Upload ./trainer_state.json with huggingface_hub

trainer_state.json

@@ -1,2734 +1,242 @@
 {
-  "best_metric": 1.2184005975723267,
-  "best_model_checkpoint": "./output/training_results/C021_random_sample_llama3-8b-base_pretrain_20240505_135320/checkpoint-800",
+  "best_metric": 1.191159963607788,
+  "best_model_checkpoint": "./output/training_results/C021_random_sample_Meta-Llama-3-8B_pretrain_20240726_033210/checkpoint-33984",
   "epoch": 4.0,
-  "eval_steps": 200,
-  "global_step": 1876,
+  "eval_steps": 3776,
+  "global_step": 37752,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.0021321961620469083,
+      "epoch": 0.00010595465140919687,
       "grad_norm": 0.0,
       "learning_rate": 0.0,
-      "loss": 1.3827,
+      "loss": 1.6347,
       "step": 1
     },
     {
-      "epoch": 0.010660980810234541,
-      "grad_norm": 2.0635243899751066,
-      "learning_rate": 2.25e-06,
-      "loss": 1.4616,
-      "step": 5
+      "epoch": 0.20004238186056367,
+      "grad_norm": 2.0841434899950975,
+      "learning_rate": 1.9947033898305087e-06,
+      "loss": 1.411,
+      "step": 1888
     },
     {
-      "epoch": 0.021321961620469083,
-      "grad_norm": 1.903018233428319,
-      "learning_rate": 6e-06,
-      "loss": 1.2299,
-      "step": 10
+      "epoch": 0.40008476372112733,
+      "grad_norm": 1.7712195190814046,
+      "learning_rate": 2.23788694222373e-06,
+      "loss": 1.3112,
+      "step": 3776
     },
     {
-      "epoch": 0.031982942430703626,
-      "grad_norm": 1.5467734081397355,
-      "learning_rate": 9.75e-06,
-      "loss": 1.2843,
-      "step": 15
+      "epoch": 0.40008476372112733,
+      "eval_loss": 1.2742130756378174,
+      "eval_runtime": 330.8268,
+      "eval_samples_per_second": 202.868,
+      "eval_steps_per_second": 1.587,
+      "step": 3776
     },
     {
-      "epoch": 0.042643923240938165,
-      "grad_norm": 1.995653146329813,
-      "learning_rate": 1.3500000000000001e-05,
-      "loss": 1.31,
-      "step": 20
-    },
-    {
-      "epoch": 0.053304904051172705,
-      "grad_norm": 2.410826051077061,
-      "learning_rate": 1.473632431655486e-05,
-      "loss": 1.3681,
-      "step": 25
-    },
-    {
-      "epoch": 0.06396588486140725,
-      "grad_norm": 1.765947093190964,
-      "learning_rate": 1.4306244585426726e-05,
-      "loss": 1.3228,
-      "step": 30
-    },
-    {
-      "epoch": 0.07462686567164178,
-      "grad_norm": 1.8242122051647833,
-      "learning_rate": 1.3887645147313336e-05,
-      "loss": 1.2664,
-      "step": 35
-    },
-    {
-      "epoch": 0.08528784648187633,
-      "grad_norm": 1.8945258342407882,
-      "learning_rate": 1.3560845370175857e-05,
-      "loss": 1.3073,
-      "step": 40
-    },
-    {
-      "epoch": 0.09594882729211088,
-      "grad_norm": 1.4272941224837,
-      "learning_rate": 1.3162217563557544e-05,
-      "loss": 1.3019,
-      "step": 45
-    },
-    {
-      "epoch": 0.10660980810234541,
-      "grad_norm": 1.4171437963555575,
-      "learning_rate": 1.2774310614975657e-05,
-      "loss": 1.3613,
-      "step": 50
-    },
-    {
-      "epoch": 0.11727078891257996,
-      "grad_norm": 1.4495644289679146,
-      "learning_rate": 1.2396864324754692e-05,
-      "loss": 1.4084,
-      "step": 55
-    },
-    {
-      "epoch": 0.1279317697228145,
-      "grad_norm": 1.34494402339979,
-      "learning_rate": 1.2029624121909564e-05,
-      "loss": 1.3002,
-      "step": 60
-    },
-    {
-      "epoch": 0.13859275053304904,
-      "grad_norm": 1.438232163080863,
-      "learning_rate": 1.1672340957315019e-05,
-      "loss": 1.3643,
-      "step": 65
-    },
-    {
-      "epoch": 0.14925373134328357,
-      "grad_norm": 1.5674663877824588,
-      "learning_rate": 1.1324771198617955e-05,
-      "loss": 1.2599,
-      "step": 70
-    },
-    {
-      "epoch": 0.15991471215351813,
-      "grad_norm": 1.3266580498364144,
-      "learning_rate": 1.0986676526868454e-05,
-      "loss": 1.307,
-      "step": 75
-    },
-    {
-      "epoch": 0.17057569296375266,
-      "grad_norm": 3.7513449392256173,
-      "learning_rate": 1.0657823834846472e-05,
-      "loss": 1.3701,
-      "step": 80
-    },
-    {
-      "epoch": 0.1812366737739872,
-      "grad_norm": 5.184018665584405,
-      "learning_rate": 1.0337985127060489e-05,
-      "loss": 1.37,
-      "step": 85
-    },
-    {
-      "epoch": 0.19189765458422176,
-      "grad_norm": 1.848934855985837,
-      "learning_rate": 1.0026937421395617e-05,
-      "loss": 1.2708,
-      "step": 90
-    },
-    {
-      "epoch": 0.2025586353944563,
-      "grad_norm": 1.579847110813568,
-      "learning_rate": 9.724462652387962e-06,
-      "loss": 1.3402,
-      "step": 95
-    },
-    {
-      "epoch": 0.21321961620469082,
-      "grad_norm": 1.5615282561022026,
-      "learning_rate": 9.430347576103212e-06,
-      "loss": 1.2961,
-      "step": 100
-    },
-    {
-      "epoch": 0.22388059701492538,
-      "grad_norm": 1.3659017705844316,
-      "learning_rate": 9.1443836765968e-06,
-      "loss": 1.2203,
-      "step": 105
-    },
-    {
-      "epoch": 0.2345415778251599,
-      "grad_norm": 1.4562735931359865,
-      "learning_rate": 8.921344243503367e-06,
-      "loss": 1.2945,
-      "step": 110
-    },
-    {
-      "epoch": 0.24520255863539445,
-      "grad_norm": 1.2855121058923065,
-      "learning_rate": 8.64954174238431e-06,
-      "loss": 1.295,
-      "step": 115
-    },
-    {
-      "epoch": 0.255863539445629,
-      "grad_norm": 1.3082074964800576,
-      "learning_rate": 8.385330949249671e-06,
-      "loss": 1.3053,
-      "step": 120
-    },
-    {
-      "epoch": 0.26652452025586354,
-      "grad_norm": 1.2815031342227978,
-      "learning_rate": 8.128520488668967e-06,
-      "loss": 1.2994,
-      "step": 125
-    },
-    {
-      "epoch": 0.2771855010660981,
-      "grad_norm": 1.3107524150250855,
-      "learning_rate": 7.878923285543024e-06,
-      "loss": 1.3281,
-      "step": 130
-    },
-    {
-      "epoch": 0.2878464818763326,
-      "grad_norm": 1.1907535460514518,
-      "learning_rate": 7.636356480312689e-06,
-      "loss": 1.2579,
-      "step": 135
-    },
-    {
-      "epoch": 0.29850746268656714,
-      "grad_norm": 1.2284678609494983,
-      "learning_rate": 7.400641345604814e-06,
-      "loss": 1.1634,
-      "step": 140
-    },
-    {
-      "epoch": 0.3091684434968017,
-      "grad_norm": 1.3826281905120297,
-      "learning_rate": 7.171603204294806e-06,
-      "loss": 1.3058,
-      "step": 145
-    },
-    {
-      "epoch": 0.31982942430703626,
-      "grad_norm": 2.431901009888026,
-      "learning_rate": 6.949071348965877e-06,
-      "loss": 1.3096,
-      "step": 150
-    },
-    {
-      "epoch": 0.3304904051172708,
-      "grad_norm": 1.4504346901041576,
-      "learning_rate": 6.732878962744999e-06,
-      "loss": 1.2943,
-      "step": 155
-    },
-    {
-      "epoch": 0.3411513859275053,
-      "grad_norm": 1.2723296431138542,
-      "learning_rate": 6.5228630414958555e-06,
-      "loss": 1.351,
-      "step": 160
-    },
-    {
-      "epoch": 0.35181236673773986,
-      "grad_norm": 1.2139070950942783,
-      "learning_rate": 6.318864317349401e-06,
-      "loss": 1.2937,
-      "step": 165
-    },
-    {
-      "epoch": 0.3624733475479744,
-      "grad_norm": 5.348465582649999,
-      "learning_rate": 6.120727183552839e-06,
-      "loss": 1.2726,
-      "step": 170
-    },
-    {
-      "epoch": 0.373134328358209,
-      "grad_norm": 1.6504356499880541,
-      "learning_rate": 5.928299620617992e-06,
-      "loss": 1.3039,
-      "step": 175
-    },
-    {
-      "epoch": 0.3837953091684435,
-      "grad_norm": 1.1981923340348997,
-      "learning_rate": 5.7414331237502024e-06,
-      "loss": 1.2027,
-      "step": 180
-    },
-    {
-      "epoch": 0.39445628997867804,
-      "grad_norm": 1.229607999022085,
-      "learning_rate": 5.559982631539405e-06,
-      "loss": 1.2637,
-      "step": 185
-    },
-    {
-      "epoch": 0.4051172707889126,
-      "grad_norm": 4.9301877176481606,
-      "learning_rate": 5.383806455894783e-06,
-      "loss": 1.2601,
-      "step": 190
-    },
-    {
-      "epoch": 0.4157782515991471,
-      "grad_norm": 1.2303183328754892,
-      "learning_rate": 5.212766213205117e-06,
-      "loss": 1.1995,
-      "step": 195
-    },
-    {
-      "epoch": 0.42643923240938164,
-      "grad_norm": 1.1512051049767256,
-      "learning_rate": 5.046726756706699e-06,
-      "loss": 1.2572,
-      "step": 200
-    },
-    {
-      "epoch": 0.42643923240938164,
-      "eval_loss": 1.261182188987732,
-      "eval_runtime": 87.341,
-      "eval_samples_per_second": 19.086,
-      "eval_steps_per_second": 0.309,
-      "step": 200
-    },
-    {
-      "epoch": 0.43710021321961623,
-      "grad_norm": 1.2483299149897367,
-      "learning_rate": 4.885556110041288e-06,
-      "loss": 1.2665,
-      "step": 205
-    },
-    {
-      "epoch": 0.44776119402985076,
-      "grad_norm": 1.1979266901021162,
-      "learning_rate": 4.7291254019864345e-06,
-      "loss": 1.1936,
-      "step": 210
-    },
-    {
-      "epoch": 0.4584221748400853,
-      "grad_norm": 1.1685251787317164,
-      "learning_rate": 4.577308802341063e-06,
-      "loss": 1.3458,
-      "step": 215
-    },
-    {
-      "epoch": 0.4690831556503198,
-      "grad_norm": 1.156653489171017,
-      "learning_rate": 4.42998345894903e-06,
-      "loss": 1.3039,
-      "step": 220
-    },
-    {
-      "epoch": 0.47974413646055436,
-      "grad_norm": 1.2415273415640915,
-      "learning_rate": 4.287029435843979e-06,
-      "loss": 1.2127,
-      "step": 225
-    },
-    {
-      "epoch": 0.4904051172707889,
-      "grad_norm": 1.0890064939371906,
-      "learning_rate": 4.148329652498597e-06,
-      "loss": 1.2844,
-      "step": 230
-    },
-    {
-      "epoch": 0.5010660980810234,
-      "grad_norm": 1.2341998642747427,
-      "learning_rate": 4.013769824161997e-06,
-      "loss": 1.2989,
-      "step": 235
-    },
-    {
-      "epoch": 0.511727078891258,
-      "grad_norm": 1.0800044414980956,
-      "learning_rate": 3.883238403268737e-06,
-      "loss": 1.18,
-      "step": 240
-    },
-    {
-      "epoch": 0.5223880597014925,
-      "grad_norm": 1.1911502497460917,
-      "learning_rate": 3.7566265219035852e-06,
-      "loss": 1.2198,
-      "step": 245
-    },
-    {
-      "epoch": 0.5330490405117271,
-      "grad_norm": 1.1875916676432052,
-      "learning_rate": 3.633827935305925e-06,
-      "loss": 1.2321,
-      "step": 250
-    },
-    {
-      "epoch": 0.5437100213219617,
-      "grad_norm": 1.1285856989543368,
-      "learning_rate": 3.5147389663983076e-06,
-      "loss": 1.2063,
-      "step": 255
-    },
-    {
-      "epoch": 0.5543710021321961,
-      "grad_norm": 1.2134517080609044,
-      "learning_rate": 3.3992584513234327e-06,
-      "loss": 1.2034,
-      "step": 260
-    },
-    {
-      "epoch": 0.5650319829424307,
-      "grad_norm": 1.1497034722164001,
-      "learning_rate": 3.2872876859744165e-06,
-      "loss": 1.2478,
-      "step": 265
-    },
-    {
-      "epoch": 0.5756929637526652,
-      "grad_norm": 1.1573953347115111,
-      "learning_rate": 3.1787303735030416e-06,
-      "loss": 1.217,
-      "step": 270
-    },
-    {
-      "epoch": 0.5863539445628998,
-      "grad_norm": 1.1278493709315967,
-      "learning_rate": 3.0734925727911652e-06,
-      "loss": 1.2251,
-      "step": 275
-    },
-    {
-      "epoch": 0.5970149253731343,
-      "grad_norm": 1.2020449597183245,
-      "learning_rate": 2.971482647870452e-06,
-      "loss": 1.1144,
-      "step": 280
-    },
-    {
-      "epoch": 0.6076759061833689,
-      "grad_norm": 1.2203652918926715,
-      "learning_rate": 2.8726112182758347e-06,
-      "loss": 1.281,
-      "step": 285
-    },
-    {
-      "epoch": 0.6183368869936035,
-      "grad_norm": 1.1541053058741007,
-      "learning_rate": 2.7767911103183137e-06,
-      "loss": 1.2344,
-      "step": 290
-    },
-    {
-      "epoch": 0.6289978678038379,
-      "grad_norm": 1.170855117533204,
-      "learning_rate": 2.6839373092628783e-06,
-      "loss": 1.1737,
-      "step": 295
-    },
-    {
-      "epoch": 0.6396588486140725,
-      "grad_norm": 1.1313568462008778,
-      "learning_rate": 2.593966912397475e-06,
-      "loss": 1.2395,
-      "step": 300
-    },
-    {
-      "epoch": 0.650319829424307,
-      "grad_norm": 1.0884221660719775,
-      "learning_rate": 2.5067990829791376e-06,
-      "loss": 1.2274,
-      "step": 305
-    },
-    {
-      "epoch": 0.6609808102345416,
-      "grad_norm": 1.1287129229662805,
-      "learning_rate": 2.4223550050435816e-06,
-      "loss": 1.2406,
-      "step": 310
-    },
-    {
-      "epoch": 0.6716417910447762,
-      "grad_norm": 1.158009309923302,
-      "learning_rate": 2.340557839064751e-06,
-      "loss": 1.2206,
-      "step": 315
-    },
-    {
-      "epoch": 0.6823027718550106,
-      "grad_norm": 1.120129915614599,
-      "learning_rate": 2.261332678450931e-06,
-      "loss": 1.1631,
-      "step": 320
-    },
-    {
-      "epoch": 0.6929637526652452,
-      "grad_norm": 1.2657256378770898,
-      "learning_rate": 2.184606506864227e-06,
-      "loss": 1.2707,
-      "step": 325
-    },
-    {
-      "epoch": 0.7036247334754797,
-      "grad_norm": 1.1637951032951959,
-      "learning_rate": 2.11030815635039e-06,
-      "loss": 1.2491,
-      "step": 330
-    },
-    {
-      "epoch": 0.7142857142857143,
-      "grad_norm": 1.2287205505399468,
-      "learning_rate": 2.0383682662661417e-06,
-      "loss": 1.22,
-      "step": 335
-    },
-    {
-      "epoch": 0.7249466950959488,
-      "grad_norm": 1.326574025149143,
-      "learning_rate": 1.9687192429912924e-06,
-      "loss": 1.2984,
-      "step": 340
-    },
-    {
-      "epoch": 0.7356076759061834,
-      "grad_norm": 1.1554026106853619,
-      "learning_rate": 1.9012952204130788e-06,
-      "loss": 1.2208,
-      "step": 345
-    },
-    {
-      "epoch": 0.746268656716418,
-      "grad_norm": 1.0875768598668099,
-      "learning_rate": 1.8360320211704227e-06,
-      "loss": 1.2204,
-      "step": 350
-    },
-    {
-      "epoch": 0.7569296375266524,
-      "grad_norm": 1.14982588486912,
-      "learning_rate": 1.772867118645806e-06,
-      "loss": 1.2091,
-      "step": 355
-    },
-    {
-      "epoch": 0.767590618336887,
-      "grad_norm": 1.1836213945640173,
-      "learning_rate": 1.7117395996927912e-06,
-      "loss": 1.2247,
-      "step": 360
-    },
-    {
-      "epoch": 0.7782515991471215,
-      "grad_norm": 1.1378213023095496,
-      "learning_rate": 1.6525901280871983e-06,
-      "loss": 1.1597,
-      "step": 365
-    },
-    {
-      "epoch": 0.7889125799573561,
-      "grad_norm": 1.146991526368693,
-      "learning_rate": 1.5953609086902758e-06,
-      "loss": 1.165,
-      "step": 370
-    },
-    {
-      "epoch": 0.7995735607675906,
-      "grad_norm": 1.0759826941866089,
-      "learning_rate": 1.539995652312188e-06,
-      "loss": 1.1934,
-      "step": 375
-    },
-    {
-      "epoch": 0.8102345415778252,
-      "grad_norm": 1.140264301703701,
-      "learning_rate": 1.486439541264454e-06,
-      "loss": 1.2476,
-      "step": 380
-    },
-    {
-      "epoch": 0.8208955223880597,
-      "grad_norm": 1.1750555438236339,
-      "learning_rate": 1.434639195589973e-06,
-      "loss": 1.2283,
-      "step": 385
-    },
-    {
-      "epoch": 0.8315565031982942,
-      "grad_norm": 1.1027256002443178,
-      "learning_rate": 1.384542639959556e-06,
-      "loss": 1.2392,
-      "step": 390
-    },
-    {
-      "epoch": 0.8422174840085288,
-      "grad_norm": 1.0969958993285855,
-      "learning_rate": 1.336099271223909e-06,
-      "loss": 1.2932,
-      "step": 395
-    },
-    {
-      "epoch": 0.8528784648187633,
-      "grad_norm": 1.1632096677157433,
-      "learning_rate": 1.28925982661026e-06,
-      "loss": 1.1754,
-      "step": 400
-    },
-    {
-      "epoch": 0.8528784648187633,
-      "eval_loss": 1.2225849628448486,
-      "eval_runtime": 86.933,
-      "eval_samples_per_second": 19.176,
-      "eval_steps_per_second": 0.311,
-      "step": 400
-    },
-    {
-      "epoch": 0.8635394456289979,
-      "grad_norm": 1.2018672364738783,
-      "learning_rate": 1.243976352552906e-06,
-      "loss": 1.1537,
-      "step": 405
-    },
-    {
-      "epoch": 0.8742004264392325,
-      "grad_norm": 1.1964994592566494,
-      "learning_rate": 1.2002021741471036e-06,
-      "loss": 1.2153,
-      "step": 410
-    },
-    {
-      "epoch": 0.8848614072494669,
-      "grad_norm": 1.2085715531383925,
-      "learning_rate": 1.1578918652158905e-06,
-      "loss": 1.2214,
-      "step": 415
-    },
-    {
-      "epoch": 0.8955223880597015,
-      "grad_norm": 1.132268550904319,
-      "learning_rate": 1.1170012189795644e-06,
-      "loss": 1.1696,
-      "step": 420
-    },
-    {
-      "epoch": 0.906183368869936,
-      "grad_norm": 1.1888002460068672,
-      "learning_rate": 1.0774872193176617e-06,
-      "loss": 1.2102,
-      "step": 425
-    },
-    {
-      "epoch": 0.9168443496801706,
-      "grad_norm": 1.081884966553129,
-      "learning_rate": 1.039308012613421e-06,
-      "loss": 1.1888,
-      "step": 430
-    },
-    {
-      "epoch": 0.9275053304904051,
-      "grad_norm": 1.1082024653231661,
-      "learning_rate": 1.0024228801708736e-06,
-      "loss": 1.3088,
-      "step": 435
-    },
-    {
-      "epoch": 0.9381663113006397,
-      "grad_norm": 1.1424196623047278,
-      "learning_rate": 9.667922111948187e-07,
-      "loss": 1.2702,
-      "step": 440
-    },
-    {
-      "epoch": 0.9488272921108742,
-      "grad_norm": 1.1515848618493325,
-      "learning_rate": 9.323774763240816e-07,
-      "loss": 1.2241,
-      "step": 445
-    },
-    {
-      "epoch": 0.9594882729211087,
-      "grad_norm": 1.133219633313401,
-      "learning_rate": 8.991412017085673e-07,
-      "loss": 1.1689,
-      "step": 450
-    },
-    {
-      "epoch": 0.9701492537313433,
-      "grad_norm": 1.1817034727115894,
-      "learning_rate": 8.670469436207778e-07,
-      "loss": 1.2429,
-      "step": 455
-    },
-    {
-      "epoch": 0.9808102345415778,
-      "grad_norm": 1.082929968453826,
-      "learning_rate": 8.360592635925826e-07,
-      "loss": 1.223,
-      "step": 460
-    },
-    {
-      "epoch": 0.9914712153518124,
-      "grad_norm": 1.0995506789476814,
-      "learning_rate": 8.061437040681486e-07,
-      "loss": 1.2493,
-      "step": 465
-    },
-    {
-      "epoch": 1.0021321961620469,
-      "grad_norm": 1.4325931391365254,
-      "learning_rate": 7.772667645640643e-07,
-      "loss": 1.2061,
-      "step": 470
-    },
-    {
-      "epoch": 1.0127931769722816,
-      "grad_norm": 1.3020611480277295,
-      "learning_rate": 7.493958783278302e-07,
-      "loss": 1.118,
-      "step": 475
-    },
-    {
-      "epoch": 1.023454157782516,
-      "grad_norm": 1.166419988285865,
-      "learning_rate": 7.22499389486003e-07,
-      "loss": 1.0997,
-      "step": 480
-    },
-    {
-      "epoch": 1.0341151385927505,
-      "grad_norm": 1.1738691779163593,
-      "learning_rate": 6.965465306734048e-07,
-      "loss": 1.0648,
-      "step": 485
-    },
-    {
-      "epoch": 1.044776119402985,
-      "grad_norm": 1.1702760912470462,
-      "learning_rate": 6.715074011349123e-07,
-      "loss": 1.125,
-      "step": 490
-    },
-    {
-      "epoch": 1.0554371002132197,
-      "grad_norm": 1.1386250553771318,
-      "learning_rate": 6.473529452915077e-07,
-      "loss": 1.12,
-      "step": 495
-    },
-    {
-      "epoch": 1.0660980810234542,
-      "grad_norm": 1.145825473750238,
-      "learning_rate": 6.240549317623237e-07,
-      "loss": 1.1182,
-      "step": 500
-    },
-    {
-      "epoch": 1.0767590618336886,
-      "grad_norm": 1.1263260101281638,
-      "learning_rate": 6.015859328346043e-07,
-      "loss": 1.0333,
-      "step": 505
-    },
-    {
-      "epoch": 1.0874200426439233,
-      "grad_norm": 1.099611354066211,
-      "learning_rate": 5.799193043735526e-07,
-      "loss": 1.0545,
-      "step": 510
-    },
-    {
-      "epoch": 1.0980810234541578,
-      "grad_norm": 1.2122056875352134,
-      "learning_rate": 5.590291661642085e-07,
-      "loss": 1.1501,
-      "step": 515
-    },
-    {
-      "epoch": 1.1087420042643923,
-      "grad_norm": 1.2108073336890413,
-      "learning_rate": 5.388903826775655e-07,
-      "loss": 1.1355,
-      "step": 520
-    },
-    {
-      "epoch": 1.1194029850746268,
-      "grad_norm": 1.1869552548231974,
-      "learning_rate": 5.194785442532827e-07,
-      "loss": 1.0177,
-      "step": 525
-    },
-    {
-      "epoch": 1.1300639658848615,
-      "grad_norm": 1.0725136686500363,
-      "learning_rate": 5.007699486914442e-07,
-      "loss": 1.0673,
-      "step": 530
-    },
-    {
-      "epoch": 1.140724946695096,
-      "grad_norm": 1.1715291176817577,
-      "learning_rate": 4.827415832459197e-07,
-      "loss": 0.9923,
-      "step": 535
-    },
-    {
-      "epoch": 1.1513859275053304,
-      "grad_norm": 1.0912818419343246,
-      "learning_rate": 4.6537110701200646e-07,
-      "loss": 1.0214,
-      "step": 540
-    },
-    {
-      "epoch": 1.1620469083155651,
-      "grad_norm": 1.2075583423687697,
-      "learning_rate": 4.486368337011321e-07,
-      "loss": 1.0799,
-      "step": 545
-    },
-    {
-      "epoch": 1.1727078891257996,
-      "grad_norm": 1.1711621672939032,
-      "learning_rate": 4.3251771479550135e-07,
-      "loss": 1.0649,
-      "step": 550
-    },
-    {
-      "epoch": 1.183368869936034,
-      "grad_norm": 1.1379570510356936,
-      "learning_rate": 4.169933230756739e-07,
-      "loss": 1.0275,
-      "step": 555
-    },
-    {
-      "epoch": 1.1940298507462686,
-      "grad_norm": 1.1805410178961084,
-      "learning_rate": 4.020438365141708e-07,
-      "loss": 1.0501,
-      "step": 560
-    },
-    {
-      "epoch": 1.2046908315565032,
-      "grad_norm": 1.0319879103444831,
-      "learning_rate": 3.876500225283097e-07,
-      "loss": 1.0616,
-      "step": 565
-    },
-    {
-      "epoch": 1.2153518123667377,
-      "grad_norm": 1.1205106100162852,
-      "learning_rate": 3.7379322258556177e-07,
-      "loss": 1.0679,
-      "step": 570
-    },
-    {
-      "epoch": 1.2260127931769722,
-      "grad_norm": 1.1518728152224136,
-      "learning_rate": 3.6045533715483096e-07,
-      "loss": 1.1601,
-      "step": 575
-    },
-    {
-      "epoch": 1.236673773987207,
-      "grad_norm": 1.10359450734736,
-      "learning_rate": 3.4761881099715165e-07,
-      "loss": 1.0466,
-      "step": 580
-    },
-    {
-      "epoch": 1.2473347547974414,
-      "grad_norm": 1.1001281681468988,
-      "learning_rate": 3.3526661878940757e-07,
-      "loss": 1.114,
-      "step": 585
-    },
-    {
-      "epoch": 1.2579957356076759,
-      "grad_norm": 1.1513825927177352,
-      "learning_rate": 3.233822510747588e-07,
-      "loss": 1.1314,
-      "step": 590
-    },
-    {
-      "epoch": 1.2686567164179103,
-      "grad_norm": 1.193698165116124,
-      "learning_rate": 3.119497005335667e-07,
-      "loss": 1.0904,
-      "step": 595
-    },
-    {
-      "epoch": 1.279317697228145,
-      "grad_norm": 1.1645268517569372,
-      "learning_rate": 3.0095344856870186e-07,
-      "loss": 1.1662,
-      "step": 600
-    },
-    {
-      "epoch": 1.279317697228145,
-      "eval_loss": 1.220153570175171,
-      "eval_runtime": 86.8011,
-      "eval_samples_per_second": 19.205,
-      "eval_steps_per_second": 0.311,
-      "step": 600
-    },
-    {
-      "epoch": 1.2899786780383795,
-      "grad_norm": 1.1865685508079404,
-      "learning_rate": 2.903784521992152e-07,
-      "loss": 1.0543,
-      "step": 605
-    },
-    {
-      "epoch": 1.3006396588486142,
-      "grad_norm": 1.0889008660105814,
-      "learning_rate": 2.802101312564394e-07,
-      "loss": 1.1059,
-      "step": 610
-    },
-    {
-      "epoch": 1.3113006396588487,
-      "grad_norm": 1.1629452168406549,
-      "learning_rate": 2.7043435587668195e-07,
-      "loss": 1.0847,
-      "step": 615
-    },
-    {
-      "epoch": 1.3219616204690832,
-      "grad_norm": 1.1225822453511047,
-      "learning_rate": 2.6103743428476387e-07,
-      "loss": 1.0547,
-      "step": 620
-    },
-    {
-      "epoch": 1.3326226012793176,
-      "grad_norm": 1.068456227087864,
-      "learning_rate": 2.520061008627476e-07,
-      "loss": 1.1331,
-      "step": 625
-    },
-    {
-      "epoch": 1.3432835820895521,
-      "grad_norm": 1.163340113599676,
-      "learning_rate": 2.4332750449828163e-07,
-      "loss": 1.0907,
-      "step": 630
-    },
-    {
-      "epoch": 1.3539445628997868,
-      "grad_norm": 1.1623906096012029,
-      "learning_rate": 2.3498919720707982e-07,
-      "loss": 1.0788,
-      "step": 635
-    },
-    {
-      "epoch": 1.3646055437100213,
-      "grad_norm": 1.1079974563308679,
-      "learning_rate": 2.269791230241396e-07,
-      "loss": 1.0701,
-      "step": 640
-    },
-    {
-      "epoch": 1.375266524520256,
-      "grad_norm": 1.2149865496581391,
-      "learning_rate": 2.1928560715838997e-07,
-      "loss": 1.0376,
-      "step": 645
-    },
-    {
-      "epoch": 1.3859275053304905,
-      "grad_norm": 1.1528488687976461,
-      "learning_rate": 2.1189734540554226e-07,
-      "loss": 1.1269,
-      "step": 650
-    },
-    {
-      "epoch": 1.396588486140725,
-      "grad_norm": 1.1288464186180127,
-      "learning_rate": 2.048033938139964e-07,
-      "loss": 1.108,
-      "step": 655
-    },
-    {
-      "epoch": 1.4072494669509594,
-      "grad_norm": 1.1817093681210782,
-      "learning_rate": 1.979931585987499e-07,
-      "loss": 1.0499,
-      "step": 660
-    },
-    {
-      "epoch": 1.417910447761194,
-      "grad_norm": 1.1321827623646836,
-      "learning_rate": 1.9145638629832107e-07,
-      "loss": 1.041,
-      "step": 665
-    },
-    {
-      "epoch": 1.4285714285714286,
-      "grad_norm": 1.15181492674688,
-      "learning_rate": 1.8518315416979456e-07,
-      "loss": 1.0994,
-      "step": 670
-    },
-    {
-      "epoch": 1.439232409381663,
-      "grad_norm": 1.1127579923456647,
-      "learning_rate": 1.7916386081716564e-07,
-      "loss": 1.0894,
-      "step": 675
-    },
-    {
-      "epoch": 1.4498933901918978,
-      "grad_norm": 1.1765219547043915,
-      "learning_rate": 1.7338921704824148e-07,
-      "loss": 1.0715,
-      "step": 680
-    },
-    {
-      "epoch": 1.4605543710021323,
-      "grad_norm": 1.2252010284139379,
-      "learning_rate": 1.678502369554362e-07,
-      "loss": 1.1498,
-      "step": 685
-    },
-    {
-      "epoch": 1.4712153518123667,
-      "grad_norm": 1.1121586595467274,
-      "learning_rate": 1.6253822921587308e-07,
-      "loss": 1.0357,
-      "step": 690
-    },
-    {
-      "epoch": 1.4818763326226012,
-      "grad_norm": 1.2326653113155133,
-      "learning_rate": 1.574447886062802e-07,
-      "loss": 1.0426,
-      "step": 695
-    },
-    {
-      "epoch": 1.4925373134328357,
-      "grad_norm": 1.2662375322966155,
-      "learning_rate": 1.5256178772824204e-07,
-      "loss": 1.1513,
-      "step": 700
-    },
-    {
-      "epoch": 1.5031982942430704,
-      "grad_norm": 1.1537609553653134,
-      "learning_rate": 1.4788136893944296e-07,
-      "loss": 1.0303,
-      "step": 705
-    },
-    {
-      "epoch": 1.5138592750533049,
-      "grad_norm": 1.0977344228272827,
-      "learning_rate": 1.4339593648661315e-07,
-      "loss": 1.0821,
-      "step": 710
-    },
-    {
-      "epoch": 1.5245202558635396,
-      "grad_norm": 1.1284164855389234,
-      "learning_rate": 1.3909814883595595e-07,
-      "loss": 1.1313,
-      "step": 715
-    },
-    {
-      "epoch": 1.535181236673774,
-      "grad_norm": 1.1938382106881487,
-      "learning_rate": 1.3498091119690798e-07,
-      "loss": 1.1446,
-      "step": 720
-    },
-    {
-      "epoch": 1.5458422174840085,
-      "grad_norm": 1.1272930245655628,
-      "learning_rate": 1.3103736823515384e-07,
-      "loss": 1.1136,
-      "step": 725
-    },
-    {
-      "epoch": 1.556503198294243,
-      "grad_norm": 1.1773751232659426,
-      "learning_rate": 1.2726089697088728e-07,
-      "loss": 1.1309,
-      "step": 730
-    },
-    {
-      "epoch": 1.5671641791044775,
-      "grad_norm": 1.3277025055368337,
-      "learning_rate": 1.2364509985837655e-07,
-      "loss": 1.0239,
-      "step": 735
-    },
-    {
-      "epoch": 1.5778251599147122,
-      "grad_norm": 1.230962189006197,
-      "learning_rate": 1.201837980429599e-07,
-      "loss": 1.058,
-      "step": 740
-    },
-    {
-      "epoch": 1.5884861407249466,
-      "grad_norm": 1.2101453511321207,
-      "learning_rate": 1.1687102479166401e-07,
-      "loss": 1.1119,
-      "step": 745
-    },
-    {
-      "epoch": 1.5991471215351813,
-      "grad_norm": 1.1936647152626743,
-      "learning_rate": 1.1370101909370577e-07,
-      "loss": 1.1556,
-      "step": 750
-    },
-    {
-      "epoch": 1.6098081023454158,
-      "grad_norm": 1.1474035015035384,
-      "learning_rate": 1.1066821942719791e-07,
-      "loss": 1.0522,
-      "step": 755
-    },
-    {
-      "epoch": 1.6204690831556503,
-      "grad_norm": 1.0637295000469429,
-      "learning_rate": 1.0776725768844699e-07,
-      "loss": 1.0549,
-      "step": 760
-    },
-    {
-      "epoch": 1.6311300639658848,
-      "grad_norm": 1.10828124909061,
-      "learning_rate": 1.0499295328029333e-07,
-      "loss": 1.043,
-      "step": 765
-    },
-    {
-      "epoch": 1.6417910447761193,
-      "grad_norm": 1.1283065722653947,
-      "learning_rate": 1.0234030735600652e-07,
-      "loss": 1.0731,
-      "step": 770
-    },
-    {
-      "epoch": 1.652452025586354,
-      "grad_norm": 1.161403848045461,
-      "learning_rate": 9.980449721530949e-08,
-      "loss": 1.161,
-      "step": 775
-    },
-    {
-      "epoch": 1.6631130063965884,
-      "grad_norm": 1.158445375876672,
-      "learning_rate": 9.738087084916553e-08,
-      "loss": 1.089,
-      "step": 780
-    },
-    {
-      "epoch": 1.6737739872068231,
-      "grad_norm": 1.1752425329283938,
-      "learning_rate": 9.506494163002325e-08,
-      "loss": 1.0479,
-      "step": 785
-    },
-    {
-      "epoch": 1.6844349680170576,
-      "grad_norm": 1.0813888422827629,
-      "learning_rate": 9.285238314427303e-08,
-      "loss": 1.0797,
-      "step": 790
-    },
-    {
-      "epoch": 1.695095948827292,
-      "grad_norm": 1.2358809521231273,
-      "learning_rate": 9.07390241637265e-08,
-      "loss": 1.0483,
-      "step": 795
-    },
-    {
-      "epoch": 1.7057569296375266,
-      "grad_norm": 1.1520098737400197,
-      "learning_rate": 8.872084375298702e-08,
-      "loss": 1.1182,
-      "step": 800
-    },
-    {
-      "epoch": 1.7057569296375266,
-      "eval_loss": 1.2184005975723267,
-      "eval_runtime": 86.9323,
-      "eval_samples_per_second": 19.176,
-      "eval_steps_per_second": 0.311,
-      "step": 800
-    },
-    {
-      "epoch": 1.716417910447761,
-      "grad_norm": 1.1409424379253437,
-      "learning_rate": 8.679396650963928e-08,
-      "loss": 1.1356,
-      "step": 805
-    },
-    {
-      "epoch": 1.7270788912579957,
-      "grad_norm": 1.1235815999344876,
-      "learning_rate": 8.49546579342372e-08,
-      "loss": 1.155,
-      "step": 810
-    },
-    {
-      "epoch": 1.7377398720682304,
-      "grad_norm": 1.1073975656892132,
-      "learning_rate": 8.319931992712862e-08,
-      "loss": 1.1137,
-      "step": 815
-    },
-    {
-      "epoch": 1.748400852878465,
-      "grad_norm": 1.2180389136635439,
-      "learning_rate": 8.15244864092076e-08,
-      "loss": 1.1422,
-      "step": 820
-    },
-    {
-      "epoch": 1.7590618336886994,
-      "grad_norm": 1.1113266470685994,
-      "learning_rate": 7.992681906373824e-08,
-      "loss": 1.062,
-      "step": 825
-    },
-    {
-      "epoch": 1.7697228144989339,
-      "grad_norm": 1.2072670273798571,
-      "learning_rate": 7.84031031964486e-08,
-      "loss": 1.0517,
-      "step": 830
-    },
-    {
-      "epoch": 1.7803837953091683,
-      "grad_norm": 1.1803556915432551,
-      "learning_rate": 7.69502437111444e-08,
-      "loss": 1.0703,
-      "step": 835
-    },
-    {
-      "epoch": 1.7910447761194028,
-      "grad_norm": 1.1844005429106652,
-      "learning_rate": 7.556526119814338e-08,
-      "loss": 1.111,
-      "step": 840
-    },
-    {
-      "epoch": 1.8017057569296375,
-      "grad_norm": 1.1384485747440536,
-      "learning_rate": 7.424528813288221e-08,
-      "loss": 1.1914,
-      "step": 845
-    },
-    {
-      "epoch": 1.8123667377398722,
-      "grad_norm": 1.2171711222670063,
-      "learning_rate": 7.298756518209771e-08,
-      "loss": 1.118,
-      "step": 850
-    },
-    {
-      "epoch": 1.8230277185501067,
-      "grad_norm": 1.1361749350155368,
-      "learning_rate": 7.178943761503399e-08,
-      "loss": 1.2114,
-      "step": 855
-    },
-    {
-      "epoch": 1.8336886993603412,
-      "grad_norm": 1.2202066733922106,
-      "learning_rate": 7.064835181717455e-08,
-      "loss": 1.0826,
-      "step": 860
-    },
-    {
-      "epoch": 1.8443496801705757,
-      "grad_norm": 1.1962899610991478,
-      "learning_rate": 6.956185190404699e-08,
-      "loss": 1.0077,
-      "step": 865
-    },
-    {
-      "epoch": 1.8550106609808101,
-      "grad_norm": 1.2308765981536185,
-      "learning_rate": 6.85275764326948e-08,
-      "loss": 1.0717,
-      "step": 870
-    },
-    {
-      "epoch": 1.8656716417910446,
-      "grad_norm": 1.1658077017002657,
-      "learning_rate": 6.75432552084579e-08,
-      "loss": 1.0958,
-      "step": 875
-    },
-    {
-      "epoch": 1.8763326226012793,
-      "grad_norm": 1.270006074619097,
-      "learning_rate": 6.660670618474825e-08,
-      "loss": 1.091,
-      "step": 880
-    },
-    {
-      "epoch": 1.886993603411514,
-      "grad_norm": 1.1582088190520023,
-      "learning_rate": 6.571583245355274e-08,
-      "loss": 1.0301,
-      "step": 885
-    },
-    {
-      "epoch": 1.8976545842217485,
-      "grad_norm": 1.0859957939920986,
-      "learning_rate": 6.486861932443996e-08,
-      "loss": 1.0629,
-      "step": 890
-    },
-    {
-      "epoch": 1.908315565031983,
-      "grad_norm": 1.0879133073671061,
-      "learning_rate": 6.406313148989106e-08,
-      "loss": 1.0293,
-      "step": 895
-    },
-    {
-      "epoch": 1.9189765458422174,
-      "grad_norm": 1.1419672670758283,
-      "learning_rate": 6.329751027481845e-08,
-      "loss": 1.1259,
-      "step": 900
-    },
-    {
-      "epoch": 1.929637526652452,
-      "grad_norm": 1.147613866551749,
-      "learning_rate": 6.256997096817788e-08,
-      "loss": 1.0128,
-      "step": 905
-    },
-    {
-      "epoch": 1.9402985074626866,
-      "grad_norm": 1.1753147588481525,
-      "learning_rate": 6.187880023462194e-08,
-      "loss": 1.0407,
-      "step": 910
-    },
-    {
-      "epoch": 1.950959488272921,
-      "grad_norm": 1.128791859643809,
-      "learning_rate": 6.122235360418442e-08,
-      "loss": 1.0722,
-      "step": 915
-    },
-    {
-      "epoch": 1.9616204690831558,
-      "grad_norm": 1.1601302162296214,
-      "learning_rate": 6.059905303802484e-08,
-      "loss": 1.1054,
-      "step": 920
-    },
-    {
-      "epoch": 1.9722814498933903,
-      "grad_norm": 1.198563476242119,
-      "learning_rate": 6.000738456830302e-08,
-      "loss": 1.0568,
-      "step": 925
-    },
-    {
-      "epoch": 1.9829424307036247,
-      "grad_norm": 1.2262616318390043,
-      "learning_rate": 5.944589601029287e-08,
-      "loss": 1.1251,
-      "step": 930
-    },
-    {
-      "epoch": 1.9936034115138592,
-      "grad_norm": 1.191995506184399,
-      "learning_rate": 5.89131947448833e-08,
-      "loss": 1.1911,
-      "step": 935
-    },
-    {
-      "epoch": 2.0042643923240937,
-      "grad_norm": 1.11244803877364,
-      "learning_rate": 5.840794556965235e-08,
-      "loss": 1.066,
-      "step": 940
-    },
-    {
-      "epoch": 2.014925373134328,
-      "grad_norm": 1.0894580884709386,
-      "learning_rate": 5.792886861673804e-08,
-      "loss": 1.0406,
-      "step": 945
-    },
-    {
-      "epoch": 2.025586353944563,
-      "grad_norm": 1.0653919782073382,
-      "learning_rate": 5.747473733576709e-08,
-      "loss": 1.0661,
-      "step": 950
-    },
-    {
-      "epoch": 2.0362473347547976,
-      "grad_norm": 1.1280141717212646,
-      "learning_rate": 5.704437654013792e-08,
-      "loss": 1.1027,
-      "step": 955
-    },
-    {
-      "epoch": 2.046908315565032,
-      "grad_norm": 1.2313007226700692,
-      "learning_rate": 5.663666051499173e-08,
-      "loss": 1.0421,
-      "step": 960
-    },
-    {
-      "epoch": 2.0575692963752665,
-      "grad_norm": 1.153128267269016,
-      "learning_rate": 5.6250511185239e-08,
-      "loss": 1.0893,
-      "step": 965
-    },
-    {
-      "epoch": 2.068230277185501,
-      "grad_norm": 1.1364871737157245,
-      "learning_rate": 5.58848963420453e-08,
-      "loss": 1.0918,
-      "step": 970
-    },
-    {
-      "epoch": 2.0788912579957355,
-      "grad_norm": 1.1584140858402872,
-      "learning_rate": 5.553882792621254e-08,
-      "loss": 1.0183,
-      "step": 975
-    },
-    {
-      "epoch": 2.08955223880597,
-      "grad_norm": 1.1608258928350752,
-      "learning_rate": 5.521136036692734e-08,
-      "loss": 1.0322,
-      "step": 980
-    },
-    {
-      "epoch": 2.100213219616205,
-      "grad_norm": 1.1873514358148416,
-      "learning_rate": 5.490158897437947e-08,
-      "loss": 1.0441,
-      "step": 985
-    },
-    {
-      "epoch": 2.1108742004264394,
-      "grad_norm": 1.202520887860229,
-      "learning_rate": 5.460864838478731e-08,
-      "loss": 1.0996,
-      "step": 990
-    },
-    {
-      "epoch": 2.121535181236674,
-      "grad_norm": 1.1559964888973397,
-      "learning_rate": 5.4331711056398124e-08,
-      "loss": 1.095,
-      "step": 995
-    },
-    {
-      "epoch": 2.1321961620469083,
-      "grad_norm": 1.0455908941955974,
-      "learning_rate": 5.4069985815063316e-08,
-      "loss": 1.046,
-      "step": 1000
-    },
-    {
-      "epoch": 2.1321961620469083,
-      "eval_loss": 1.2190189361572266,
-      "eval_runtime": 86.8993,
-      "eval_samples_per_second": 19.183,
-      "eval_steps_per_second": 0.311,
-      "step": 1000
-    },
-    {
-      "epoch": 2.142857142857143,
-      "grad_norm": 1.1258833797291203,
-      "learning_rate": 5.3822716448019046e-08,
-      "loss": 1.049,
-      "step": 1005
-    },
-    {
-      "epoch": 2.1535181236673773,
-      "grad_norm": 1.162003500965812,
-      "learning_rate": 5.35891803445336e-08,
-      "loss": 1.094,
-      "step": 1010
-    },
-    {
-      "epoch": 2.1641791044776117,
-      "grad_norm": 1.1823939265510066,
-      "learning_rate": 5.3368687182112245e-08,
-      "loss": 1.0556,
-      "step": 1015
-    },
-    {
-      "epoch": 2.1748400852878467,
-      "grad_norm": 1.1584053741977514,
-      "learning_rate": 5.3160577656980396e-08,
-      "loss": 1.1292,
-      "step": 1020
-    },
-    {
-      "epoch": 2.185501066098081,
-      "grad_norm": 1.1507740539032782,
-      "learning_rate": 5.296422225759416e-08,
-      "loss": 1.1023,
-      "step": 1025
-    },
-    {
-      "epoch": 2.1961620469083156,
-      "grad_norm": 1.1594756240093598,
-      "learning_rate": 5.277902007995615e-08,
-      "loss": 1.107,
-      "step": 1030
-    },
-    {
-      "epoch": 2.20682302771855,
-      "grad_norm": 1.10929463120434,
-      "learning_rate": 5.2604397683542377e-08,
-      "loss": 1.1102,
-      "step": 1035
-    },
-    {
-      "epoch": 2.2174840085287846,
-      "grad_norm": 1.1432334524639647,
-      "learning_rate": 5.243980798667319e-08,
-      "loss": 1.1,
-      "step": 1040
-    },
-    {
-      "epoch": 2.228144989339019,
-      "grad_norm": 1.1904483860983543,
-      "learning_rate": 5.228472920018869e-08,
-      "loss": 1.09,
-      "step": 1045
-    },
-    {
-      "epoch": 2.2388059701492535,
-      "grad_norm": 1.167622664957689,
-      "learning_rate": 5.213866379831496e-08,
-      "loss": 1.0814,
-      "step": 1050
-    },
-    {
-      "epoch": 2.2494669509594885,
-      "grad_norm": 1.0895757958922552,
-      "learning_rate": 5.200113752563408e-08,
-      "loss": 1.1161,
-      "step": 1055
-    },
-    {
-      "epoch": 2.260127931769723,
-      "grad_norm": 1.1558659882895932,
-      "learning_rate": 5.187169843909606e-08,
-      "loss": 1.0795,
-      "step": 1060
-    },
-    {
-      "epoch": 2.2707889125799574,
-      "grad_norm": 1.1344999555588917,
-      "learning_rate": 5.17499159840361e-08,
-      "loss": 1.0706,
-      "step": 1065
-    },
-    {
-      "epoch": 2.281449893390192,
-      "grad_norm": 1.1415261926993812,
-      "learning_rate": 5.163538010318527e-08,
-      "loss": 1.0967,
-      "step": 1070
-    },
-    {
-      "epoch": 2.2921108742004264,
-      "grad_norm": 1.1381490852319662,
-      "learning_rate": 5.152770037768703e-08,
-      "loss": 1.056,
-      "step": 1075
-    },
-    {
-      "epoch": 2.302771855010661,
-      "grad_norm": 1.1347196474571326,
-      "learning_rate": 5.142650519915549e-08,
-      "loss": 1.0654,
-      "step": 1080
-    },
-    {
-      "epoch": 2.3134328358208958,
-      "grad_norm": 1.140691001232856,
-      "learning_rate": 5.133144097183511e-08,
-      "loss": 1.0555,
-      "step": 1085
-    },
-    {
-      "epoch": 2.3240938166311302,
-      "grad_norm": 1.1321293656672715,
-      "learning_rate": 5.1242171343944076e-08,
-      "loss": 1.0617,
-      "step": 1090
-    },
-    {
-      "epoch": 2.3347547974413647,
-      "grad_norm": 1.0730144292666668,
-      "learning_rate": 5.115837646730644e-08,
-      "loss": 1.0129,
-      "step": 1095
-    },
-    {
-      "epoch": 2.345415778251599,
-      "grad_norm": 1.1568722548151742,
-      "learning_rate": 5.107975228439987e-08,
-      "loss": 1.1102,
-      "step": 1100
-    },
-    {
-      "epoch": 2.3560767590618337,
-      "grad_norm": 1.1165138012512623,
-      "learning_rate": 5.100600984196796e-08,
-      "loss": 1.0316,
-      "step": 1105
-    },
-    {
-      "epoch": 2.366737739872068,
-      "grad_norm": 1.1003671418770582,
-      "learning_rate": 5.093687463036684e-08,
-      "loss": 1.0387,
-      "step": 1110
-    },
-    {
-      "epoch": 2.3773987206823026,
-      "grad_norm": 1.1858341121422311,
-      "learning_rate": 5.087208594783712e-08,
-      "loss": 1.1381,
-      "step": 1115
-    },
-    {
-      "epoch": 2.388059701492537,
-      "grad_norm": 1.1724486836357029,
-      "learning_rate": 5.0811396288912494e-08,
-      "loss": 1.1484,
-      "step": 1120
-    },
-    {
-      "epoch": 2.398720682302772,
-      "grad_norm": 1.1477785681404549,
-      "learning_rate": 5.075457075619642e-08,
-      "loss": 1.0488,
-      "step": 1125
-    },
-    {
-      "epoch": 2.4093816631130065,
-      "grad_norm": 1.1926168382576205,
-      "learning_rate": 5.070138649475803e-08,
-      "loss": 1.0335,
-      "step": 1130
-    },
-    {
-      "epoch": 2.420042643923241,
-      "grad_norm": 1.1592591417618787,
-      "learning_rate": 5.065163214841775e-08,
-      "loss": 1.0867,
-      "step": 1135
-    },
-    {
-      "epoch": 2.4307036247334755,
-      "grad_norm": 1.1492395746059412,
-      "learning_rate": 5.0605107337212035e-08,
-      "loss": 1.0934,
-      "step": 1140
-    },
-    {
-      "epoch": 2.44136460554371,
-      "grad_norm": 1.1808862259382695,
-      "learning_rate": 5.0561622155345265e-08,
-      "loss": 1.0804,
-      "step": 1145
-    },
-    {
-      "epoch": 2.4520255863539444,
-      "grad_norm": 1.2148315415559092,
-      "learning_rate": 5.052099668895486e-08,
-      "loss": 1.1264,
-      "step": 1150
-    },
-    {
-      "epoch": 2.4626865671641793,
-      "grad_norm": 1.0677698199987926,
-      "learning_rate": 5.0483060553033895e-08,
-      "loss": 1.029,
-      "step": 1155
-    },
-    {
-      "epoch": 2.473347547974414,
-      "grad_norm": 1.1698932019494643,
-      "learning_rate": 5.0447652446872465e-08,
-      "loss": 1.028,
-      "step": 1160
-    },
-    {
-      "epoch": 2.4840085287846483,
-      "grad_norm": 1.1287646587327143,
-      "learning_rate": 5.0414619727396756e-08,
-      "loss": 1.0884,
-      "step": 1165
-    },
-    {
-      "epoch": 2.4946695095948828,
-      "grad_norm": 1.16856773725038,
-      "learning_rate": 5.0383817999801064e-08,
-      "loss": 1.0204,
-      "step": 1170
-    },
-    {
-      "epoch": 2.5053304904051172,
-      "grad_norm": 1.141453152240354,
-      "learning_rate": 5.035511072488479e-08,
-      "loss": 1.0523,
-      "step": 1175
-    },
-    {
-      "epoch": 2.5159914712153517,
-      "grad_norm": 1.230429011852803,
-      "learning_rate": 5.032836884252246e-08,
-      "loss": 1.0617,
-      "step": 1180
-    },
-    {
-      "epoch": 2.526652452025586,
-      "grad_norm": 1.151378135213026,
-      "learning_rate": 5.03034704107104e-08,
-      "loss": 1.048,
-      "step": 1185
-    },
-    {
-      "epoch": 2.5373134328358207,
-      "grad_norm": 1.1297760039941285,
-      "learning_rate": 5.028030025964957e-08,
-      "loss": 1.0614,
-      "step": 1190
-    },
-    {
-      "epoch": 2.5479744136460556,
-      "grad_norm": 1.149687991991871,
-      "learning_rate": 5.025874966033864e-08,
-      "loss": 1.1042,
-      "step": 1195
-    },
-    {
-      "epoch": 2.55863539445629,
-      "grad_norm": 1.1935154857243262,
-      "learning_rate": 5.023871600716678e-08,
-      "loss": 1.0772,
-      "step": 1200
-    },
-    {
-      "epoch": 2.55863539445629,
-      "eval_loss": 1.2190290689468384,
-      "eval_runtime": 86.4625,
-      "eval_samples_per_second": 19.28,
-      "eval_steps_per_second": 0.312,
-      "step": 1200
-    },
-    {
-      "epoch": 2.5692963752665245,
-      "grad_norm": 1.1485570257886584,
-      "learning_rate": 5.022010251400949e-08,
-      "loss": 1.1083,
-      "step": 1205
-    },
-    {
-      "epoch": 2.579957356076759,
-      "grad_norm": 1.180867296316843,
-      "learning_rate": 5.0202817923345707e-08,
-      "loss": 1.062,
-      "step": 1210
-    },
-    {
-      "epoch": 2.5906183368869935,
-      "grad_norm": 1.1667090859024647,
-      "learning_rate": 5.0186776227927486e-08,
-      "loss": 1.0414,
-      "step": 1215
-    },
-    {
-      "epoch": 2.6012793176972284,
-      "grad_norm": 1.1582105008250059,
-      "learning_rate": 5.017189640454793e-08,
-      "loss": 1.0663,
-      "step": 1220
-    },
-    {
-      "epoch": 2.611940298507463,
-      "grad_norm": 1.0799260851167845,
-      "learning_rate": 5.01581021594656e-08,
-      "loss": 1.102,
-      "step": 1225
-    },
-    {
-      "epoch": 2.6226012793176974,
-      "grad_norm": 1.1531786446908774,
-      "learning_rate": 5.01453216850572e-08,
-      "loss": 1.1336,
-      "step": 1230
-    },
-    {
-      "epoch": 2.633262260127932,
-      "grad_norm": 1.1432906025568284,
-      "learning_rate": 5.0133487427282597e-08,
-      "loss": 1.0256,
-      "step": 1235
-    },
-    {
-      "epoch": 2.6439232409381663,
-      "grad_norm": 1.1673336163640968,
-      "learning_rate": 5.0122535863558936e-08,
-      "loss": 1.1471,
-      "step": 1240
-    },
-    {
-      "epoch": 2.654584221748401,
-      "grad_norm": 1.1845243824938692,
-      "learning_rate": 5.011240729065255e-08,
-      "loss": 1.0642,
-      "step": 1245
-    },
-    {
-      "epoch": 2.6652452025586353,
-      "grad_norm": 1.1256336089648642,
-      "learning_rate": 5.010304562220935e-08,
-      "loss": 1.1023,
-      "step": 1250
-    },
-    {
-      "epoch": 2.6759061833688698,
-      "grad_norm": 1.1635762261874374,
-      "learning_rate": 5.00943981955558e-08,
-      "loss": 1.0605,
-      "step": 1255
-    },
-    {
-      "epoch": 2.6865671641791042,
-      "grad_norm": 1.1747405452856283,
-      "learning_rate": 5.0086415587414125e-08,
-      "loss": 1.0897,
-      "step": 1260
-    },
-    {
-      "epoch": 2.697228144989339,
-      "grad_norm": 1.086387379481192,
-      "learning_rate": 5.00790514381863e-08,
-      "loss": 1.0893,
-      "step": 1265
-    },
-    {
-      "epoch": 2.7078891257995736,
-      "grad_norm": 1.1923401716553281,
-      "learning_rate": 5.007226228447204e-08,
-      "loss": 1.0473,
-      "step": 1270
-    },
-    {
-      "epoch": 2.718550106609808,
-      "grad_norm": 1.205935131659565,
-      "learning_rate": 5.006600739949697e-08,
-      "loss": 1.0231,
-      "step": 1275
-    },
-    {
-      "epoch": 2.7292110874200426,
-      "grad_norm": 1.1391126635720465,
-      "learning_rate": 5.006024864113683e-08,
-      "loss": 1.0403,
-      "step": 1280
-    },
-    {
-      "epoch": 2.739872068230277,
-      "grad_norm": 1.1345121679379757,
-      "learning_rate": 5.005495030723431e-08,
-      "loss": 1.0764,
-      "step": 1285
-    },
-    {
-      "epoch": 2.750533049040512,
-      "grad_norm": 1.1850716557115912,
-      "learning_rate": 5.005007899791418e-08,
-      "loss": 0.9945,
-      "step": 1290
-    },
-    {
-      "epoch": 2.7611940298507465,
-      "grad_norm": 1.1874286830736782,
-      "learning_rate": 5.00456034846126e-08,
-      "loss": 1.085,
-      "step": 1295
-    },
-    {
-      "epoch": 2.771855010660981,
-      "grad_norm": 1.1901764193670399,
-      "learning_rate": 5.004149458554536e-08,
-      "loss": 1.0092,
-      "step": 1300
-    },
-    {
-      "epoch": 2.7825159914712154,
-      "grad_norm": 1.0961826150665088,
-      "learning_rate": 5.0037725047349065e-08,
-      "loss": 1.1071,
-      "step": 1305
-    },
-    {
-      "epoch": 2.79317697228145,
-      "grad_norm": 1.1628718738526371,
-      "learning_rate": 5.0034269432638164e-08,
-      "loss": 1.1233,
-      "step": 1310
-    },
-    {
-      "epoch": 2.8038379530916844,
-      "grad_norm": 1.1626738934739451,
-      "learning_rate": 5.0031104013229344e-08,
-      "loss": 1.1894,
-      "step": 1315
-    },
-    {
-      "epoch": 2.814498933901919,
-      "grad_norm": 1.1011603049613505,
-      "learning_rate": 5.002820666879318e-08,
-      "loss": 1.1494,
-      "step": 1320
-    },
-    {
-      "epoch": 2.8251599147121533,
-      "grad_norm": 1.1717668628457665,
-      "learning_rate": 5.002555679070117e-08,
-      "loss": 1.0514,
-      "step": 1325
-    },
-    {
-      "epoch": 2.835820895522388,
-      "grad_norm": 1.1774225127445404,
-      "learning_rate": 5.0023135190844316e-08,
-      "loss": 1.1027,
-      "step": 1330
-    },
-    {
-      "epoch": 2.8464818763326227,
-      "grad_norm": 1.1717401192939292,
-      "learning_rate": 5.0020924015207164e-08,
-      "loss": 1.063,
-      "step": 1335
-    },
-    {
-      "epoch": 2.857142857142857,
-      "grad_norm": 1.1587594052870365,
-      "learning_rate": 5.001890666198876e-08,
-      "loss": 1.1021,
-      "step": 1340
-    },
-    {
-      "epoch": 2.8678038379530917,
-      "grad_norm": 1.1525118578792972,
-      "learning_rate": 5.001706770406945e-08,
-      "loss": 1.1019,
-      "step": 1345
-    },
-    {
-      "epoch": 2.878464818763326,
-      "grad_norm": 1.37326829503098,
-      "learning_rate": 5.001539281562949e-08,
-      "loss": 1.0355,
-      "step": 1350
-    },
-    {
-      "epoch": 2.8891257995735606,
-      "grad_norm": 1.1331048627302174,
-      "learning_rate": 5.001386870273264e-08,
-      "loss": 1.0899,
-      "step": 1355
-    },
-    {
-      "epoch": 2.8997867803837956,
-      "grad_norm": 1.1984672662153428,
-      "learning_rate": 5.0012483037694405e-08,
-      "loss": 1.1161,
-      "step": 1360
-    },
-    {
-      "epoch": 2.91044776119403,
-      "grad_norm": 1.1109425480042538,
-      "learning_rate": 5.0011224397061566e-08,
-      "loss": 1.0231,
-      "step": 1365
-    },
-    {
-      "epoch": 2.9211087420042645,
-      "grad_norm": 1.1795722740533605,
-      "learning_rate": 5.0010082203035725e-08,
-      "loss": 1.1211,
-      "step": 1370
-    },
-    {
-      "epoch": 2.931769722814499,
-      "grad_norm": 1.0937372807822898,
-      "learning_rate": 5.0009046668180024e-08,
-      "loss": 1.0735,
-      "step": 1375
-    },
-    {
-      "epoch": 2.9424307036247335,
-      "grad_norm": 1.1405185259805586,
-      "learning_rate": 5.000810874325427e-08,
-      "loss": 1.0276,
-      "step": 1380
-    },
-    {
-      "epoch": 2.953091684434968,
-      "grad_norm": 1.20416756854441,
-      "learning_rate": 5.0007260068029507e-08,
-      "loss": 1.0596,
-      "step": 1385
-    },
-    {
-      "epoch": 2.9637526652452024,
-      "grad_norm": 1.1625071717485667,
-      "learning_rate": 5.000649292493887e-08,
-      "loss": 1.0697,
-      "step": 1390
-    },
-    {
-      "epoch": 2.974413646055437,
-      "grad_norm": 1.1863714583502374,
-      "learning_rate": 5.000580019542706e-08,
-      "loss": 1.1505,
-      "step": 1395
-    },
-    {
-      "epoch": 2.9850746268656714,
-      "grad_norm": 1.1942905354245756,
-      "learning_rate": 5.000517531886622e-08,
-      "loss": 1.0326,
-      "step": 1400
-    },
-    {
-      "epoch": 2.9850746268656714,
-      "eval_loss": 1.2187978029251099,
-      "eval_runtime": 86.8658,
-      "eval_samples_per_second": 19.191,
-      "eval_steps_per_second": 0.311,
-      "step": 1400
-    },
-    {
-      "epoch": 2.9957356076759063,
-      "grad_norm": 1.1519642562669563,
-      "learning_rate": 5.0004612253911274e-08,
-      "loss": 1.0479,
-      "step": 1405
-    },
-    {
-      "epoch": 3.0063965884861408,
-      "grad_norm": 1.1665923395094933,
-      "learning_rate": 5.000410544217268e-08,
-      "loss": 1.0174,
-      "step": 1410
-    },
-    {
-      "epoch": 3.0170575692963753,
-      "grad_norm": 1.1796085918771502,
-      "learning_rate": 5.0003649774089725e-08,
-      "loss": 1.0737,
-      "step": 1415
-    },
-    {
-      "epoch": 3.0277185501066097,
-      "grad_norm": 1.1757596964828239,
-      "learning_rate": 5.0003240556892086e-08,
-      "loss": 1.0738,
-      "step": 1420
-    },
-    {
-      "epoch": 3.038379530916844,
-      "grad_norm": 1.1395407440291647,
-      "learning_rate": 5.000287348454201e-08,
-      "loss": 1.094,
-      "step": 1425
-    },
-    {
-      "epoch": 3.0490405117270787,
-      "grad_norm": 1.1323357375626646,
-      "learning_rate": 5.000254460955416e-08,
-      "loss": 1.0684,
-      "step": 1430
-    },
-    {
-      "epoch": 3.0597014925373136,
-      "grad_norm": 1.1177035566781262,
-      "learning_rate": 5.0002250316594135e-08,
-      "loss": 1.038,
-      "step": 1435
-    },
-    {
-      "epoch": 3.070362473347548,
-      "grad_norm": 1.1118810589333648,
-      "learning_rate": 5.0001987297761256e-08,
-      "loss": 1.0114,
-      "step": 1440
-    },
-    {
-      "epoch": 3.0810234541577826,
-      "grad_norm": 1.153651371893197,
-      "learning_rate": 5.00017525294651e-08,
-      "loss": 1.1564,
-      "step": 1445
-    },
-    {
-      "epoch": 3.091684434968017,
-      "grad_norm": 1.117017616812222,
-      "learning_rate": 5.000154325080907e-08,
-      "loss": 1.0421,
-      "step": 1450
-    },
-    {
-      "epoch": 3.1023454157782515,
-      "grad_norm": 1.1455567386664758,
-      "learning_rate": 5.0001356943398416e-08,
-      "loss": 1.0494,
-      "step": 1455
-    },
-    {
-      "epoch": 3.113006396588486,
-      "grad_norm": 1.1745918907121704,
-      "learning_rate": 5.000119131249349e-08,
-      "loss": 1.0673,
-      "step": 1460
+      "epoch": 0.600127145581691,
+      "grad_norm": 1.8384043597752484,
+      "learning_rate": 1.2167168103937e-06,
+      "loss": 1.2493,
+      "step": 5664
     },
     {
-      "epoch": 3.1236673773987205,
-      "grad_norm": 1.0945498527723858,
-      "learning_rate": 5.000104426943277e-08,
-      "loss": 1.15,
-      "step": 1465
+      "epoch": 0.8001695274422547,
+      "grad_norm": 1.7725919716418708,
+      "learning_rate": 6.493264000872854e-07,
+      "loss": 1.2264,
+      "step": 7552
     },
     {
-      "epoch": 3.1343283582089554,
-      "grad_norm": 1.2775793960973583,
-      "learning_rate": 5.0000913915253465e-08,
-      "loss": 1.0434,
-      "step": 1470
+      "epoch": 0.8001695274422547,
+      "eval_loss": 1.214762568473816,
+      "eval_runtime": 327.7212,
+      "eval_samples_per_second": 204.79,
+      "eval_steps_per_second": 1.602,
+      "step": 7552
     },
     {
-      "epoch": 3.14498933901919,
-      "grad_norm": 1.1275573880105167,
-      "learning_rate": 5.000079852544101e-08,
-      "loss": 1.0726,
-      "step": 1475
+      "epoch": 1.0002119093028183,
+      "grad_norm": 1.8972637627842048,
+      "learning_rate": 3.4492111928598874e-07,
+      "loss": 1.211,
+      "step": 9440
     },
     {
-      "epoch": 3.1556503198294243,
-      "grad_norm": 1.0900224885696637,
-      "learning_rate": 5.0000696535741776e-08,
-      "loss": 1.0439,
-      "step": 1480
+      "epoch": 1.200254291163382,
+      "grad_norm": 1.8787635251660875,
+      "learning_rate": 1.8831885638765917e-07,
+      "loss": 1.1558,
+      "step": 11328
     },
     {
-      "epoch": 3.166311300639659,
-      "grad_norm": 1.1943472978175196,
-      "learning_rate": 5.0000606528976535e-08,
-      "loss": 1.0778,
-      "step": 1485
+      "epoch": 1.200254291163382,
+      "eval_loss": 1.2016733884811401,
+      "eval_runtime": 327.5932,
+      "eval_samples_per_second": 204.87,
+      "eval_steps_per_second": 1.603,
+      "step": 11328
     },
     {
-      "epoch": 3.1769722814498933,
-      "grad_norm": 1.1656854953422446,
-      "learning_rate": 5.000052722279511e-08,
-      "loss": 1.0236,
-      "step": 1490
+      "epoch": 1.4002966730239457,
+      "grad_norm": 2.0582903963596952,
+      "learning_rate": 1.1128443649930022e-07,
+      "loss": 1.1568,
+      "step": 13216
     },
     {
-      "epoch": 3.1876332622601278,
-      "grad_norm": 1.7099219299352593,
-      "learning_rate": 5.0000457458315606e-08,
-      "loss": 1.048,
-      "step": 1495
+      "epoch": 1.6003390548845093,
+      "grad_norm": 1.8665084344236207,
+      "learning_rate": 7.542962298572276e-08,
+      "loss": 1.1544,
+      "step": 15104
     },
     {
-      "epoch": 3.1982942430703627,
-      "grad_norm": 1.2368961315589513,
-      "learning_rate": 5.0000396189594195e-08,
-      "loss": 1.0601,
-      "step": 1500
+      "epoch": 1.6003390548845093,
+      "eval_loss": 1.1974143981933594,
+      "eval_runtime": 327.7112,
+      "eval_samples_per_second": 204.796,
+      "eval_steps_per_second": 1.602,
+      "step": 15104
     },
     {
-      "epoch": 3.208955223880597,
-      "grad_norm": 1.1053541732810632,
-      "learning_rate": 5.000034247387439e-08,
+      "epoch": 1.8003814367450732,
+      "grad_norm": 1.7528553673770804,
+      "learning_rate": 5.978512044954537e-08,
       "loss": 1.1574,
-      "step": 1505
-    },
-    {
-      "epoch": 3.2196162046908317,
-      "grad_norm": 1.2088122913217223,
-      "learning_rate": 5.0000295462567e-08,
-      "loss": 1.021,
-      "step": 1510
-    },
-    {
-      "epoch": 3.230277185501066,
-      "grad_norm": 1.1822420881980706,
-      "learning_rate": 5.000025439291453e-08,
-      "loss": 1.0577,
-      "step": 1515
-    },
-    {
-      "epoch": 3.2409381663113006,
-      "grad_norm": 1.1982389728811844,
-      "learning_rate": 5.0000218580296324e-08,
-      "loss": 1.1241,
-      "step": 1520
-    },
-    {
-      "epoch": 3.251599147121535,
-      "grad_norm": 1.1836468791922983,
-      "learning_rate": 5.0000187411132714e-08,
-      "loss": 1.106,
-      "step": 1525
-    },
-    {
-      "epoch": 3.2622601279317696,
-      "grad_norm": 1.1758919894146582,
-      "learning_rate": 5.0000160336348885e-08,
-      "loss": 1.2091,
-      "step": 1530
-    },
-    {
-      "epoch": 3.272921108742004,
-      "grad_norm": 1.1300862552793904,
-      "learning_rate": 5.0000136865361153e-08,
-      "loss": 1.0345,
-      "step": 1535
-    },
-    {
-      "epoch": 3.283582089552239,
-      "grad_norm": 1.1470416574016102,
-      "learning_rate": 5.000011656055037e-08,
-      "loss": 1.0357,
-      "step": 1540
-    },
-    {
-      "epoch": 3.2942430703624734,
-      "grad_norm": 1.18001188067462,
-      "learning_rate": 5.000009903218913e-08,
-      "loss": 1.0896,
-      "step": 1545
-    },
-    {
-      "epoch": 3.304904051172708,
-      "grad_norm": 1.1316402765789904,
-      "learning_rate": 5.0000083933791284e-08,
-      "loss": 1.0557,
-      "step": 1550
-    },
-    {
-      "epoch": 3.3155650319829424,
-      "grad_norm": 1.2685713000625365,
-      "learning_rate": 5.0000070957854027e-08,
-      "loss": 1.0442,
-      "step": 1555
-    },
-    {
-      "epoch": 3.326226012793177,
-      "grad_norm": 1.3173532606855904,
-      "learning_rate": 5.000005983196463e-08,
-      "loss": 1.1031,
-      "step": 1560
-    },
-    {
-      "epoch": 3.3368869936034113,
-      "grad_norm": 1.1645251930023506,
-      "learning_rate": 5.0000050315245216e-08,
-      "loss": 1.0327,
-      "step": 1565
-    },
-    {
-      "epoch": 3.3475479744136463,
-      "grad_norm": 1.1526305321387222,
-      "learning_rate": 5.000004219511104e-08,
-      "loss": 1.1047,
-      "step": 1570
-    },
-    {
-      "epoch": 3.3582089552238807,
-      "grad_norm": 1.1956529776682534,
-      "learning_rate": 5.0000035284318585e-08,
-      "loss": 1.0513,
-      "step": 1575
-    },
-    {
-      "epoch": 3.368869936034115,
-      "grad_norm": 1.1723882568539148,
-      "learning_rate": 5.00000294182818e-08,
-      "loss": 1.19,
-      "step": 1580
-    },
-    {
-      "epoch": 3.3795309168443497,
-      "grad_norm": 1.1986098281392215,
-      "learning_rate": 5.00000244526357e-08,
-      "loss": 1.077,
-      "step": 1585
-    },
-    {
-      "epoch": 3.390191897654584,
-      "grad_norm": 1.1822052046386484,
-      "learning_rate": 5.0000020261028096e-08,
-      "loss": 0.9985,
-      "step": 1590
-    },
-    {
-      "epoch": 3.4008528784648187,
-      "grad_norm": 1.1965544250131317,
-      "learning_rate": 5.0000016733121405e-08,
-      "loss": 1.1462,
-      "step": 1595
-    },
-    {
-      "epoch": 3.411513859275053,
-      "grad_norm": 1.233729124712432,
-      "learning_rate": 5.00000137727875e-08,
-      "loss": 1.013,
-      "step": 1600
-    },
-    {
-      "epoch": 3.411513859275053,
-      "eval_loss": 1.2190721035003662,
-      "eval_runtime": 86.8673,
-      "eval_samples_per_second": 19.19,
-      "eval_steps_per_second": 0.311,
-      "step": 1600
-    },
-    {
-      "epoch": 3.4221748400852876,
-      "grad_norm": 1.1940390976157107,
-      "learning_rate": 5.00000112964799e-08,
-      "loss": 1.0902,
-      "step": 1605
-    },
-    {
-      "epoch": 3.4328358208955225,
-      "grad_norm": 1.1852233751297125,
-      "learning_rate": 5.000000923176855e-08,
-      "loss": 1.0622,
-      "step": 1610
-    },
-    {
-      "epoch": 3.443496801705757,
-      "grad_norm": 1.174182970082022,
-      "learning_rate": 5.0000007516023255e-08,
-      "loss": 1.1077,
-      "step": 1615
-    },
-    {
-      "epoch": 3.4541577825159915,
-      "grad_norm": 1.2175665950720118,
-      "learning_rate": 5.0000006095233244e-08,
-      "loss": 1.0463,
-      "step": 1620
-    },
-    {
-      "epoch": 3.464818763326226,
-      "grad_norm": 1.1910402507833668,
-      "learning_rate": 5.0000004922950696e-08,
-      "loss": 1.0485,
-      "step": 1625
-    },
-    {
-      "epoch": 3.4754797441364604,
-      "grad_norm": 1.1756720560395666,
-      "learning_rate": 5.000000395934729e-08,
-      "loss": 1.0855,
-      "step": 1630
-    },
-    {
-      "epoch": 3.486140724946695,
-      "grad_norm": 1.1357035763961498,
-      "learning_rate": 5.0000003170373506e-08,
-      "loss": 1.0552,
-      "step": 1635
-    },
-    {
-      "epoch": 3.49680170575693,
-      "grad_norm": 1.12995191245614,
-      "learning_rate": 5.0000002527011166e-08,
-      "loss": 1.1139,
-      "step": 1640
-    },
-    {
-      "epoch": 3.5074626865671643,
-      "grad_norm": 1.2169670051086863,
-      "learning_rate": 5.000000200461043e-08,
-      "loss": 1.0746,
-      "step": 1645
-    },
-    {
-      "epoch": 3.518123667377399,
-      "grad_norm": 1.1477244123007122,
-      "learning_rate": 5.000000158230317e-08,
-      "loss": 1.0328,
-      "step": 1650
-    },
-    {
-      "epoch": 3.5287846481876333,
-      "grad_norm": 1.146160022657433,
-      "learning_rate": 5.000000124248513e-08,
-      "loss": 1.0818,
-      "step": 1655
-    },
-    {
-      "epoch": 3.5394456289978677,
-      "grad_norm": 1.1725432339751227,
-      "learning_rate": 5.0000000970360234e-08,
-      "loss": 1.0631,
-      "step": 1660
-    },
-    {
-      "epoch": 3.550106609808102,
-      "grad_norm": 1.1437856255430832,
-      "learning_rate": 5.000000075354043e-08,
-      "loss": 1.03,
-      "step": 1665
-    },
-    {
-      "epoch": 3.5607675906183367,
-      "grad_norm": 1.1949291330530631,
-      "learning_rate": 5.000000058169555e-08,
-      "loss": 1.0766,
-      "step": 1670
-    },
-    {
-      "epoch": 3.571428571428571,
-      "grad_norm": 1.1521311383486394,
-      "learning_rate": 5.000000044624768e-08,
-      "loss": 1.1008,
-      "step": 1675
-    },
-    {
-      "epoch": 3.582089552238806,
-      "grad_norm": 1.1470924970059302,
-      "learning_rate": 5.000000034010534e-08,
-      "loss": 1.0803,
-      "step": 1680
-    },
-    {
-      "epoch": 3.5927505330490406,
-      "grad_norm": 1.163957344487391,
-      "learning_rate": 5.0000000257433004e-08,
-      "loss": 1.0318,
-      "step": 1685
-    },
-    {
-      "epoch": 3.603411513859275,
-      "grad_norm": 1.1289250902982475,
-      "learning_rate": 5.000000019345195e-08,
-      "loss": 1.0714,
-      "step": 1690
-    },
-    {
-      "epoch": 3.6140724946695095,
-      "grad_norm": 1.1396900843621998,
-      "learning_rate": 5.0000000144268826e-08,
-      "loss": 1.0785,
-      "step": 1695
-    },
-    {
-      "epoch": 3.624733475479744,
-      "grad_norm": 1.1928003735738484,
-      "learning_rate": 5.000000010672866e-08,
-      "loss": 1.0476,
-      "step": 1700
-    },
-    {
-      "epoch": 3.635394456289979,
-      "grad_norm": 1.1813297878533093,
-      "learning_rate": 5.000000007828923e-08,
-      "loss": 1.1059,
-      "step": 1705
-    },
-    {
-      "epoch": 3.6460554371002134,
-      "grad_norm": 1.1580070228375894,
-      "learning_rate": 5.000000005691433e-08,
-      "loss": 1.0804,
-      "step": 1710
-    },
-    {
-      "epoch": 3.656716417910448,
-      "grad_norm": 1.1877112723964363,
-      "learning_rate": 5.0000000040983296e-08,
-      "loss": 1.0926,
-      "step": 1715
-    },
-    {
-      "epoch": 3.6673773987206824,
-      "grad_norm": 1.162988885771696,
-      "learning_rate": 5.0000000029214826e-08,
-      "loss": 1.134,
-      "step": 1720
-    },
-    {
-      "epoch": 3.678038379530917,
-      "grad_norm": 1.2167801349477705,
-      "learning_rate": 5.000000002060308e-08,
-      "loss": 1.1055,
-      "step": 1725
-    },
-    {
-      "epoch": 3.6886993603411513,
-      "grad_norm": 1.1458862089420816,
-      "learning_rate": 5.000000001436439e-08,
-      "loss": 1.0211,
-      "step": 1730
-    },
-    {
-      "epoch": 3.699360341151386,
-      "grad_norm": 1.0873296281789822,
-      "learning_rate": 5.000000000989308e-08,
-      "loss": 1.0072,
-      "step": 1735
-    },
-    {
-      "epoch": 3.7100213219616203,
-      "grad_norm": 1.1575880345419727,
-      "learning_rate": 5.000000000672499e-08,
-      "loss": 1.1223,
-      "step": 1740
-    },
-    {
-      "epoch": 3.7206823027718547,
-      "grad_norm": 1.1887408531399297,
-      "learning_rate": 5.0000000004507695e-08,
-      "loss": 1.0693,
-      "step": 1745
-    },
-    {
-      "epoch": 3.7313432835820897,
-      "grad_norm": 1.1370169892656834,
-      "learning_rate": 5.000000000297619e-08,
-      "loss": 0.9994,
-      "step": 1750
-    },
-    {
-      "epoch": 3.742004264392324,
-      "grad_norm": 1.1964357258341392,
-      "learning_rate": 5.000000000193327e-08,
-      "loss": 1.1289,
-      "step": 1755
-    },
-    {
-      "epoch": 3.7526652452025586,
-      "grad_norm": 1.1471490215463733,
-      "learning_rate": 5.000000000123389e-08,
-      "loss": 1.0828,
-      "step": 1760
-    },
-    {
-      "epoch": 3.763326226012793,
-      "grad_norm": 1.1835965960152477,
-      "learning_rate": 5.0000000000772604e-08,
-      "loss": 1.1045,
-      "step": 1765
-    },
-    {
-      "epoch": 3.7739872068230276,
-      "grad_norm": 1.122980946834769,
-      "learning_rate": 5.000000000047381e-08,
-      "loss": 1.0586,
-      "step": 1770
-    },
-    {
-      "epoch": 3.7846481876332625,
-      "grad_norm": 1.1509727983825737,
-      "learning_rate": 5.000000000028403e-08,
-      "loss": 1.1015,
-      "step": 1775
-    },
-    {
-      "epoch": 3.795309168443497,
-      "grad_norm": 1.1806503665236887,
-      "learning_rate": 5.000000000016606e-08,
-      "loss": 1.1195,
-      "step": 1780
-    },
-    {
-      "epoch": 3.8059701492537314,
-      "grad_norm": 1.1840945960267957,
-      "learning_rate": 5.0000000000094455e-08,
-      "loss": 1.0452,
-      "step": 1785
-    },
-    {
-      "epoch": 3.816631130063966,
-      "grad_norm": 1.1373535444813456,
-      "learning_rate": 5.000000000005211e-08,
-      "loss": 1.0538,
-      "step": 1790
-    },
-    {
-      "epoch": 3.8272921108742004,
-      "grad_norm": 1.1164448190518461,
-      "learning_rate": 5.000000000002779e-08,
-      "loss": 1.0796,
-      "step": 1795
-    },
-    {
-      "epoch": 3.837953091684435,
-      "grad_norm": 1.2618135197926215,
-      "learning_rate": 5.0000000000014265e-08,
-      "loss": 1.1103,
-      "step": 1800
-    },
-    {
-      "epoch": 3.837953091684435,
-      "eval_loss": 1.2187809944152832,
-      "eval_runtime": 87.3426,
-      "eval_samples_per_second": 19.086,
-      "eval_steps_per_second": 0.309,
-      "step": 1800
+      "step": 16992
     },
     {
-      "epoch": 3.8486140724946694,
-      "grad_norm": 1.166670525467647,
-      "learning_rate": 5.000000000000701e-08,
-      "loss": 1.0504,
-      "step": 1805
+      "epoch": 2.0004238186056367,
+      "grad_norm": 1.8861106343782263,
+      "learning_rate": 5.343416604131081e-08,
+      "loss": 1.1562,
+      "step": 18880
     },
     {
-      "epoch": 3.859275053304904,
-      "grad_norm": 1.1358229871637773,
-      "learning_rate": 5.000000000000328e-08,
-      "loss": 1.031,
-      "step": 1810
+      "epoch": 2.0004238186056367,
+      "eval_loss": 1.19540274143219,
+      "eval_runtime": 328.044,
+      "eval_samples_per_second": 204.588,
+      "eval_steps_per_second": 1.6,
+      "step": 18880
     },
     {
-      "epoch": 3.8699360341151388,
-      "grad_norm": 1.1553072990278574,
-      "learning_rate": 5.0000000000001454e-08,
-      "loss": 1.0852,
-      "step": 1815
+      "epoch": 2.2004662004662006,
+      "grad_norm": 1.9003730335704914,
+      "learning_rate": 5.1080120038479373e-08,
+      "loss": 1.1431,
+      "step": 20768
     },
     {
-      "epoch": 3.8805970149253732,
-      "grad_norm": 1.1514532203092147,
-      "learning_rate": 5.00000000000006e-08,
-      "loss": 1.0682,
-      "step": 1820
+      "epoch": 2.400508582326764,
+      "grad_norm": 1.941099899013768,
+      "learning_rate": 5.029649877564949e-08,
+      "loss": 1.1416,
+      "step": 22656
     },
     {
-      "epoch": 3.8912579957356077,
-      "grad_norm": 1.1884972946614185,
-      "learning_rate": 5.000000000000023e-08,
-      "loss": 1.0012,
-      "step": 1825
+      "epoch": 2.400508582326764,
+      "eval_loss": 1.1947613954544067,
+      "eval_runtime": 327.5948,
+      "eval_samples_per_second": 204.869,
+      "eval_steps_per_second": 1.603,
+      "step": 22656
     },
     {
-      "epoch": 3.901918976545842,
-      "grad_norm": 1.1704188735844283,
-      "learning_rate": 5.000000000000008e-08,
-      "loss": 1.0772,
-      "step": 1830
+      "epoch": 2.600550964187328,
+      "grad_norm": 1.9472750123018565,
+      "learning_rate": 5.0068500018386117e-08,
+      "loss": 1.1461,
+      "step": 24544
     },
     {
-      "epoch": 3.9125799573560767,
-      "grad_norm": 1.128126951495901,
-      "learning_rate": 5.0000000000000024e-08,
-      "loss": 1.0497,
-      "step": 1835
+      "epoch": 2.8005933460478913,
+      "grad_norm": 1.893076803756726,
+      "learning_rate": 5.0012641588155837e-08,
+      "loss": 1.1449,
+      "step": 26432
     },
     {
-      "epoch": 3.923240938166311,
-      "grad_norm": 1.14600117506389,
-      "learning_rate": 5.0000000000000004e-08,
-      "loss": 1.0813,
-      "step": 1840
+      "epoch": 2.8005933460478913,
+      "eval_loss": 1.1934314966201782,
+      "eval_runtime": 328.7941,
+      "eval_samples_per_second": 204.122,
+      "eval_steps_per_second": 1.597,
+      "step": 26432
     },
     {
-      "epoch": 3.933901918976546,
-      "grad_norm": 1.1304423202109097,
-      "learning_rate": 5e-08,
-      "loss": 1.1111,
-      "step": 1845
+      "epoch": 3.0006357279084552,
+      "grad_norm": 2.082824554026489,
+      "learning_rate": 5.0001709336153935e-08,
+      "loss": 1.1464,
+      "step": 28320
     },
     {
-      "epoch": 3.9445628997867805,
-      "grad_norm": 1.127824200062695,
-      "learning_rate": 5e-08,
-      "loss": 1.0618,
-      "step": 1850
+      "epoch": 3.2006781097690187,
+      "grad_norm": 1.9988580640926128,
+      "learning_rate": 5.000014828801133e-08,
+      "loss": 1.1401,
+      "step": 30208
     },
     {
-      "epoch": 3.955223880597015,
-      "grad_norm": 1.1479631723327013,
-      "learning_rate": 5e-08,
-      "loss": 1.0155,
-      "step": 1855
+      "epoch": 3.2006781097690187,
+      "eval_loss": 1.1925288438796997,
+      "eval_runtime": 327.8373,
+      "eval_samples_per_second": 204.717,
+      "eval_steps_per_second": 1.601,
+      "step": 30208
     },
     {
-      "epoch": 3.9658848614072495,
-      "grad_norm": 1.2107343511697226,
-      "learning_rate": 5e-08,
-      "loss": 1.1179,
-      "step": 1860
+      "epoch": 3.4007204916295826,
+      "grad_norm": 1.8907403603158888,
+      "learning_rate": 5.0000006391422415e-08,
+      "loss": 1.1397,
+      "step": 32096
     },
     {
-      "epoch": 3.976545842217484,
-      "grad_norm": 1.1701981970918207,
-      "learning_rate": 5e-08,
-      "loss": 1.0525,
-      "step": 1865
+      "epoch": 3.6007628734901465,
+      "grad_norm": 1.9407599933229671,
+      "learning_rate": 5.0000000076061e-08,
+      "loss": 1.1382,
+      "step": 33984
     },
     {
-      "epoch": 3.9872068230277184,
-      "grad_norm": 1.2066704843958889,
-      "learning_rate": 5e-08,
-      "loss": 1.0012,
-      "step": 1870
+      "epoch": 3.6007628734901465,
+      "eval_loss": 1.191159963607788,
+      "eval_runtime": 327.6723,
+      "eval_samples_per_second": 204.82,
+      "eval_steps_per_second": 1.602,
+      "step": 33984
     },
     {
-      "epoch": 3.997867803837953,
-      "grad_norm": 1.1997617373616742,
-      "learning_rate": 5e-08,
-      "loss": 1.0131,
-      "step": 1875
+      "epoch": 3.80080525535071,
+      "grad_norm": 1.9516245761095283,
+      "learning_rate": 5.000000000004096e-08,
+      "loss": 1.1404,
+      "step": 35872
     },
     {
       "epoch": 4.0,
-      "step": 1876,
-      "total_flos": 391958715432960.0,
-      "train_loss": 1.1227505109203395,
-      "train_runtime": 12501.281,
-      "train_samples_per_second": 4.798,
-      "train_steps_per_second": 0.15
+      "step": 37752,
+      "total_flos": 3947644028190720.0,
+      "train_loss": 1.1804671342891009,
+      "train_runtime": 53865.2221,
+      "train_samples_per_second": 44.854,
+      "train_steps_per_second": 0.701
     }
   ],
-  "logging_steps": 5,
-  "max_steps": 1876,
+  "logging_steps": 1888,
+  "max_steps": 37752,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 4,
-  "save_steps": 200,
-  "total_flos": 391958715432960.0,
-  "train_batch_size": 4,
+  "save_steps": 3776,
+  "total_flos": 3947644028190720.0,
+  "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null
 }