memorag-mistral-7b-inst / trainer_state.json
chienqian
init
578867d
raw
history blame contribute delete
No virus
49.5 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 15260,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"grad_norm": 0.9230163097381592,
"learning_rate": 4.984598243544371e-05,
"loss": 2.0529,
"step": 50
},
{
"epoch": 0.01,
"grad_norm": 1.205999732017517,
"learning_rate": 4.968213396251148e-05,
"loss": 1.8376,
"step": 100
},
{
"epoch": 0.01,
"grad_norm": 0.5816870331764221,
"learning_rate": 4.951828548957924e-05,
"loss": 1.7882,
"step": 150
},
{
"epoch": 0.01,
"grad_norm": 0.45057499408721924,
"learning_rate": 4.935443701664701e-05,
"loss": 1.8113,
"step": 200
},
{
"epoch": 0.02,
"grad_norm": 0.5159332752227783,
"learning_rate": 4.9190588543714776e-05,
"loss": 1.8183,
"step": 250
},
{
"epoch": 0.02,
"grad_norm": 0.42658549547195435,
"learning_rate": 4.9026740070782544e-05,
"loss": 1.8049,
"step": 300
},
{
"epoch": 0.02,
"grad_norm": 0.6193262338638306,
"learning_rate": 4.8862891597850306e-05,
"loss": 1.8195,
"step": 350
},
{
"epoch": 0.03,
"grad_norm": 0.40126094222068787,
"learning_rate": 4.8699043124918075e-05,
"loss": 1.8445,
"step": 400
},
{
"epoch": 0.03,
"grad_norm": 0.4089142978191376,
"learning_rate": 4.853519465198584e-05,
"loss": 1.7967,
"step": 450
},
{
"epoch": 0.03,
"grad_norm": 0.3923342525959015,
"learning_rate": 4.837134617905361e-05,
"loss": 1.7941,
"step": 500
},
{
"epoch": 0.04,
"grad_norm": 0.415017694234848,
"learning_rate": 4.820749770612138e-05,
"loss": 1.7999,
"step": 550
},
{
"epoch": 0.04,
"grad_norm": 0.5846891403198242,
"learning_rate": 4.804364923318915e-05,
"loss": 1.8064,
"step": 600
},
{
"epoch": 0.04,
"grad_norm": 0.3614323139190674,
"learning_rate": 4.787980076025692e-05,
"loss": 1.7865,
"step": 650
},
{
"epoch": 0.05,
"grad_norm": 0.37756502628326416,
"learning_rate": 4.7715952287324686e-05,
"loss": 1.7745,
"step": 700
},
{
"epoch": 0.05,
"grad_norm": 0.3763517141342163,
"learning_rate": 4.7552103814392455e-05,
"loss": 1.8188,
"step": 750
},
{
"epoch": 0.05,
"grad_norm": 0.41242149472236633,
"learning_rate": 4.7388255341460216e-05,
"loss": 1.8238,
"step": 800
},
{
"epoch": 0.06,
"grad_norm": 0.4072262644767761,
"learning_rate": 4.7224406868527985e-05,
"loss": 1.838,
"step": 850
},
{
"epoch": 0.06,
"grad_norm": 0.5368711352348328,
"learning_rate": 4.7060558395595753e-05,
"loss": 1.8121,
"step": 900
},
{
"epoch": 0.06,
"grad_norm": 0.3726208508014679,
"learning_rate": 4.689670992266352e-05,
"loss": 1.7872,
"step": 950
},
{
"epoch": 0.07,
"grad_norm": 0.3755347728729248,
"learning_rate": 4.673286144973129e-05,
"loss": 1.8164,
"step": 1000
},
{
"epoch": 0.07,
"grad_norm": 0.6106815934181213,
"learning_rate": 4.656901297679906e-05,
"loss": 1.8034,
"step": 1050
},
{
"epoch": 0.07,
"grad_norm": 0.4525647461414337,
"learning_rate": 4.640516450386683e-05,
"loss": 1.8174,
"step": 1100
},
{
"epoch": 0.08,
"grad_norm": 0.4508289396762848,
"learning_rate": 4.6241316030934596e-05,
"loss": 1.7958,
"step": 1150
},
{
"epoch": 0.08,
"grad_norm": 0.5615507960319519,
"learning_rate": 4.6077467558002365e-05,
"loss": 1.807,
"step": 1200
},
{
"epoch": 0.08,
"grad_norm": 0.5430244207382202,
"learning_rate": 4.5913619085070133e-05,
"loss": 1.7694,
"step": 1250
},
{
"epoch": 0.09,
"grad_norm": 0.4304679334163666,
"learning_rate": 4.57497706121379e-05,
"loss": 1.9162,
"step": 1300
},
{
"epoch": 0.09,
"grad_norm": 0.4160442650318146,
"learning_rate": 4.5585922139205664e-05,
"loss": 1.8106,
"step": 1350
},
{
"epoch": 0.09,
"grad_norm": 0.5771602988243103,
"learning_rate": 4.542207366627343e-05,
"loss": 1.7815,
"step": 1400
},
{
"epoch": 0.1,
"grad_norm": 0.6032801866531372,
"learning_rate": 4.52582251933412e-05,
"loss": 1.8171,
"step": 1450
},
{
"epoch": 0.1,
"grad_norm": 0.4265432059764862,
"learning_rate": 4.509437672040897e-05,
"loss": 1.7537,
"step": 1500
},
{
"epoch": 0.1,
"grad_norm": 0.6046428084373474,
"learning_rate": 4.493052824747673e-05,
"loss": 1.8251,
"step": 1550
},
{
"epoch": 0.1,
"grad_norm": 0.4501182436943054,
"learning_rate": 4.47666797745445e-05,
"loss": 1.8186,
"step": 1600
},
{
"epoch": 0.11,
"grad_norm": 0.5386430621147156,
"learning_rate": 4.460283130161227e-05,
"loss": 1.82,
"step": 1650
},
{
"epoch": 0.11,
"grad_norm": 0.549318790435791,
"learning_rate": 4.443898282868004e-05,
"loss": 1.9232,
"step": 1700
},
{
"epoch": 0.11,
"grad_norm": 0.432767778635025,
"learning_rate": 4.4275134355747806e-05,
"loss": 1.7913,
"step": 1750
},
{
"epoch": 0.12,
"grad_norm": 0.5354149341583252,
"learning_rate": 4.4111285882815574e-05,
"loss": 1.7933,
"step": 1800
},
{
"epoch": 0.12,
"grad_norm": 0.46177294850349426,
"learning_rate": 4.394743740988334e-05,
"loss": 1.7801,
"step": 1850
},
{
"epoch": 0.12,
"grad_norm": 4.336757183074951,
"learning_rate": 4.378358893695111e-05,
"loss": 1.774,
"step": 1900
},
{
"epoch": 0.13,
"grad_norm": 0.44567111134529114,
"learning_rate": 4.361974046401888e-05,
"loss": 1.7708,
"step": 1950
},
{
"epoch": 0.13,
"grad_norm": 0.4324832260608673,
"learning_rate": 4.345589199108664e-05,
"loss": 1.7743,
"step": 2000
},
{
"epoch": 0.13,
"grad_norm": 0.35905638337135315,
"learning_rate": 4.329204351815441e-05,
"loss": 1.8115,
"step": 2050
},
{
"epoch": 0.14,
"grad_norm": 0.6442646980285645,
"learning_rate": 4.312819504522218e-05,
"loss": 1.7691,
"step": 2100
},
{
"epoch": 0.14,
"grad_norm": 0.385105162858963,
"learning_rate": 4.296434657228995e-05,
"loss": 1.7991,
"step": 2150
},
{
"epoch": 0.14,
"grad_norm": 0.4879704713821411,
"learning_rate": 4.2800498099357716e-05,
"loss": 1.7906,
"step": 2200
},
{
"epoch": 0.15,
"grad_norm": 0.3042909801006317,
"learning_rate": 4.2636649626425485e-05,
"loss": 1.7957,
"step": 2250
},
{
"epoch": 0.15,
"grad_norm": 0.41942423582077026,
"learning_rate": 4.247280115349325e-05,
"loss": 1.791,
"step": 2300
},
{
"epoch": 0.15,
"grad_norm": 0.40257078409194946,
"learning_rate": 4.230895268056102e-05,
"loss": 1.7815,
"step": 2350
},
{
"epoch": 0.16,
"grad_norm": 0.4478552043437958,
"learning_rate": 4.214510420762879e-05,
"loss": 1.803,
"step": 2400
},
{
"epoch": 0.16,
"grad_norm": 284.23187255859375,
"learning_rate": 4.198125573469656e-05,
"loss": 1.8267,
"step": 2450
},
{
"epoch": 0.16,
"grad_norm": 0.3262351453304291,
"learning_rate": 4.181740726176433e-05,
"loss": 1.8991,
"step": 2500
},
{
"epoch": 0.17,
"grad_norm": 0.4118567109107971,
"learning_rate": 4.165355878883209e-05,
"loss": 1.8224,
"step": 2550
},
{
"epoch": 0.17,
"grad_norm": 1.5378553867340088,
"learning_rate": 4.148971031589986e-05,
"loss": 1.7887,
"step": 2600
},
{
"epoch": 0.17,
"grad_norm": 0.3631303012371063,
"learning_rate": 4.1325861842967626e-05,
"loss": 1.7825,
"step": 2650
},
{
"epoch": 0.18,
"grad_norm": 0.3822169303894043,
"learning_rate": 4.116201337003539e-05,
"loss": 1.7608,
"step": 2700
},
{
"epoch": 0.18,
"grad_norm": 0.41518792510032654,
"learning_rate": 4.099816489710316e-05,
"loss": 1.8161,
"step": 2750
},
{
"epoch": 0.18,
"grad_norm": 0.517413318157196,
"learning_rate": 4.0834316424170925e-05,
"loss": 1.8081,
"step": 2800
},
{
"epoch": 0.19,
"grad_norm": 0.3587147891521454,
"learning_rate": 4.0670467951238694e-05,
"loss": 1.8077,
"step": 2850
},
{
"epoch": 0.19,
"grad_norm": 0.39981338381767273,
"learning_rate": 4.050661947830646e-05,
"loss": 1.7991,
"step": 2900
},
{
"epoch": 0.19,
"grad_norm": 0.35814446210861206,
"learning_rate": 4.034277100537423e-05,
"loss": 1.7933,
"step": 2950
},
{
"epoch": 0.2,
"grad_norm": 0.317147433757782,
"learning_rate": 4.0178922532442e-05,
"loss": 1.8309,
"step": 3000
},
{
"epoch": 0.2,
"grad_norm": 0.4601675271987915,
"learning_rate": 4.001507405950977e-05,
"loss": 1.7635,
"step": 3050
},
{
"epoch": 0.2,
"grad_norm": 0.3678904175758362,
"learning_rate": 3.985122558657754e-05,
"loss": 1.7948,
"step": 3100
},
{
"epoch": 0.21,
"grad_norm": 0.37515679001808167,
"learning_rate": 3.9687377113645305e-05,
"loss": 1.823,
"step": 3150
},
{
"epoch": 0.21,
"grad_norm": 0.4639309346675873,
"learning_rate": 3.952352864071307e-05,
"loss": 1.7888,
"step": 3200
},
{
"epoch": 0.21,
"grad_norm": 0.45223185420036316,
"learning_rate": 3.9359680167780836e-05,
"loss": 1.7366,
"step": 3250
},
{
"epoch": 0.22,
"grad_norm": 0.40405765175819397,
"learning_rate": 3.9195831694848604e-05,
"loss": 1.7682,
"step": 3300
},
{
"epoch": 0.22,
"grad_norm": 0.45404157042503357,
"learning_rate": 3.903198322191637e-05,
"loss": 1.798,
"step": 3350
},
{
"epoch": 0.22,
"grad_norm": 0.5335371494293213,
"learning_rate": 3.886813474898414e-05,
"loss": 1.8179,
"step": 3400
},
{
"epoch": 0.23,
"grad_norm": 0.4759540557861328,
"learning_rate": 3.870428627605191e-05,
"loss": 1.7829,
"step": 3450
},
{
"epoch": 0.23,
"grad_norm": 0.5050774216651917,
"learning_rate": 3.854043780311968e-05,
"loss": 1.7953,
"step": 3500
},
{
"epoch": 0.23,
"grad_norm": 0.44674813747406006,
"learning_rate": 3.837658933018745e-05,
"loss": 1.7945,
"step": 3550
},
{
"epoch": 0.24,
"grad_norm": 0.34432777762413025,
"learning_rate": 3.8212740857255216e-05,
"loss": 1.7968,
"step": 3600
},
{
"epoch": 0.24,
"grad_norm": 0.4797002077102661,
"learning_rate": 3.8048892384322984e-05,
"loss": 1.7275,
"step": 3650
},
{
"epoch": 0.24,
"grad_norm": 0.40203359723091125,
"learning_rate": 3.788504391139075e-05,
"loss": 1.7905,
"step": 3700
},
{
"epoch": 0.25,
"grad_norm": 0.3257518708705902,
"learning_rate": 3.7721195438458515e-05,
"loss": 1.8077,
"step": 3750
},
{
"epoch": 0.25,
"grad_norm": 0.38829201459884644,
"learning_rate": 3.755734696552628e-05,
"loss": 1.7965,
"step": 3800
},
{
"epoch": 0.25,
"grad_norm": 0.3530164062976837,
"learning_rate": 3.739349849259405e-05,
"loss": 1.7925,
"step": 3850
},
{
"epoch": 0.26,
"grad_norm": 0.521428644657135,
"learning_rate": 3.7229650019661814e-05,
"loss": 1.7921,
"step": 3900
},
{
"epoch": 0.26,
"grad_norm": 0.3673698902130127,
"learning_rate": 3.706580154672958e-05,
"loss": 1.7981,
"step": 3950
},
{
"epoch": 0.26,
"grad_norm": 0.3477175235748291,
"learning_rate": 3.690195307379735e-05,
"loss": 1.7857,
"step": 4000
},
{
"epoch": 0.27,
"grad_norm": 0.3988274037837982,
"learning_rate": 3.673810460086512e-05,
"loss": 1.8014,
"step": 4050
},
{
"epoch": 0.27,
"grad_norm": 0.48254644870758057,
"learning_rate": 3.657425612793289e-05,
"loss": 1.8317,
"step": 4100
},
{
"epoch": 0.27,
"grad_norm": 0.9297109842300415,
"learning_rate": 3.6410407655000656e-05,
"loss": 1.7732,
"step": 4150
},
{
"epoch": 0.28,
"grad_norm": 0.33822837471961975,
"learning_rate": 3.6246559182068425e-05,
"loss": 1.8004,
"step": 4200
},
{
"epoch": 0.28,
"grad_norm": 0.3119085431098938,
"learning_rate": 3.6082710709136193e-05,
"loss": 1.7709,
"step": 4250
},
{
"epoch": 0.28,
"grad_norm": 0.3552871346473694,
"learning_rate": 3.591886223620396e-05,
"loss": 1.7865,
"step": 4300
},
{
"epoch": 0.29,
"grad_norm": 0.5379615426063538,
"learning_rate": 3.575501376327173e-05,
"loss": 1.8026,
"step": 4350
},
{
"epoch": 0.29,
"grad_norm": 0.3345562815666199,
"learning_rate": 3.559116529033949e-05,
"loss": 1.8146,
"step": 4400
},
{
"epoch": 0.29,
"grad_norm": 0.5010606646537781,
"learning_rate": 3.542731681740726e-05,
"loss": 1.8117,
"step": 4450
},
{
"epoch": 0.29,
"grad_norm": 0.46487560868263245,
"learning_rate": 3.526346834447503e-05,
"loss": 1.7583,
"step": 4500
},
{
"epoch": 0.3,
"grad_norm": 0.4831026792526245,
"learning_rate": 3.50996198715428e-05,
"loss": 1.768,
"step": 4550
},
{
"epoch": 0.3,
"grad_norm": 0.5314655303955078,
"learning_rate": 3.493577139861057e-05,
"loss": 1.8041,
"step": 4600
},
{
"epoch": 0.3,
"grad_norm": 0.4257725477218628,
"learning_rate": 3.4771922925678335e-05,
"loss": 1.7929,
"step": 4650
},
{
"epoch": 0.31,
"grad_norm": 0.4357529878616333,
"learning_rate": 3.4608074452746104e-05,
"loss": 1.7826,
"step": 4700
},
{
"epoch": 0.31,
"grad_norm": 0.4604736268520355,
"learning_rate": 3.444422597981387e-05,
"loss": 1.8099,
"step": 4750
},
{
"epoch": 0.31,
"grad_norm": 0.44079703092575073,
"learning_rate": 3.428037750688164e-05,
"loss": 1.7825,
"step": 4800
},
{
"epoch": 0.32,
"grad_norm": 0.4325021207332611,
"learning_rate": 3.411652903394941e-05,
"loss": 1.7546,
"step": 4850
},
{
"epoch": 0.32,
"grad_norm": 0.3814845681190491,
"learning_rate": 3.395268056101718e-05,
"loss": 1.8295,
"step": 4900
},
{
"epoch": 0.32,
"grad_norm": 0.3426695764064789,
"learning_rate": 3.378883208808494e-05,
"loss": 1.8149,
"step": 4950
},
{
"epoch": 0.33,
"grad_norm": 0.3510643541812897,
"learning_rate": 3.362498361515271e-05,
"loss": 1.8042,
"step": 5000
},
{
"epoch": 0.33,
"grad_norm": 0.4605468213558197,
"learning_rate": 3.346113514222048e-05,
"loss": 1.7828,
"step": 5050
},
{
"epoch": 0.33,
"grad_norm": 0.3923262059688568,
"learning_rate": 3.329728666928824e-05,
"loss": 1.8252,
"step": 5100
},
{
"epoch": 0.34,
"grad_norm": 0.35685479640960693,
"learning_rate": 3.313343819635601e-05,
"loss": 1.8354,
"step": 5150
},
{
"epoch": 0.34,
"grad_norm": 0.6825860738754272,
"learning_rate": 3.2969589723423776e-05,
"loss": 1.7901,
"step": 5200
},
{
"epoch": 0.34,
"grad_norm": 0.6853482723236084,
"learning_rate": 3.2805741250491545e-05,
"loss": 1.8072,
"step": 5250
},
{
"epoch": 0.35,
"grad_norm": 0.5828983783721924,
"learning_rate": 3.264189277755931e-05,
"loss": 1.7666,
"step": 5300
},
{
"epoch": 0.35,
"grad_norm": 0.4155603051185608,
"learning_rate": 3.247804430462708e-05,
"loss": 1.8165,
"step": 5350
},
{
"epoch": 0.35,
"grad_norm": 0.3048844635486603,
"learning_rate": 3.231419583169485e-05,
"loss": 1.7664,
"step": 5400
},
{
"epoch": 0.36,
"grad_norm": 0.786605715751648,
"learning_rate": 3.215034735876262e-05,
"loss": 1.8138,
"step": 5450
},
{
"epoch": 0.36,
"grad_norm": 0.9150344133377075,
"learning_rate": 3.198649888583039e-05,
"loss": 1.8106,
"step": 5500
},
{
"epoch": 0.36,
"grad_norm": 0.43113359808921814,
"learning_rate": 3.1822650412898156e-05,
"loss": 1.7895,
"step": 5550
},
{
"epoch": 0.37,
"grad_norm": 0.4392976760864258,
"learning_rate": 3.165880193996592e-05,
"loss": 1.7897,
"step": 5600
},
{
"epoch": 0.37,
"grad_norm": 0.5544856786727905,
"learning_rate": 3.1494953467033686e-05,
"loss": 1.783,
"step": 5650
},
{
"epoch": 0.37,
"grad_norm": 0.30507203936576843,
"learning_rate": 3.1331104994101455e-05,
"loss": 1.819,
"step": 5700
},
{
"epoch": 0.38,
"grad_norm": 0.39048048853874207,
"learning_rate": 3.1167256521169224e-05,
"loss": 1.8011,
"step": 5750
},
{
"epoch": 0.38,
"grad_norm": 0.3265933096408844,
"learning_rate": 3.100340804823699e-05,
"loss": 1.7867,
"step": 5800
},
{
"epoch": 0.38,
"grad_norm": 0.8230869174003601,
"learning_rate": 3.083955957530476e-05,
"loss": 1.8035,
"step": 5850
},
{
"epoch": 0.39,
"grad_norm": 0.5455942153930664,
"learning_rate": 3.067571110237253e-05,
"loss": 1.7662,
"step": 5900
},
{
"epoch": 0.39,
"grad_norm": 0.4801247715950012,
"learning_rate": 3.0511862629440298e-05,
"loss": 1.7932,
"step": 5950
},
{
"epoch": 0.39,
"grad_norm": 0.3930888772010803,
"learning_rate": 3.0348014156508063e-05,
"loss": 1.7954,
"step": 6000
},
{
"epoch": 0.4,
"grad_norm": 0.3690837323665619,
"learning_rate": 3.018416568357583e-05,
"loss": 1.8081,
"step": 6050
},
{
"epoch": 0.4,
"grad_norm": 0.45955517888069153,
"learning_rate": 3.00203172106436e-05,
"loss": 1.7859,
"step": 6100
},
{
"epoch": 0.4,
"grad_norm": 0.3670261800289154,
"learning_rate": 2.985646873771137e-05,
"loss": 1.7891,
"step": 6150
},
{
"epoch": 0.41,
"grad_norm": 0.5386211276054382,
"learning_rate": 2.9692620264779137e-05,
"loss": 1.8026,
"step": 6200
},
{
"epoch": 0.41,
"grad_norm": 0.5107685327529907,
"learning_rate": 2.95287717918469e-05,
"loss": 1.8281,
"step": 6250
},
{
"epoch": 0.41,
"grad_norm": 0.3285304009914398,
"learning_rate": 2.9364923318914668e-05,
"loss": 1.788,
"step": 6300
},
{
"epoch": 0.42,
"grad_norm": 0.699974775314331,
"learning_rate": 2.9201074845982436e-05,
"loss": 1.8138,
"step": 6350
},
{
"epoch": 0.42,
"grad_norm": 0.3917068839073181,
"learning_rate": 2.90372263730502e-05,
"loss": 1.7874,
"step": 6400
},
{
"epoch": 0.42,
"grad_norm": 0.4506165087223053,
"learning_rate": 2.887337790011797e-05,
"loss": 1.8026,
"step": 6450
},
{
"epoch": 0.43,
"grad_norm": 0.5004333257675171,
"learning_rate": 2.870952942718574e-05,
"loss": 1.7651,
"step": 6500
},
{
"epoch": 0.43,
"grad_norm": 0.4514022171497345,
"learning_rate": 2.8545680954253507e-05,
"loss": 1.7393,
"step": 6550
},
{
"epoch": 0.43,
"grad_norm": 0.4520861804485321,
"learning_rate": 2.8381832481321276e-05,
"loss": 1.7651,
"step": 6600
},
{
"epoch": 0.44,
"grad_norm": 0.5931407809257507,
"learning_rate": 2.821798400838904e-05,
"loss": 1.7623,
"step": 6650
},
{
"epoch": 0.44,
"grad_norm": 0.518058180809021,
"learning_rate": 2.805413553545681e-05,
"loss": 1.7582,
"step": 6700
},
{
"epoch": 0.44,
"grad_norm": 0.34412890672683716,
"learning_rate": 2.7890287062524578e-05,
"loss": 1.8062,
"step": 6750
},
{
"epoch": 0.45,
"grad_norm": 0.4060100018978119,
"learning_rate": 2.7726438589592347e-05,
"loss": 1.8026,
"step": 6800
},
{
"epoch": 0.45,
"grad_norm": 0.5105322003364563,
"learning_rate": 2.7562590116660115e-05,
"loss": 1.7946,
"step": 6850
},
{
"epoch": 0.45,
"grad_norm": 0.4951707422733307,
"learning_rate": 2.7398741643727884e-05,
"loss": 1.7421,
"step": 6900
},
{
"epoch": 0.46,
"grad_norm": 0.5833694338798523,
"learning_rate": 2.723489317079565e-05,
"loss": 1.774,
"step": 6950
},
{
"epoch": 0.46,
"grad_norm": 0.3745187520980835,
"learning_rate": 2.7071044697863417e-05,
"loss": 1.7682,
"step": 7000
},
{
"epoch": 0.46,
"grad_norm": 0.4256528615951538,
"learning_rate": 2.6907196224931186e-05,
"loss": 1.7371,
"step": 7050
},
{
"epoch": 0.47,
"grad_norm": 0.3425038754940033,
"learning_rate": 2.6743347751998955e-05,
"loss": 1.7922,
"step": 7100
},
{
"epoch": 0.47,
"grad_norm": 0.5452577471733093,
"learning_rate": 2.6579499279066723e-05,
"loss": 1.8093,
"step": 7150
},
{
"epoch": 0.47,
"grad_norm": 0.4865332245826721,
"learning_rate": 2.641565080613449e-05,
"loss": 1.7855,
"step": 7200
},
{
"epoch": 0.48,
"grad_norm": 0.513595461845398,
"learning_rate": 2.6251802333202257e-05,
"loss": 1.7647,
"step": 7250
},
{
"epoch": 0.48,
"grad_norm": 0.508758544921875,
"learning_rate": 2.6087953860270025e-05,
"loss": 1.8106,
"step": 7300
},
{
"epoch": 0.48,
"grad_norm": 0.8143779635429382,
"learning_rate": 2.5924105387337794e-05,
"loss": 1.7777,
"step": 7350
},
{
"epoch": 0.48,
"grad_norm": 0.44763991236686707,
"learning_rate": 2.5760256914405563e-05,
"loss": 1.7972,
"step": 7400
},
{
"epoch": 0.49,
"grad_norm": 0.6493588089942932,
"learning_rate": 2.5596408441473324e-05,
"loss": 1.8002,
"step": 7450
},
{
"epoch": 0.49,
"grad_norm": 0.4406678378582001,
"learning_rate": 2.5432559968541093e-05,
"loss": 1.7884,
"step": 7500
},
{
"epoch": 0.49,
"grad_norm": 0.500289797782898,
"learning_rate": 2.526871149560886e-05,
"loss": 1.7323,
"step": 7550
},
{
"epoch": 0.5,
"grad_norm": 0.4473750591278076,
"learning_rate": 2.5104863022676627e-05,
"loss": 1.7923,
"step": 7600
},
{
"epoch": 0.5,
"grad_norm": 0.3302167057991028,
"learning_rate": 2.49410145497444e-05,
"loss": 1.7537,
"step": 7650
},
{
"epoch": 0.5,
"grad_norm": 0.5885249376296997,
"learning_rate": 2.4777166076812167e-05,
"loss": 1.769,
"step": 7700
},
{
"epoch": 0.51,
"grad_norm": 0.3272787034511566,
"learning_rate": 2.4613317603879932e-05,
"loss": 1.7479,
"step": 7750
},
{
"epoch": 0.51,
"grad_norm": 0.38857901096343994,
"learning_rate": 2.44494691309477e-05,
"loss": 1.7615,
"step": 7800
},
{
"epoch": 0.51,
"grad_norm": 0.7423263192176819,
"learning_rate": 2.4285620658015466e-05,
"loss": 1.7882,
"step": 7850
},
{
"epoch": 0.52,
"grad_norm": 0.3933955729007721,
"learning_rate": 2.4121772185083235e-05,
"loss": 1.804,
"step": 7900
},
{
"epoch": 0.52,
"grad_norm": 0.3640703856945038,
"learning_rate": 2.3957923712151003e-05,
"loss": 1.7408,
"step": 7950
},
{
"epoch": 0.52,
"grad_norm": 0.7845476865768433,
"learning_rate": 2.3794075239218772e-05,
"loss": 1.8048,
"step": 8000
},
{
"epoch": 0.53,
"grad_norm": 0.46670156717300415,
"learning_rate": 2.363022676628654e-05,
"loss": 1.7889,
"step": 8050
},
{
"epoch": 0.53,
"grad_norm": 0.5479481220245361,
"learning_rate": 2.346637829335431e-05,
"loss": 1.7725,
"step": 8100
},
{
"epoch": 0.53,
"grad_norm": 0.4924238324165344,
"learning_rate": 2.3302529820422074e-05,
"loss": 1.7956,
"step": 8150
},
{
"epoch": 0.54,
"grad_norm": 0.5267847776412964,
"learning_rate": 2.3138681347489843e-05,
"loss": 1.7683,
"step": 8200
},
{
"epoch": 0.54,
"grad_norm": 0.42921149730682373,
"learning_rate": 2.297483287455761e-05,
"loss": 1.7662,
"step": 8250
},
{
"epoch": 0.54,
"grad_norm": 0.4606122374534607,
"learning_rate": 2.281098440162538e-05,
"loss": 1.7773,
"step": 8300
},
{
"epoch": 0.55,
"grad_norm": 0.4239887595176697,
"learning_rate": 2.2647135928693145e-05,
"loss": 1.7753,
"step": 8350
},
{
"epoch": 0.55,
"grad_norm": 0.46374326944351196,
"learning_rate": 2.2483287455760914e-05,
"loss": 1.814,
"step": 8400
},
{
"epoch": 0.55,
"grad_norm": 0.3622525632381439,
"learning_rate": 2.231943898282868e-05,
"loss": 1.7955,
"step": 8450
},
{
"epoch": 0.56,
"grad_norm": 0.7099259495735168,
"learning_rate": 2.2155590509896447e-05,
"loss": 1.8063,
"step": 8500
},
{
"epoch": 0.56,
"grad_norm": 0.4248027801513672,
"learning_rate": 2.1991742036964216e-05,
"loss": 1.7714,
"step": 8550
},
{
"epoch": 0.56,
"grad_norm": 0.3817599415779114,
"learning_rate": 2.1827893564031985e-05,
"loss": 1.8087,
"step": 8600
},
{
"epoch": 0.57,
"grad_norm": 0.4455879032611847,
"learning_rate": 2.1664045091099753e-05,
"loss": 1.7811,
"step": 8650
},
{
"epoch": 0.57,
"grad_norm": 0.4604770243167877,
"learning_rate": 2.1500196618167522e-05,
"loss": 1.7511,
"step": 8700
},
{
"epoch": 0.57,
"grad_norm": 0.43301329016685486,
"learning_rate": 2.1336348145235287e-05,
"loss": 1.762,
"step": 8750
},
{
"epoch": 0.58,
"grad_norm": 0.4077916741371155,
"learning_rate": 2.1172499672303055e-05,
"loss": 1.7512,
"step": 8800
},
{
"epoch": 0.58,
"grad_norm": 0.3406469225883484,
"learning_rate": 2.1008651199370824e-05,
"loss": 1.8008,
"step": 8850
},
{
"epoch": 0.58,
"grad_norm": 1.0516856908798218,
"learning_rate": 2.084480272643859e-05,
"loss": 1.7822,
"step": 8900
},
{
"epoch": 0.59,
"grad_norm": 0.4626096189022064,
"learning_rate": 2.0680954253506358e-05,
"loss": 1.7883,
"step": 8950
},
{
"epoch": 0.59,
"grad_norm": 0.9625229239463806,
"learning_rate": 2.0517105780574126e-05,
"loss": 1.7423,
"step": 9000
},
{
"epoch": 0.59,
"grad_norm": 0.5017744302749634,
"learning_rate": 2.035325730764189e-05,
"loss": 1.7559,
"step": 9050
},
{
"epoch": 0.6,
"grad_norm": 0.41525375843048096,
"learning_rate": 2.018940883470966e-05,
"loss": 1.7961,
"step": 9100
},
{
"epoch": 0.6,
"grad_norm": 0.48959624767303467,
"learning_rate": 2.002556036177743e-05,
"loss": 1.797,
"step": 9150
},
{
"epoch": 0.6,
"grad_norm": 0.5421375036239624,
"learning_rate": 1.9861711888845197e-05,
"loss": 1.7637,
"step": 9200
},
{
"epoch": 0.61,
"grad_norm": 0.47492170333862305,
"learning_rate": 1.9697863415912966e-05,
"loss": 1.7559,
"step": 9250
},
{
"epoch": 0.61,
"grad_norm": 0.8118923902511597,
"learning_rate": 1.9534014942980734e-05,
"loss": 1.8089,
"step": 9300
},
{
"epoch": 0.61,
"grad_norm": 0.6577257513999939,
"learning_rate": 1.93701664700485e-05,
"loss": 1.7974,
"step": 9350
},
{
"epoch": 0.62,
"grad_norm": 0.3861309587955475,
"learning_rate": 1.9206317997116268e-05,
"loss": 1.7601,
"step": 9400
},
{
"epoch": 0.62,
"grad_norm": 0.4553438723087311,
"learning_rate": 1.9042469524184037e-05,
"loss": 1.8411,
"step": 9450
},
{
"epoch": 0.62,
"grad_norm": 0.5294969081878662,
"learning_rate": 1.8878621051251802e-05,
"loss": 1.8056,
"step": 9500
},
{
"epoch": 0.63,
"grad_norm": 0.4538947641849518,
"learning_rate": 1.871477257831957e-05,
"loss": 1.7922,
"step": 9550
},
{
"epoch": 0.63,
"grad_norm": 0.40422508120536804,
"learning_rate": 1.855092410538734e-05,
"loss": 1.7954,
"step": 9600
},
{
"epoch": 0.63,
"grad_norm": 0.9616673588752747,
"learning_rate": 1.8387075632455104e-05,
"loss": 1.7896,
"step": 9650
},
{
"epoch": 0.64,
"grad_norm": 0.3476395606994629,
"learning_rate": 1.8223227159522873e-05,
"loss": 1.8015,
"step": 9700
},
{
"epoch": 0.64,
"grad_norm": 0.388485312461853,
"learning_rate": 1.805937868659064e-05,
"loss": 1.7918,
"step": 9750
},
{
"epoch": 0.64,
"grad_norm": 0.5417673587799072,
"learning_rate": 1.789553021365841e-05,
"loss": 1.7444,
"step": 9800
},
{
"epoch": 0.65,
"grad_norm": 0.3682622015476227,
"learning_rate": 1.773168174072618e-05,
"loss": 1.7803,
"step": 9850
},
{
"epoch": 0.65,
"grad_norm": 0.6654282212257385,
"learning_rate": 1.7567833267793947e-05,
"loss": 1.7559,
"step": 9900
},
{
"epoch": 0.65,
"grad_norm": 0.5030877590179443,
"learning_rate": 1.7403984794861712e-05,
"loss": 1.7657,
"step": 9950
},
{
"epoch": 0.66,
"grad_norm": 0.627768874168396,
"learning_rate": 1.724013632192948e-05,
"loss": 1.7261,
"step": 10000
},
{
"epoch": 0.66,
"grad_norm": 0.4117398262023926,
"learning_rate": 1.707628784899725e-05,
"loss": 1.7837,
"step": 10050
},
{
"epoch": 0.66,
"grad_norm": 0.40106311440467834,
"learning_rate": 1.6912439376065015e-05,
"loss": 1.7961,
"step": 10100
},
{
"epoch": 0.67,
"grad_norm": 0.32165274024009705,
"learning_rate": 1.6748590903132783e-05,
"loss": 1.7629,
"step": 10150
},
{
"epoch": 0.67,
"grad_norm": 0.47640106081962585,
"learning_rate": 1.6584742430200552e-05,
"loss": 1.777,
"step": 10200
},
{
"epoch": 0.67,
"grad_norm": 0.45250073075294495,
"learning_rate": 1.6420893957268317e-05,
"loss": 1.771,
"step": 10250
},
{
"epoch": 0.67,
"grad_norm": 0.45913493633270264,
"learning_rate": 1.6257045484336086e-05,
"loss": 1.7971,
"step": 10300
},
{
"epoch": 0.68,
"grad_norm": 0.5237872004508972,
"learning_rate": 1.6093197011403854e-05,
"loss": 1.7989,
"step": 10350
},
{
"epoch": 0.68,
"grad_norm": 0.47246378660202026,
"learning_rate": 1.5929348538471623e-05,
"loss": 1.7867,
"step": 10400
},
{
"epoch": 0.68,
"grad_norm": 0.3215338885784149,
"learning_rate": 1.576550006553939e-05,
"loss": 1.7864,
"step": 10450
},
{
"epoch": 0.69,
"grad_norm": 1.0621765851974487,
"learning_rate": 1.560165159260716e-05,
"loss": 1.7482,
"step": 10500
},
{
"epoch": 0.69,
"grad_norm": 0.5376123785972595,
"learning_rate": 1.5437803119674925e-05,
"loss": 1.7635,
"step": 10550
},
{
"epoch": 0.69,
"grad_norm": 0.8421134352684021,
"learning_rate": 1.5273954646742694e-05,
"loss": 1.7854,
"step": 10600
},
{
"epoch": 0.7,
"grad_norm": 0.4911418855190277,
"learning_rate": 1.5110106173810462e-05,
"loss": 1.7514,
"step": 10650
},
{
"epoch": 0.7,
"grad_norm": 0.32285892963409424,
"learning_rate": 1.4946257700878227e-05,
"loss": 1.7701,
"step": 10700
},
{
"epoch": 0.7,
"grad_norm": 0.6105034947395325,
"learning_rate": 1.4782409227945996e-05,
"loss": 1.7726,
"step": 10750
},
{
"epoch": 0.71,
"grad_norm": 0.36749064922332764,
"learning_rate": 1.4618560755013763e-05,
"loss": 1.8059,
"step": 10800
},
{
"epoch": 0.71,
"grad_norm": 0.46773606538772583,
"learning_rate": 1.4454712282081531e-05,
"loss": 1.8167,
"step": 10850
},
{
"epoch": 0.71,
"grad_norm": 0.5864791870117188,
"learning_rate": 1.42908638091493e-05,
"loss": 1.7503,
"step": 10900
},
{
"epoch": 0.72,
"grad_norm": 0.5605257153511047,
"learning_rate": 1.4127015336217067e-05,
"loss": 1.8021,
"step": 10950
},
{
"epoch": 0.72,
"grad_norm": 0.6443638205528259,
"learning_rate": 1.3963166863284835e-05,
"loss": 1.7622,
"step": 11000
},
{
"epoch": 0.72,
"grad_norm": 0.323830246925354,
"learning_rate": 1.3799318390352602e-05,
"loss": 1.7814,
"step": 11050
},
{
"epoch": 0.73,
"grad_norm": 0.5781140327453613,
"learning_rate": 1.363546991742037e-05,
"loss": 1.7693,
"step": 11100
},
{
"epoch": 0.73,
"grad_norm": 0.7093366980552673,
"learning_rate": 1.347162144448814e-05,
"loss": 1.7426,
"step": 11150
},
{
"epoch": 0.73,
"grad_norm": 0.4438339173793793,
"learning_rate": 1.3307772971555906e-05,
"loss": 1.7855,
"step": 11200
},
{
"epoch": 0.74,
"grad_norm": 0.4285770356655121,
"learning_rate": 1.3143924498623675e-05,
"loss": 1.7898,
"step": 11250
},
{
"epoch": 0.74,
"grad_norm": 0.6678707599639893,
"learning_rate": 1.298007602569144e-05,
"loss": 1.7668,
"step": 11300
},
{
"epoch": 0.74,
"grad_norm": 0.5140306353569031,
"learning_rate": 1.2816227552759209e-05,
"loss": 1.7804,
"step": 11350
},
{
"epoch": 0.75,
"grad_norm": 0.3694643974304199,
"learning_rate": 1.2652379079826975e-05,
"loss": 1.7633,
"step": 11400
},
{
"epoch": 0.75,
"grad_norm": 0.3662973940372467,
"learning_rate": 1.2488530606894744e-05,
"loss": 1.758,
"step": 11450
},
{
"epoch": 0.75,
"grad_norm": 0.41093963384628296,
"learning_rate": 1.2324682133962513e-05,
"loss": 1.7501,
"step": 11500
},
{
"epoch": 0.76,
"grad_norm": 0.4775155782699585,
"learning_rate": 1.216083366103028e-05,
"loss": 1.7986,
"step": 11550
},
{
"epoch": 0.76,
"grad_norm": 0.36493539810180664,
"learning_rate": 1.1996985188098048e-05,
"loss": 1.7555,
"step": 11600
},
{
"epoch": 0.76,
"grad_norm": 0.9100409746170044,
"learning_rate": 1.1833136715165815e-05,
"loss": 1.7505,
"step": 11650
},
{
"epoch": 0.77,
"grad_norm": 0.41831496357917786,
"learning_rate": 1.1669288242233583e-05,
"loss": 1.7612,
"step": 11700
},
{
"epoch": 0.77,
"grad_norm": 0.4716850221157074,
"learning_rate": 1.150543976930135e-05,
"loss": 1.7838,
"step": 11750
},
{
"epoch": 0.77,
"grad_norm": 0.3609485924243927,
"learning_rate": 1.1341591296369119e-05,
"loss": 1.7926,
"step": 11800
},
{
"epoch": 0.78,
"grad_norm": 0.3157222270965576,
"learning_rate": 1.1177742823436886e-05,
"loss": 1.7448,
"step": 11850
},
{
"epoch": 0.78,
"grad_norm": 0.46687546372413635,
"learning_rate": 1.1013894350504654e-05,
"loss": 1.7975,
"step": 11900
},
{
"epoch": 0.78,
"grad_norm": 0.6575301289558411,
"learning_rate": 1.0850045877572421e-05,
"loss": 1.736,
"step": 11950
},
{
"epoch": 0.79,
"grad_norm": 1.4544192552566528,
"learning_rate": 1.0686197404640188e-05,
"loss": 1.7763,
"step": 12000
},
{
"epoch": 0.79,
"grad_norm": 0.5222745537757874,
"learning_rate": 1.0522348931707957e-05,
"loss": 1.7786,
"step": 12050
},
{
"epoch": 0.79,
"grad_norm": 0.6186469197273254,
"learning_rate": 1.0358500458775725e-05,
"loss": 1.7963,
"step": 12100
},
{
"epoch": 0.8,
"grad_norm": 1.1090551614761353,
"learning_rate": 1.0194651985843492e-05,
"loss": 1.808,
"step": 12150
},
{
"epoch": 0.8,
"grad_norm": 0.46629172563552856,
"learning_rate": 1.003080351291126e-05,
"loss": 1.7754,
"step": 12200
},
{
"epoch": 0.8,
"grad_norm": 1.5671145915985107,
"learning_rate": 9.866955039979028e-06,
"loss": 1.7846,
"step": 12250
},
{
"epoch": 0.81,
"grad_norm": 0.5020501613616943,
"learning_rate": 9.703106567046794e-06,
"loss": 1.753,
"step": 12300
},
{
"epoch": 0.81,
"grad_norm": 0.34436970949172974,
"learning_rate": 9.539258094114563e-06,
"loss": 1.8023,
"step": 12350
},
{
"epoch": 0.81,
"grad_norm": 0.4563136100769043,
"learning_rate": 9.375409621182332e-06,
"loss": 1.7644,
"step": 12400
},
{
"epoch": 0.82,
"grad_norm": 0.3465505540370941,
"learning_rate": 9.211561148250098e-06,
"loss": 1.7819,
"step": 12450
},
{
"epoch": 0.82,
"grad_norm": 0.4823606610298157,
"learning_rate": 9.047712675317867e-06,
"loss": 1.7626,
"step": 12500
},
{
"epoch": 0.82,
"grad_norm": 0.5832303166389465,
"learning_rate": 8.883864202385634e-06,
"loss": 1.724,
"step": 12550
},
{
"epoch": 0.83,
"grad_norm": 0.590023934841156,
"learning_rate": 8.7200157294534e-06,
"loss": 1.7708,
"step": 12600
},
{
"epoch": 0.83,
"grad_norm": 0.3506929278373718,
"learning_rate": 8.55616725652117e-06,
"loss": 1.7979,
"step": 12650
},
{
"epoch": 0.83,
"grad_norm": 0.5510967373847961,
"learning_rate": 8.392318783588938e-06,
"loss": 1.7531,
"step": 12700
},
{
"epoch": 0.84,
"grad_norm": 0.4131988286972046,
"learning_rate": 8.228470310656705e-06,
"loss": 1.7869,
"step": 12750
},
{
"epoch": 0.84,
"grad_norm": 0.3922310173511505,
"learning_rate": 8.064621837724473e-06,
"loss": 1.8283,
"step": 12800
},
{
"epoch": 0.84,
"grad_norm": 0.4114150106906891,
"learning_rate": 7.90077336479224e-06,
"loss": 1.8061,
"step": 12850
},
{
"epoch": 0.85,
"grad_norm": 0.5551263093948364,
"learning_rate": 7.736924891860007e-06,
"loss": 1.8083,
"step": 12900
},
{
"epoch": 0.85,
"grad_norm": 0.4596497714519501,
"learning_rate": 7.573076418927776e-06,
"loss": 1.768,
"step": 12950
},
{
"epoch": 0.85,
"grad_norm": 0.5840251445770264,
"learning_rate": 7.409227945995543e-06,
"loss": 1.7793,
"step": 13000
},
{
"epoch": 0.86,
"grad_norm": 0.6608572602272034,
"learning_rate": 7.245379473063311e-06,
"loss": 1.7433,
"step": 13050
},
{
"epoch": 0.86,
"grad_norm": 0.4415280818939209,
"learning_rate": 7.08153100013108e-06,
"loss": 1.7555,
"step": 13100
},
{
"epoch": 0.86,
"grad_norm": 0.8738279938697815,
"learning_rate": 6.9176825271988474e-06,
"loss": 1.7655,
"step": 13150
},
{
"epoch": 0.87,
"grad_norm": 0.5214579105377197,
"learning_rate": 6.753834054266614e-06,
"loss": 1.7626,
"step": 13200
},
{
"epoch": 0.87,
"grad_norm": 0.42428484559059143,
"learning_rate": 6.589985581334382e-06,
"loss": 1.754,
"step": 13250
},
{
"epoch": 0.87,
"grad_norm": 0.4472542703151703,
"learning_rate": 6.42613710840215e-06,
"loss": 1.8004,
"step": 13300
},
{
"epoch": 0.87,
"grad_norm": 0.44398799538612366,
"learning_rate": 6.2622886354699175e-06,
"loss": 1.7692,
"step": 13350
},
{
"epoch": 0.88,
"grad_norm": 0.39280277490615845,
"learning_rate": 6.098440162537685e-06,
"loss": 1.756,
"step": 13400
},
{
"epoch": 0.88,
"grad_norm": 0.5377512574195862,
"learning_rate": 5.934591689605453e-06,
"loss": 1.8203,
"step": 13450
},
{
"epoch": 0.88,
"grad_norm": 0.384995698928833,
"learning_rate": 5.770743216673221e-06,
"loss": 1.7671,
"step": 13500
},
{
"epoch": 0.89,
"grad_norm": 0.5518258810043335,
"learning_rate": 5.606894743740988e-06,
"loss": 1.7892,
"step": 13550
},
{
"epoch": 0.89,
"grad_norm": 0.5481269955635071,
"learning_rate": 5.443046270808756e-06,
"loss": 1.7621,
"step": 13600
},
{
"epoch": 0.89,
"grad_norm": 0.6983498930931091,
"learning_rate": 5.279197797876524e-06,
"loss": 1.7699,
"step": 13650
},
{
"epoch": 0.9,
"grad_norm": 0.42530539631843567,
"learning_rate": 5.1153493249442916e-06,
"loss": 1.781,
"step": 13700
},
{
"epoch": 0.9,
"grad_norm": 0.5558788776397705,
"learning_rate": 4.951500852012059e-06,
"loss": 1.7386,
"step": 13750
},
{
"epoch": 0.9,
"grad_norm": 0.40052852034568787,
"learning_rate": 4.787652379079827e-06,
"loss": 1.7761,
"step": 13800
},
{
"epoch": 0.91,
"grad_norm": 0.3250432312488556,
"learning_rate": 4.623803906147595e-06,
"loss": 1.7782,
"step": 13850
},
{
"epoch": 0.91,
"grad_norm": 0.38817328214645386,
"learning_rate": 4.4599554332153624e-06,
"loss": 1.7407,
"step": 13900
},
{
"epoch": 0.91,
"grad_norm": 0.49112117290496826,
"learning_rate": 4.29610696028313e-06,
"loss": 1.7229,
"step": 13950
},
{
"epoch": 0.92,
"grad_norm": 2.6235454082489014,
"learning_rate": 4.132258487350898e-06,
"loss": 1.7547,
"step": 14000
},
{
"epoch": 0.92,
"grad_norm": 0.4624078571796417,
"learning_rate": 3.968410014418666e-06,
"loss": 1.8014,
"step": 14050
},
{
"epoch": 0.92,
"grad_norm": 0.3169557452201843,
"learning_rate": 3.8045615414864338e-06,
"loss": 1.7387,
"step": 14100
},
{
"epoch": 0.93,
"grad_norm": 0.6099672913551331,
"learning_rate": 3.640713068554201e-06,
"loss": 1.8081,
"step": 14150
},
{
"epoch": 0.93,
"grad_norm": 0.4258570969104767,
"learning_rate": 3.4768645956219688e-06,
"loss": 1.7663,
"step": 14200
},
{
"epoch": 0.93,
"grad_norm": 2.3566250801086426,
"learning_rate": 3.313016122689737e-06,
"loss": 1.7435,
"step": 14250
},
{
"epoch": 0.94,
"grad_norm": 0.4608743190765381,
"learning_rate": 3.1491676497575042e-06,
"loss": 1.7605,
"step": 14300
},
{
"epoch": 0.94,
"grad_norm": 0.38116034865379333,
"learning_rate": 2.985319176825272e-06,
"loss": 1.7673,
"step": 14350
},
{
"epoch": 0.94,
"grad_norm": 0.5122969746589661,
"learning_rate": 2.8214707038930397e-06,
"loss": 1.767,
"step": 14400
},
{
"epoch": 0.95,
"grad_norm": 0.5967180132865906,
"learning_rate": 2.657622230960808e-06,
"loss": 1.8025,
"step": 14450
},
{
"epoch": 0.95,
"grad_norm": 0.4944264590740204,
"learning_rate": 2.493773758028575e-06,
"loss": 1.7947,
"step": 14500
},
{
"epoch": 0.95,
"grad_norm": 0.353904128074646,
"learning_rate": 2.329925285096343e-06,
"loss": 1.795,
"step": 14550
},
{
"epoch": 0.96,
"grad_norm": 0.531599223613739,
"learning_rate": 2.166076812164111e-06,
"loss": 1.7888,
"step": 14600
},
{
"epoch": 0.96,
"grad_norm": 0.43692663311958313,
"learning_rate": 2.0022283392318783e-06,
"loss": 1.7695,
"step": 14650
},
{
"epoch": 0.96,
"grad_norm": 0.489036500453949,
"learning_rate": 1.838379866299646e-06,
"loss": 1.7895,
"step": 14700
},
{
"epoch": 0.97,
"grad_norm": 0.41665858030319214,
"learning_rate": 1.674531393367414e-06,
"loss": 1.7757,
"step": 14750
},
{
"epoch": 0.97,
"grad_norm": 0.507135272026062,
"learning_rate": 1.5106829204351817e-06,
"loss": 1.7717,
"step": 14800
},
{
"epoch": 0.97,
"grad_norm": 0.4490343928337097,
"learning_rate": 1.3468344475029494e-06,
"loss": 1.7953,
"step": 14850
},
{
"epoch": 0.98,
"grad_norm": 0.7222400307655334,
"learning_rate": 1.182985974570717e-06,
"loss": 1.7899,
"step": 14900
},
{
"epoch": 0.98,
"grad_norm": 0.46795913577079773,
"learning_rate": 1.0191375016384849e-06,
"loss": 1.7826,
"step": 14950
},
{
"epoch": 0.98,
"grad_norm": 1.0008032321929932,
"learning_rate": 8.552890287062526e-07,
"loss": 1.7784,
"step": 15000
},
{
"epoch": 0.99,
"grad_norm": 0.5528744459152222,
"learning_rate": 6.914405557740202e-07,
"loss": 1.7827,
"step": 15050
},
{
"epoch": 0.99,
"grad_norm": 0.38615551590919495,
"learning_rate": 5.275920828417879e-07,
"loss": 1.784,
"step": 15100
},
{
"epoch": 0.99,
"grad_norm": 0.44470590353012085,
"learning_rate": 3.6374360990955564e-07,
"loss": 1.8012,
"step": 15150
},
{
"epoch": 1.0,
"grad_norm": 0.39151865243911743,
"learning_rate": 1.998951369773234e-07,
"loss": 1.7987,
"step": 15200
},
{
"epoch": 1.0,
"grad_norm": 0.4639741778373718,
"learning_rate": 3.60466640450911e-08,
"loss": 1.7547,
"step": 15250
}
],
"logging_steps": 50,
"max_steps": 15260,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 3815,
"total_flos": 1.0144210157797009e+20,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}