { + "epoch": 3.5345673152982124, + "grad_norm": 9.740511894226074, + "learning_rate": 2.975476108511817e-06, + "loss": 0.4435, + "step": 89500 + }, + { + "epoch": 3.5385223856984656, + "grad_norm": 8.59338665008545, + "learning_rate": 2.9604967010714583e-06, + "loss": 0.43, + "step": 89600 + }, + { + "epoch": 3.5424774560987187, + "grad_norm": 11.468289375305176, + "learning_rate": 2.9455458168000685e-06, + "loss": 0.4368, + "step": 89700 + }, + { + "epoch": 3.546432526498972, + "grad_norm": 9.191219329833984, + "learning_rate": 2.9306235496388993e-06, + "loss": 0.4495, + "step": 89800 + }, + { + "epoch": 3.550387596899225, + "grad_norm": 10.588215827941895, + "learning_rate": 2.91572999334938e-06, + "loss": 0.443, + "step": 89900 + }, + { + "epoch": 3.554342667299478, + "grad_norm": 8.200735092163086, + "learning_rate": 2.9008652415125554e-06, + "loss": 0.4623, + "step": 90000 + }, + { + "epoch": 3.558297737699731, + "grad_norm": 8.150663375854492, + "learning_rate": 2.8860293875284642e-06, + "loss": 0.4423, + "step": 90100 + }, + { + "epoch": 3.5622528080999842, + "grad_norm": 8.815475463867188, + "learning_rate": 2.8712225246155885e-06, + "loss": 0.4275, + "step": 90200 + }, + { + "epoch": 3.5662078785002373, + "grad_norm": 9.670441627502441, + "learning_rate": 2.856444745810235e-06, + "loss": 0.4411, + "step": 90300 + }, + { + "epoch": 3.5701629489004905, + "grad_norm": 11.169974327087402, + "learning_rate": 2.8416961439659787e-06, + "loss": 0.4527, + "step": 90400 + }, + { + "epoch": 3.5741180193007436, + "grad_norm": 9.71986198425293, + "learning_rate": 2.8269768117530577e-06, + "loss": 0.424, + "step": 90500 + }, + { + "epoch": 3.5780730897009967, + "grad_norm": 8.910456657409668, + "learning_rate": 2.8122868416578013e-06, + "loss": 0.4291, + "step": 90600 + }, + { + "epoch": 3.58202816010125, + "grad_norm": 9.892890930175781, + "learning_rate": 2.797626325982046e-06, + "loss": 0.427, + "step": 90700 + }, + { + "epoch": 3.585983230501503, + "grad_norm": 8.138750076293945, + "learning_rate": 2.7829953568425623e-06, + "loss": 0.4279, + "step": 90800 + }, + { + "epoch": 3.589938300901756, + "grad_norm": 10.201644897460938, + "learning_rate": 2.7683940261704656e-06, + "loss": 0.4333, + "step": 90900 + }, + { + "epoch": 3.593893371302009, + "grad_norm": 9.497387886047363, + "learning_rate": 2.753822425710642e-06, + "loss": 0.4275, + "step": 91000 + }, + { + "epoch": 3.5978484417022623, + "grad_norm": 11.225207328796387, + "learning_rate": 2.739280647021174e-06, + "loss": 0.4401, + "step": 91100 + }, + { + "epoch": 3.6018035121025154, + "grad_norm": 8.41219711303711, + "learning_rate": 2.724768781472769e-06, + "loss": 0.4213, + "step": 91200 + }, + { + "epoch": 3.6057585825027685, + "grad_norm": 8.49161148071289, + "learning_rate": 2.7102869202481757e-06, + "loss": 0.4298, + "step": 91300 + }, + { + "epoch": 3.6097136529030216, + "grad_norm": 7.584569931030273, + "learning_rate": 2.695835154341616e-06, + "loss": 0.4455, + "step": 91400 + }, + { + "epoch": 3.6136687233032747, + "grad_norm": 9.672845840454102, + "learning_rate": 2.6814135745582114e-06, + "loss": 0.4161, + "step": 91500 + }, + { + "epoch": 3.617623793703528, + "grad_norm": 11.135618209838867, + "learning_rate": 2.6670222715134223e-06, + "loss": 0.4158, + "step": 91600 + }, + { + "epoch": 3.621578864103781, + "grad_norm": 10.18213176727295, + "learning_rate": 2.652661335632462e-06, + "loss": 0.4273, + "step": 91700 + }, + { + "epoch": 3.625533934504034, + "grad_norm": 9.0409517288208, + "learning_rate": 2.6383308571497417e-06, + "loss": 0.4403, + "step": 91800 + }, + { + "epoch": 3.629489004904287, + "grad_norm": 10.783854484558105, + "learning_rate": 2.6240309261082944e-06, + "loss": 0.4412, + "step": 91900 + }, + { + "epoch": 3.6334440753045403, + "grad_norm": 8.486966133117676, + "learning_rate": 2.609761632359221e-06, + "loss": 0.4397, + "step": 92000 + }, + { + "epoch": 3.6373991457047934, + "grad_norm": 10.486551284790039, + "learning_rate": 2.5955230655611086e-06, + "loss": 0.4348, + "step": 92100 + }, + { + "epoch": 3.6413542161050465, + "grad_norm": 8.568268775939941, + "learning_rate": 2.5813153151794913e-06, + "loss": 0.4267, + "step": 92200 + }, + { + "epoch": 3.6453092865052996, + "grad_norm": 9.572600364685059, + "learning_rate": 2.567138470486257e-06, + "loss": 0.4165, + "step": 92300 + }, + { + "epoch": 3.6492643569055527, + "grad_norm": 9.160146713256836, + "learning_rate": 2.5529926205591102e-06, + "loss": 0.4161, + "step": 92400 + }, + { + "epoch": 3.653219427305806, + "grad_norm": 7.692389488220215, + "learning_rate": 2.538877854281012e-06, + "loss": 0.3944, + "step": 92500 + }, + { + "epoch": 3.657174497706059, + "grad_norm": 10.177444458007812, + "learning_rate": 2.5247942603396016e-06, + "loss": 0.4301, + "step": 92600 + }, + { + "epoch": 3.661129568106312, + "grad_norm": 7.599040508270264, + "learning_rate": 2.5107419272266704e-06, + "loss": 0.4338, + "step": 92700 + }, + { + "epoch": 3.665084638506565, + "grad_norm": 9.309414863586426, + "learning_rate": 2.4967209432375644e-06, + "loss": 0.4286, + "step": 92800 + }, + { + "epoch": 3.6690397089068183, + "grad_norm": 8.994830131530762, + "learning_rate": 2.4827313964706737e-06, + "loss": 0.4082, + "step": 92900 + }, + { + "epoch": 3.6729947793070714, + "grad_norm": 9.818035125732422, + "learning_rate": 2.4687733748268454e-06, + "loss": 0.41, + "step": 93000 + }, + { + "epoch": 3.6769498497073245, + "grad_norm": 9.885952949523926, + "learning_rate": 2.4548469660088533e-06, + "loss": 0.4075, + "step": 93100 + }, + { + "epoch": 3.6809049201075776, + "grad_norm": 8.293312072753906, + "learning_rate": 2.4409522575208308e-06, + "loss": 0.3905, + "step": 93200 + }, + { + "epoch": 3.684859990507831, + "grad_norm": 9.278348922729492, + "learning_rate": 2.4270893366677303e-06, + "loss": 0.4255, + "step": 93300 + }, + { + "epoch": 3.6888150609080843, + "grad_norm": 9.701757431030273, + "learning_rate": 2.4132582905547684e-06, + "loss": 0.3935, + "step": 93400 + }, + { + "epoch": 3.6927701313083374, + "grad_norm": 10.177591323852539, + "learning_rate": 2.3994592060868913e-06, + "loss": 0.4118, + "step": 93500 + }, + { + "epoch": 3.6967252017085905, + "grad_norm": 10.166543960571289, + "learning_rate": 2.3856921699682113e-06, + "loss": 0.4117, + "step": 93600 + }, + { + "epoch": 3.7006802721088436, + "grad_norm": 9.657312393188477, + "learning_rate": 2.371957268701475e-06, + "loss": 0.4194, + "step": 93700 + }, + { + "epoch": 3.7046353425090968, + "grad_norm": 9.794391632080078, + "learning_rate": 2.35825458858751e-06, + "loss": 0.4267, + "step": 93800 + }, + { + "epoch": 3.70859041290935, + "grad_norm": 8.307416915893555, + "learning_rate": 2.344584215724699e-06, + "loss": 0.3905, + "step": 93900 + }, + { + "epoch": 3.712545483309603, + "grad_norm": 9.993391036987305, + "learning_rate": 2.3309462360084153e-06, + "loss": 0.4013, + "step": 94000 + }, + { + "epoch": 3.716500553709856, + "grad_norm": 9.117337226867676, + "learning_rate": 2.3173407351305035e-06, + "loss": 0.4329, + "step": 94100 + }, + { + "epoch": 3.720455624110109, + "grad_norm": 7.99422550201416, + "learning_rate": 2.303767798578727e-06, + "loss": 0.3951, + "step": 94200 + }, + { + "epoch": 3.7244106945103623, + "grad_norm": 7.283392429351807, + "learning_rate": 2.2902275116362455e-06, + "loss": 0.4213, + "step": 94300 + }, + { + "epoch": 3.7283657649106154, + "grad_norm": 7.147285461425781, + "learning_rate": 2.2767199593810643e-06, + "loss": 0.4145, + "step": 94400 + }, + { + "epoch": 3.7323208353108686, + "grad_norm": 7.999619007110596, + "learning_rate": 2.263245226685507e-06, + "loss": 0.4164, + "step": 94500 + }, + { + "epoch": 3.7362759057111217, + "grad_norm": 7.173412322998047, + "learning_rate": 2.2498033982156786e-06, + "loss": 0.4375, + "step": 94600 + }, + { + "epoch": 3.740230976111375, + "grad_norm": 9.304818153381348, + "learning_rate": 2.236394558430945e-06, + "loss": 0.392, + "step": 94700 + }, + { + "epoch": 3.744186046511628, + "grad_norm": 11.311491966247559, + "learning_rate": 2.2230187915833827e-06, + "loss": 0.3951, + "step": 94800 + }, + { + "epoch": 3.748141116911881, + "grad_norm": 8.677116394042969, + "learning_rate": 2.2096761817172735e-06, + "loss": 0.406, + "step": 94900 + }, + { + "epoch": 3.752096187312134, + "grad_norm": 8.232828140258789, + "learning_rate": 2.196366812668544e-06, + "loss": 0.411, + "step": 95000 + }, + { + "epoch": 3.7560512577123872, + "grad_norm": 7.7508440017700195, + "learning_rate": 2.1830907680642782e-06, + "loss": 0.425, + "step": 95100 + }, + { + "epoch": 3.7600063281126404, + "grad_norm": 9.16795539855957, + "learning_rate": 2.1698481313221564e-06, + "loss": 0.4124, + "step": 95200 + }, + { + "epoch": 3.7639613985128935, + "grad_norm": 9.288763046264648, + "learning_rate": 2.156638985649958e-06, + "loss": 0.3808, + "step": 95300 + }, + { + "epoch": 3.7679164689131466, + "grad_norm": 8.190459251403809, + "learning_rate": 2.1434634140450197e-06, + "loss": 0.41, + "step": 95400 + }, + { + "epoch": 3.7718715393133997, + "grad_norm": 9.141033172607422, + "learning_rate": 2.1303214992937243e-06, + "loss": 0.3768, + "step": 95500 + }, + { + "epoch": 3.775826609713653, + "grad_norm": 10.411824226379395, + "learning_rate": 2.117213323970973e-06, + "loss": 0.3955, + "step": 95600 + }, + { + "epoch": 3.779781680113906, + "grad_norm": 8.365334510803223, + "learning_rate": 2.1041389704396817e-06, + "loss": 0.4249, + "step": 95700 + }, + { + "epoch": 3.783736750514159, + "grad_norm": 7.489829063415527, + "learning_rate": 2.0910985208502444e-06, + "loss": 0.4026, + "step": 95800 + }, + { + "epoch": 3.787691820914412, + "grad_norm": 10.931512832641602, + "learning_rate": 2.078092057140026e-06, + "loss": 0.3997, + "step": 95900 + }, + { + "epoch": 3.7916468913146653, + "grad_norm": 7.846652507781982, + "learning_rate": 2.065119661032849e-06, + "loss": 0.395, + "step": 96000 + }, + { + "epoch": 3.7956019617149184, + "grad_norm": 9.316243171691895, + "learning_rate": 2.0521814140384813e-06, + "loss": 0.41, + "step": 96100 + }, + { + "epoch": 3.7995570321151715, + "grad_norm": 7.889522552490234, + "learning_rate": 2.0392773974521163e-06, + "loss": 0.3711, + "step": 96200 + }, + { + "epoch": 3.803512102515425, + "grad_norm": 8.276979446411133, + "learning_rate": 2.026407692353868e-06, + "loss": 0.3864, + "step": 96300 + }, + { + "epoch": 3.807467172915678, + "grad_norm": 7.635178565979004, + "learning_rate": 2.0135723796082577e-06, + "loss": 0.3854, + "step": 96400 + }, + { + "epoch": 3.8114222433159313, + "grad_norm": 10.312528610229492, + "learning_rate": 2.0007715398637154e-06, + "loss": 0.4005, + "step": 96500 + }, + { + "epoch": 3.8153773137161844, + "grad_norm": 9.488860130310059, + "learning_rate": 1.9880052535520586e-06, + "loss": 0.3974, + "step": 96600 + }, + { + "epoch": 3.8193323841164375, + "grad_norm": 7.628594875335693, + "learning_rate": 1.975273600888002e-06, + "loss": 0.386, + "step": 96700 + }, + { + "epoch": 3.8232874545166906, + "grad_norm": 8.586271286010742, + "learning_rate": 1.9625766618686344e-06, + "loss": 0.4291, + "step": 96800 + }, + { + "epoch": 3.8272425249169437, + "grad_norm": 12.247819900512695, + "learning_rate": 1.9499145162729406e-06, + "loss": 0.3928, + "step": 96900 + }, + { + "epoch": 3.831197595317197, + "grad_norm": 10.7797269821167, + "learning_rate": 1.9372872436612757e-06, + "loss": 0.3932, + "step": 97000 + }, + { + "epoch": 3.83515266571745, + "grad_norm": 9.973164558410645, + "learning_rate": 1.92469492337489e-06, + "loss": 0.4081, + "step": 97100 + }, + { + "epoch": 3.839107736117703, + "grad_norm": 8.402750015258789, + "learning_rate": 1.912137634535398e-06, + "loss": 0.3671, + "step": 97200 + }, + { + "epoch": 3.843062806517956, + "grad_norm": 7.674239158630371, + "learning_rate": 1.8996154560443199e-06, + "loss": 0.426, + "step": 97300 + }, + { + "epoch": 3.8470178769182093, + "grad_norm": 12.181373596191406, + "learning_rate": 1.8871284665825514e-06, + "loss": 0.3842, + "step": 97400 + }, + { + "epoch": 3.8509729473184624, + "grad_norm": 10.594054222106934, + "learning_rate": 1.8746767446098952e-06, + "loss": 0.3929, + "step": 97500 + }, + { + "epoch": 3.8549280177187155, + "grad_norm": 9.901941299438477, + "learning_rate": 1.86226036836455e-06, + "loss": 0.3947, + "step": 97600 + }, + { + "epoch": 3.8588830881189686, + "grad_norm": 8.486770629882812, + "learning_rate": 1.8498794158626267e-06, + "loss": 0.3876, + "step": 97700 + }, + { + "epoch": 3.8628381585192217, + "grad_norm": 9.486550331115723, + "learning_rate": 1.8375339648976577e-06, + "loss": 0.3769, + "step": 97800 + }, + { + "epoch": 3.866793228919475, + "grad_norm": 8.092984199523926, + "learning_rate": 1.8252240930401123e-06, + "loss": 0.3927, + "step": 97900 + }, + { + "epoch": 3.870748299319728, + "grad_norm": 11.440380096435547, + "learning_rate": 1.8129498776368985e-06, + "loss": 0.3913, + "step": 98000 + }, + { + "epoch": 3.874703369719981, + "grad_norm": 8.888964653015137, + "learning_rate": 1.8007113958108853e-06, + "loss": 0.3608, + "step": 98100 + }, + { + "epoch": 3.878658440120234, + "grad_norm": 9.67924690246582, + "learning_rate": 1.7885087244604153e-06, + "loss": 0.3869, + "step": 98200 + }, + { + "epoch": 3.8826135105204873, + "grad_norm": 8.57454776763916, + "learning_rate": 1.7763419402588263e-06, + "loss": 0.3739, + "step": 98300 + }, + { + "epoch": 3.8865685809207404, + "grad_norm": 10.284082412719727, + "learning_rate": 1.7642111196539621e-06, + "loss": 0.3835, + "step": 98400 + }, + { + "epoch": 3.8905236513209935, + "grad_norm": 10.843579292297363, + "learning_rate": 1.7521163388676956e-06, + "loss": 0.4116, + "step": 98500 + }, + { + "epoch": 3.8944787217212467, + "grad_norm": 9.84794807434082, + "learning_rate": 1.7400576738954468e-06, + "loss": 0.4025, + "step": 98600 + }, + { + "epoch": 3.8984337921214998, + "grad_norm": 8.804971694946289, + "learning_rate": 1.728035200505718e-06, + "loss": 0.3527, + "step": 98700 + }, + { + "epoch": 3.902388862521753, + "grad_norm": 9.877714157104492, + "learning_rate": 1.716048994239598e-06, + "loss": 0.3746, + "step": 98800 + }, + { + "epoch": 3.906343932922006, + "grad_norm": 8.774710655212402, + "learning_rate": 1.7040991304103059e-06, + "loss": 0.37, + "step": 98900 + }, + { + "epoch": 3.910299003322259, + "grad_norm": 9.542379379272461, + "learning_rate": 1.692185684102699e-06, + "loss": 0.3734, + "step": 99000 + }, + { + "epoch": 3.914254073722512, + "grad_norm": 10.886748313903809, + "learning_rate": 1.6803087301728216e-06, + "loss": 0.3437, + "step": 99100 + }, + { + "epoch": 3.9182091441227653, + "grad_norm": 9.636738777160645, + "learning_rate": 1.6684683432474177e-06, + "loss": 0.3882, + "step": 99200 + }, + { + "epoch": 3.9221642145230184, + "grad_norm": 7.8506269454956055, + "learning_rate": 1.656664597723476e-06, + "loss": 0.3702, + "step": 99300 + }, + { + "epoch": 3.9261192849232716, + "grad_norm": 9.247983932495117, + "learning_rate": 1.64489756776775e-06, + "loss": 0.3562, + "step": 99400 + }, + { + "epoch": 3.9300743553235247, + "grad_norm": 9.73788070678711, + "learning_rate": 1.633167327316297e-06, + "loss": 0.3913, + "step": 99500 + }, + { + "epoch": 3.934029425723778, + "grad_norm": 9.978226661682129, + "learning_rate": 1.6214739500740142e-06, + "loss": 0.3628, + "step": 99600 + }, + { + "epoch": 3.937984496124031, + "grad_norm": 8.844416618347168, + "learning_rate": 1.6098175095141815e-06, + "loss": 0.3713, + "step": 99700 + }, + { + "epoch": 3.941939566524284, + "grad_norm": 8.160172462463379, + "learning_rate": 1.5981980788779863e-06, + "loss": 0.3669, + "step": 99800 + }, + { + "epoch": 3.945894636924537, + "grad_norm": 11.400806427001953, + "learning_rate": 1.5866157311740733e-06, + "loss": 0.3758, + "step": 99900 + }, + { + "epoch": 3.9498497073247902, + "grad_norm": 10.661243438720703, + "learning_rate": 1.5750705391780833e-06, + "loss": 0.3869, + "step": 100000 + }, + { + "epoch": 3.9538047777250434, + "grad_norm": 6.112812519073486, + "learning_rate": 1.5635625754321935e-06, + "loss": 0.3829, + "step": 100100 + }, + { + "epoch": 3.9577598481252965, + "grad_norm": 9.569903373718262, + "learning_rate": 1.55209191224467e-06, + "loss": 0.3836, + "step": 100200 + }, + { + "epoch": 3.9617149185255496, + "grad_norm": 10.391584396362305, + "learning_rate": 1.5406586216894019e-06, + "loss": 0.3854, + "step": 100300 + }, + { + "epoch": 3.9656699889258027, + "grad_norm": 9.75516414642334, + "learning_rate": 1.5292627756054542e-06, + "loss": 0.3837, + "step": 100400 + }, + { + "epoch": 3.969625059326056, + "grad_norm": 7.6981706619262695, + "learning_rate": 1.517904445596617e-06, + "loss": 0.396, + "step": 100500 + }, + { + "epoch": 3.973580129726309, + "grad_norm": 9.208215713500977, + "learning_rate": 1.506583703030958e-06, + "loss": 0.3652, + "step": 100600 + }, + { + "epoch": 3.977535200126562, + "grad_norm": 11.380250930786133, + "learning_rate": 1.495300619040367e-06, + "loss": 0.3833, + "step": 100700 + }, + { + "epoch": 3.981490270526815, + "grad_norm": 8.134552955627441, + "learning_rate": 1.4840552645201126e-06, + "loss": 0.3723, + "step": 100800 + }, + { + "epoch": 3.9854453409270683, + "grad_norm": 10.506682395935059, + "learning_rate": 1.4728477101283977e-06, + "loss": 0.3825, + "step": 100900 + }, + { + "epoch": 3.9894004113273214, + "grad_norm": 9.069189071655273, + "learning_rate": 1.4616780262859177e-06, + "loss": 0.3804, + "step": 101000 + } + ], + "logging_steps": 100, + "max_steps": 126330, + "num_input_tokens_seen": 0, + "num_train_epochs": 5, + "save_steps": 25266, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 1.1903543369015073e+20, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}