{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9995747929973436,
  "eval_steps": 500,
  "global_step": 1708,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0005852311434410677,
      "grad_norm": 0.515625,
      "learning_rate": 1.1695906432748538e-06,
      "loss": 1.737,
      "step": 1
    },
    {
      "epoch": 0.0029261557172053382,
      "grad_norm": 0.5234375,
      "learning_rate": 5.8479532163742686e-06,
      "loss": 1.7512,
      "step": 5
    },
    {
      "epoch": 0.0058523114344106765,
      "grad_norm": 0.48046875,
      "learning_rate": 1.1695906432748537e-05,
      "loss": 1.7378,
      "step": 10
    },
    {
      "epoch": 0.008778467151616015,
      "grad_norm": 0.474609375,
      "learning_rate": 1.7543859649122806e-05,
      "loss": 1.7494,
      "step": 15
    },
    {
      "epoch": 0.011704622868821353,
      "grad_norm": 0.4765625,
      "learning_rate": 2.3391812865497074e-05,
      "loss": 1.7101,
      "step": 20
    },
    {
      "epoch": 0.014630778586026691,
      "grad_norm": 0.482421875,
      "learning_rate": 2.9239766081871346e-05,
      "loss": 1.6696,
      "step": 25
    },
    {
      "epoch": 0.01755693430323203,
      "grad_norm": 0.34765625,
      "learning_rate": 3.508771929824561e-05,
      "loss": 1.6369,
      "step": 30
    },
    {
      "epoch": 0.02048309002043737,
      "grad_norm": 0.287109375,
      "learning_rate": 4.093567251461988e-05,
      "loss": 1.6058,
      "step": 35
    },
    {
      "epoch": 0.023409245737642706,
      "grad_norm": 0.2333984375,
      "learning_rate": 4.678362573099415e-05,
      "loss": 1.5587,
      "step": 40
    },
    {
      "epoch": 0.026335401454848046,
      "grad_norm": 0.1708984375,
      "learning_rate": 5.2631578947368424e-05,
      "loss": 1.5149,
      "step": 45
    },
    {
      "epoch": 0.029261557172053382,
      "grad_norm": 0.140625,
      "learning_rate": 5.847953216374269e-05,
      "loss": 1.4769,
      "step": 50
    },
    {
      "epoch": 0.03218771288925872,
      "grad_norm": 0.12060546875,
      "learning_rate": 6.432748538011695e-05,
      "loss": 1.4573,
      "step": 55
    },
    {
      "epoch": 0.03511386860646406,
      "grad_norm": 0.09228515625,
      "learning_rate": 7.017543859649122e-05,
      "loss": 1.4219,
      "step": 60
    },
    {
      "epoch": 0.0380400243236694,
      "grad_norm": 0.09375,
      "learning_rate": 7.602339181286549e-05,
      "loss": 1.398,
      "step": 65
    },
    {
      "epoch": 0.04096618004087474,
      "grad_norm": 0.06884765625,
      "learning_rate": 8.187134502923976e-05,
      "loss": 1.3912,
      "step": 70
    },
    {
      "epoch": 0.043892335758080075,
      "grad_norm": 0.06787109375,
      "learning_rate": 8.771929824561403e-05,
      "loss": 1.3633,
      "step": 75
    },
    {
      "epoch": 0.04681849147528541,
      "grad_norm": 0.06103515625,
      "learning_rate": 9.35672514619883e-05,
      "loss": 1.3546,
      "step": 80
    },
    {
      "epoch": 0.049744647192490755,
      "grad_norm": 0.05712890625,
      "learning_rate": 9.941520467836257e-05,
      "loss": 1.3309,
      "step": 85
    },
    {
      "epoch": 0.05267080290969609,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.00010526315789473685,
      "loss": 1.3234,
      "step": 90
    },
    {
      "epoch": 0.05559695862690143,
      "grad_norm": 0.048828125,
      "learning_rate": 0.00011111111111111112,
      "loss": 1.3061,
      "step": 95
    },
    {
      "epoch": 0.058523114344106765,
      "grad_norm": 0.048828125,
      "learning_rate": 0.00011695906432748539,
      "loss": 1.2844,
      "step": 100
    },
    {
      "epoch": 0.06144927006131211,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.00012280701754385965,
      "loss": 1.2973,
      "step": 105
    },
    {
      "epoch": 0.06437542577851744,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0001286549707602339,
      "loss": 1.2836,
      "step": 110
    },
    {
      "epoch": 0.06730158149572278,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0001345029239766082,
      "loss": 1.2723,
      "step": 115
    },
    {
      "epoch": 0.07022773721292812,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.00014035087719298245,
      "loss": 1.2634,
      "step": 120
    },
    {
      "epoch": 0.07315389293013345,
      "grad_norm": 0.05078125,
      "learning_rate": 0.00014619883040935673,
      "loss": 1.2355,
      "step": 125
    },
    {
      "epoch": 0.0760800486473388,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00015204678362573098,
      "loss": 1.2494,
      "step": 130
    },
    {
      "epoch": 0.07900620436454414,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.00015789473684210527,
      "loss": 1.253,
      "step": 135
    },
    {
      "epoch": 0.08193236008174948,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.00016374269005847952,
      "loss": 1.2499,
      "step": 140
    },
    {
      "epoch": 0.08485851579895481,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0001695906432748538,
      "loss": 1.2193,
      "step": 145
    },
    {
      "epoch": 0.08778467151616015,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.00017543859649122806,
      "loss": 1.2339,
      "step": 150
    },
    {
      "epoch": 0.09071082723336549,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.00018128654970760234,
      "loss": 1.2427,
      "step": 155
    },
    {
      "epoch": 0.09363698295057082,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0001871345029239766,
      "loss": 1.2184,
      "step": 160
    },
    {
      "epoch": 0.09656313866777616,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.00019298245614035088,
      "loss": 1.2306,
      "step": 165
    },
    {
      "epoch": 0.09948929438498151,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.00019883040935672513,
      "loss": 1.214,
      "step": 170
    },
    {
      "epoch": 0.10241545010218685,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.00019999665774502696,
      "loss": 1.2176,
      "step": 175
    },
    {
      "epoch": 0.10534160581939218,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0001999830802170989,
      "loss": 1.204,
      "step": 180
    },
    {
      "epoch": 0.10826776153659752,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.00019995905994229593,
      "loss": 1.2153,
      "step": 185
    },
    {
      "epoch": 0.11119391725380286,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.00019992459942941906,
      "loss": 1.1936,
      "step": 190
    },
    {
      "epoch": 0.1141200729710082,
      "grad_norm": 0.064453125,
      "learning_rate": 0.00019987970227770135,
      "loss": 1.1987,
      "step": 195
    },
    {
      "epoch": 0.11704622868821353,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.00019982437317643217,
      "loss": 1.2065,
      "step": 200
    },
    {
      "epoch": 0.11997238440541888,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.00019975861790446722,
      "loss": 1.2088,
      "step": 205
    },
    {
      "epoch": 0.12289854012262422,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.0001996824433296252,
      "loss": 1.2082,
      "step": 210
    },
    {
      "epoch": 0.12582469583982955,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.00019959585740797028,
      "loss": 1.2062,
      "step": 215
    },
    {
      "epoch": 0.1287508515570349,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0001994988691829812,
      "loss": 1.2046,
      "step": 220
    },
    {
      "epoch": 0.13167700727424023,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.00019939148878460677,
      "loss": 1.195,
      "step": 225
    },
    {
      "epoch": 0.13460316299144556,
      "grad_norm": 0.0859375,
      "learning_rate": 0.00019927372742820779,
      "loss": 1.1807,
      "step": 230
    },
    {
      "epoch": 0.1375293187086509,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0001991455974133857,
      "loss": 1.1887,
      "step": 235
    },
    {
      "epoch": 0.14045547442585624,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.0001990071121226979,
      "loss": 1.189,
      "step": 240
    },
    {
      "epoch": 0.14338163014306157,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0001988582860202601,
      "loss": 1.172,
      "step": 245
    },
    {
      "epoch": 0.1463077858602669,
      "grad_norm": 0.078125,
      "learning_rate": 0.00019869913465023548,
      "loss": 1.1738,
      "step": 250
    },
    {
      "epoch": 0.14923394157747225,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.00019852967463521124,
      "loss": 1.1947,
      "step": 255
    },
    {
      "epoch": 0.1521600972946776,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0001983499236744625,
      "loss": 1.1789,
      "step": 260
    },
    {
      "epoch": 0.15508625301188295,
      "grad_norm": 0.078125,
      "learning_rate": 0.00019815990054210361,
      "loss": 1.1878,
      "step": 265
    },
    {
      "epoch": 0.15801240872908828,
      "grad_norm": 0.072265625,
      "learning_rate": 0.00019795962508512742,
      "loss": 1.1825,
      "step": 270
    },
    {
      "epoch": 0.16093856444629362,
      "grad_norm": 0.103515625,
      "learning_rate": 0.00019774911822133216,
      "loss": 1.1848,
      "step": 275
    },
    {
      "epoch": 0.16386472016349896,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001975284019371368,
      "loss": 1.1634,
      "step": 280
    },
    {
      "epoch": 0.1667908758807043,
      "grad_norm": 0.078125,
      "learning_rate": 0.0001972974992852847,
      "loss": 1.1539,
      "step": 285
    },
    {
      "epoch": 0.16971703159790963,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00019705643438243584,
      "loss": 1.1656,
      "step": 290
    },
    {
      "epoch": 0.17264318731511497,
      "grad_norm": 0.0859375,
      "learning_rate": 0.00019680523240664786,
      "loss": 1.1923,
      "step": 295
    },
    {
      "epoch": 0.1755693430323203,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.00019654391959474647,
      "loss": 1.1651,
      "step": 300
    },
    {
      "epoch": 0.17849549874952564,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.00019627252323958504,
      "loss": 1.1501,
      "step": 305
    },
    {
      "epoch": 0.18142165446673097,
      "grad_norm": 0.0859375,
      "learning_rate": 0.00019599107168719412,
      "loss": 1.1581,
      "step": 310
    },
    {
      "epoch": 0.1843478101839363,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001956995943338206,
      "loss": 1.1785,
      "step": 315
    },
    {
      "epoch": 0.18727396590114165,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.00019539812162285767,
      "loss": 1.1691,
      "step": 320
    },
    {
      "epoch": 0.19020012161834698,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.00019508668504166505,
      "loss": 1.1758,
      "step": 325
    },
    {
      "epoch": 0.19312627733555232,
      "grad_norm": 0.0703125,
      "learning_rate": 0.00019476531711828027,
      "loss": 1.1582,
      "step": 330
    },
    {
      "epoch": 0.19605243305275769,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001944340514180212,
      "loss": 1.1767,
      "step": 335
    },
    {
      "epoch": 0.19897858876996302,
      "grad_norm": 0.0703125,
      "learning_rate": 0.00019409292253998062,
      "loss": 1.1392,
      "step": 340
    },
    {
      "epoch": 0.20190474448716836,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.0001937419661134121,
      "loss": 1.1626,
      "step": 345
    },
    {
      "epoch": 0.2048309002043737,
      "grad_norm": 0.08203125,
      "learning_rate": 0.00019338121879400896,
      "loss": 1.1551,
      "step": 350
    },
    {
      "epoch": 0.20775705592157903,
      "grad_norm": 0.080078125,
      "learning_rate": 0.00019301071826007576,
      "loss": 1.1495,
      "step": 355
    },
    {
      "epoch": 0.21068321163878437,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.00019263050320859283,
      "loss": 1.1514,
      "step": 360
    },
    {
      "epoch": 0.2136093673559897,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.00019224061335117472,
      "loss": 1.1649,
      "step": 365
    },
    {
      "epoch": 0.21653552307319504,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0001918410894099224,
      "loss": 1.1433,
      "step": 370
    },
    {
      "epoch": 0.21946167879040038,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.00019143197311317014,
      "loss": 1.1275,
      "step": 375
    },
    {
      "epoch": 0.22238783450760571,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.00019101330719112705,
      "loss": 1.1684,
      "step": 380
    },
    {
      "epoch": 0.22531399022481105,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.00019058513537141428,
      "loss": 1.1606,
      "step": 385
    },
    {
      "epoch": 0.2282401459420164,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0001901475023744977,
      "loss": 1.148,
      "step": 390
    },
    {
      "epoch": 0.23116630165922172,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.00018970045390901728,
      "loss": 1.1626,
      "step": 395
    },
    {
      "epoch": 0.23409245737642706,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.00018924403666701286,
      "loss": 1.1575,
      "step": 400
    },
    {
      "epoch": 0.23701861309363242,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.00018877829831904746,
      "loss": 1.1637,
      "step": 405
    },
    {
      "epoch": 0.23994476881083776,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0001883032875092283,
      "loss": 1.1441,
      "step": 410
    },
    {
      "epoch": 0.2428709245280431,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.00018781905385012627,
      "loss": 1.1615,
      "step": 415
    },
    {
      "epoch": 0.24579708024524843,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.000187325647917594,
      "loss": 1.1536,
      "step": 420
    },
    {
      "epoch": 0.24872323596245377,
      "grad_norm": 0.08203125,
      "learning_rate": 0.00018682312124548346,
      "loss": 1.1512,
      "step": 425
    },
    {
      "epoch": 0.2516493916796591,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.00018631152632026364,
      "loss": 1.1397,
      "step": 430
    },
    {
      "epoch": 0.25457554739686444,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.00018579091657553844,
      "loss": 1.1585,
      "step": 435
    },
    {
      "epoch": 0.2575017031140698,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.00018526134638646583,
      "loss": 1.1612,
      "step": 440
    },
    {
      "epoch": 0.2604278588312751,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.00018472287106407876,
      "loss": 1.1272,
      "step": 445
    },
    {
      "epoch": 0.26335401454848045,
      "grad_norm": 0.09375,
      "learning_rate": 0.00018417554684950794,
      "loss": 1.1413,
      "step": 450
    },
    {
      "epoch": 0.2662801702656858,
      "grad_norm": 0.0703125,
      "learning_rate": 0.00018361943090810796,
      "loss": 1.1489,
      "step": 455
    },
    {
      "epoch": 0.2692063259828911,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.00018305458132348657,
      "loss": 1.1575,
      "step": 460
    },
    {
      "epoch": 0.27213248170009646,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.00018248105709143799,
      "loss": 1.136,
      "step": 465
    },
    {
      "epoch": 0.2750586374173018,
      "grad_norm": 0.08203125,
      "learning_rate": 0.00018189891811378137,
      "loss": 1.1369,
      "step": 470
    },
    {
      "epoch": 0.27798479313450714,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001813082251921041,
      "loss": 1.1255,
      "step": 475
    },
    {
      "epoch": 0.28091094885171247,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001807090400214114,
      "loss": 1.1288,
      "step": 480
    },
    {
      "epoch": 0.2838371045689178,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.00018010142518368278,
      "loss": 1.1233,
      "step": 485
    },
    {
      "epoch": 0.28676326028612315,
      "grad_norm": 0.078125,
      "learning_rate": 0.00017948544414133534,
      "loss": 1.1475,
      "step": 490
    },
    {
      "epoch": 0.2896894160033285,
      "grad_norm": 0.080078125,
      "learning_rate": 0.00017886116123059574,
      "loss": 1.1356,
      "step": 495
    },
    {
      "epoch": 0.2926155717205338,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.00017822864165478034,
      "loss": 1.1553,
      "step": 500
    },
    {
      "epoch": 0.29554172743773915,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.00017758795147748523,
      "loss": 1.1188,
      "step": 505
    },
    {
      "epoch": 0.2984678831549445,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.00017693915761568608,
      "loss": 1.1388,
      "step": 510
    },
    {
      "epoch": 0.3013940388721499,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.000176282327832749,
      "loss": 1.1267,
      "step": 515
    },
    {
      "epoch": 0.3043201945893552,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001756175307313531,
      "loss": 1.1341,
      "step": 520
    },
    {
      "epoch": 0.30724635030656056,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.00017494483574632513,
      "loss": 1.1365,
      "step": 525
    },
    {
      "epoch": 0.3101725060237659,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.00017426431313738734,
      "loss": 1.1335,
      "step": 530
    },
    {
      "epoch": 0.31309866174097123,
      "grad_norm": 0.0703125,
      "learning_rate": 0.00017357603398181936,
      "loss": 1.1484,
      "step": 535
    },
    {
      "epoch": 0.31602481745817657,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.00017288007016703444,
      "loss": 1.1186,
      "step": 540
    },
    {
      "epoch": 0.3189509731753819,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00017217649438307106,
      "loss": 1.1442,
      "step": 545
    },
    {
      "epoch": 0.32187712889258724,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.00017146538011500093,
      "loss": 1.1284,
      "step": 550
    },
    {
      "epoch": 0.3248032846097926,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.00017074680163525375,
      "loss": 1.1331,
      "step": 555
    },
    {
      "epoch": 0.3277294403269979,
      "grad_norm": 0.07421875,
      "learning_rate": 0.00017002083399586,
      "loss": 1.1255,
      "step": 560
    },
    {
      "epoch": 0.33065559604420325,
      "grad_norm": 0.07421875,
      "learning_rate": 0.00016928755302061173,
      "loss": 1.1354,
      "step": 565
    },
    {
      "epoch": 0.3335817517614086,
      "grad_norm": 0.078125,
      "learning_rate": 0.0001685470352971437,
      "loss": 1.1333,
      "step": 570
    },
    {
      "epoch": 0.3365079074786139,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.00016779935816893353,
      "loss": 1.1376,
      "step": 575
    },
    {
      "epoch": 0.33943406319581926,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00016704459972722414,
      "loss": 1.1249,
      "step": 580
    },
    {
      "epoch": 0.3423602189130246,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.00016628283880286703,
      "loss": 1.1451,
      "step": 585
    },
    {
      "epoch": 0.34528637463022993,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.00016551415495808915,
      "loss": 1.1195,
      "step": 590
    },
    {
      "epoch": 0.34821253034743527,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.00016473862847818277,
      "loss": 1.146,
      "step": 595
    },
    {
      "epoch": 0.3511386860646406,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.00016395634036312013,
      "loss": 1.1327,
      "step": 600
    },
    {
      "epoch": 0.35406484178184594,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.00016316737231909342,
      "loss": 1.1176,
      "step": 605
    },
    {
      "epoch": 0.3569909974990513,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.000162371806749981,
      "loss": 1.1208,
      "step": 610
    },
    {
      "epoch": 0.3599171532162566,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.00016156972674874056,
      "loss": 1.1315,
      "step": 615
    },
    {
      "epoch": 0.36284330893346195,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00016076121608873072,
      "loss": 1.1455,
      "step": 620
    },
    {
      "epoch": 0.3657694646506673,
      "grad_norm": 0.072265625,
      "learning_rate": 0.000159946359214961,
      "loss": 1.1234,
      "step": 625
    },
    {
      "epoch": 0.3686956203678726,
      "grad_norm": 0.07421875,
      "learning_rate": 0.00015912524123527221,
      "loss": 1.1185,
      "step": 630
    },
    {
      "epoch": 0.37162177608507796,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0001582979479114472,
      "loss": 1.1208,
      "step": 635
    },
    {
      "epoch": 0.3745479318022833,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0001574645656502536,
      "loss": 1.1257,
      "step": 640
    },
    {
      "epoch": 0.37747408751948863,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0001566251814944188,
      "loss": 1.1317,
      "step": 645
    },
    {
      "epoch": 0.38040024323669397,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.00015577988311353904,
      "loss": 1.1431,
      "step": 650
    },
    {
      "epoch": 0.3833263989538993,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0001549287587949226,
      "loss": 1.1253,
      "step": 655
    },
    {
      "epoch": 0.38625255467110464,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.00015407189743436864,
      "loss": 1.1314,
      "step": 660
    },
    {
      "epoch": 0.38917871038831003,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.00015320938852688248,
      "loss": 1.1148,
      "step": 665
    },
    {
      "epoch": 0.39210486610551537,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.00015234132215732822,
      "loss": 1.141,
      "step": 670
    },
    {
      "epoch": 0.3950310218227207,
      "grad_norm": 0.072265625,
      "learning_rate": 0.00015146778899102,
      "loss": 1.1222,
      "step": 675
    },
    {
      "epoch": 0.39795717753992604,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.00015058888026425212,
      "loss": 1.1177,
      "step": 680
    },
    {
      "epoch": 0.4008833332571314,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.00014970468777477026,
      "loss": 1.1181,
      "step": 685
    },
    {
      "epoch": 0.4038094889743367,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.00014881530387218325,
      "loss": 1.1417,
      "step": 690
    },
    {
      "epoch": 0.40673564469154205,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.00014792082144831793,
      "loss": 1.1302,
      "step": 695
    },
    {
      "epoch": 0.4096618004087474,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.00014702133392751688,
      "loss": 1.122,
      "step": 700
    },
    {
      "epoch": 0.4125879561259527,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.00014611693525688066,
      "loss": 1.1268,
      "step": 705
    },
    {
      "epoch": 0.41551411184315806,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00014520771989645563,
      "loss": 1.1238,
      "step": 710
    },
    {
      "epoch": 0.4184402675603634,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.00014429378280936804,
      "loss": 1.119,
      "step": 715
    },
    {
      "epoch": 0.42136642327756874,
      "grad_norm": 0.0703125,
      "learning_rate": 0.0001433752194519054,
      "loss": 1.1187,
      "step": 720
    },
    {
      "epoch": 0.42429257899477407,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.00014245212576354682,
      "loss": 1.122,
      "step": 725
    },
    {
      "epoch": 0.4272187347119794,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0001415245981569424,
      "loss": 1.1267,
      "step": 730
    },
    {
      "epoch": 0.43014489042918475,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.00014059273350784342,
      "loss": 1.1273,
      "step": 735
    },
    {
      "epoch": 0.4330710461463901,
      "grad_norm": 0.076171875,
      "learning_rate": 0.00013965662914498428,
      "loss": 1.1267,
      "step": 740
    },
    {
      "epoch": 0.4359972018635954,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.00013871638283991677,
      "loss": 1.1175,
      "step": 745
    },
    {
      "epoch": 0.43892335758080075,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0001377720927967985,
      "loss": 1.1211,
      "step": 750
    },
    {
      "epoch": 0.4418495132980061,
      "grad_norm": 0.1884765625,
      "learning_rate": 0.00013682385764213572,
      "loss": 1.1319,
      "step": 755
    },
    {
      "epoch": 0.44477566901521143,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.00013587177641448265,
      "loss": 1.1233,
      "step": 760
    },
    {
      "epoch": 0.44770182473241676,
      "grad_norm": 0.078125,
      "learning_rate": 0.00013491594855409697,
      "loss": 1.1385,
      "step": 765
    },
    {
      "epoch": 0.4506279804496221,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.00013395647389255396,
      "loss": 1.1189,
      "step": 770
    },
    {
      "epoch": 0.45355413616682744,
      "grad_norm": 0.072265625,
      "learning_rate": 0.00013299345264231957,
      "loss": 1.1157,
      "step": 775
    },
    {
      "epoch": 0.4564802918840328,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.00013202698538628376,
      "loss": 1.1224,
      "step": 780
    },
    {
      "epoch": 0.4594064476012381,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.00013105717306725501,
      "loss": 1.1283,
      "step": 785
    },
    {
      "epoch": 0.46233260331844345,
      "grad_norm": 0.0703125,
      "learning_rate": 0.0001300841169774174,
      "loss": 1.131,
      "step": 790
    },
    {
      "epoch": 0.4652587590356488,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.000129107918747751,
      "loss": 1.1175,
      "step": 795
    },
    {
      "epoch": 0.4681849147528541,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.00012812868033741724,
      "loss": 1.138,
      "step": 800
    },
    {
      "epoch": 0.47111107047005946,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.00012714650402310967,
      "loss": 1.1344,
      "step": 805
    },
    {
      "epoch": 0.47403722618726485,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.00012616149238837146,
      "loss": 1.1195,
      "step": 810
    },
    {
      "epoch": 0.4769633819044702,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.00012517374831288146,
      "loss": 1.1005,
      "step": 815
    },
    {
      "epoch": 0.4798895376216755,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.00012418337496170842,
      "loss": 1.1158,
      "step": 820
    },
    {
      "epoch": 0.48281569333888086,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.00012319047577453638,
      "loss": 1.1181,
      "step": 825
    },
    {
      "epoch": 0.4857418490560862,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.00012219515445486054,
      "loss": 1.1321,
      "step": 830
    },
    {
      "epoch": 0.48866800477329153,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.00012119751495915617,
      "loss": 1.1309,
      "step": 835
    },
    {
      "epoch": 0.49159416049049687,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.00012019766148602062,
      "loss": 1.1276,
      "step": 840
    },
    {
      "epoch": 0.4945203162077022,
      "grad_norm": 0.068359375,
      "learning_rate": 0.00011919569846529057,
      "loss": 1.1173,
      "step": 845
    },
    {
      "epoch": 0.49744647192490754,
      "grad_norm": 0.0703125,
      "learning_rate": 0.00011819173054713466,
      "loss": 1.1111,
      "step": 850
    },
    {
      "epoch": 0.5003726276421129,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.00011718586259112326,
      "loss": 1.1137,
      "step": 855
    },
    {
      "epoch": 0.5032987833593182,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0001161781996552765,
      "loss": 1.1157,
      "step": 860
    },
    {
      "epoch": 0.5062249390765236,
      "grad_norm": 0.0703125,
      "learning_rate": 0.00011516884698509143,
      "loss": 1.1136,
      "step": 865
    },
    {
      "epoch": 0.5091510947937289,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.00011415791000254964,
      "loss": 1.1217,
      "step": 870
    },
    {
      "epoch": 0.5120772505109342,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.0001131454942951065,
      "loss": 1.119,
      "step": 875
    },
    {
      "epoch": 0.5150034062281396,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0001121317056046629,
      "loss": 1.1122,
      "step": 880
    },
    {
      "epoch": 0.5179295619453449,
      "grad_norm": 0.072265625,
      "learning_rate": 0.00011111664981652121,
      "loss": 1.1137,
      "step": 885
    },
    {
      "epoch": 0.5208557176625502,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.00011010043294832601,
      "loss": 1.1132,
      "step": 890
    },
    {
      "epoch": 0.5237818733797556,
      "grad_norm": 0.072265625,
      "learning_rate": 0.00010908316113899097,
      "loss": 1.1373,
      "step": 895
    },
    {
      "epoch": 0.5267080290969609,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.00010806494063761335,
      "loss": 1.1165,
      "step": 900
    },
    {
      "epoch": 0.5296341848141662,
      "grad_norm": 0.0703125,
      "learning_rate": 0.00010704587779237654,
      "loss": 1.1149,
      "step": 905
    },
    {
      "epoch": 0.5325603405313716,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00010602607903944279,
      "loss": 1.1244,
      "step": 910
    },
    {
      "epoch": 0.5354864962485769,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.00010500565089183627,
      "loss": 1.1141,
      "step": 915
    },
    {
      "epoch": 0.5384126519657823,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.00010398469992831832,
      "loss": 1.1031,
      "step": 920
    },
    {
      "epoch": 0.5413388076829876,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.00010296333278225599,
      "loss": 1.1072,
      "step": 925
    },
    {
      "epoch": 0.5442649634001929,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00010194165613048444,
      "loss": 1.0993,
      "step": 930
    },
    {
      "epoch": 0.5471911191173983,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.00010091977668216524,
      "loss": 1.1089,
      "step": 935
    },
    {
      "epoch": 0.5501172748346036,
      "grad_norm": 0.068359375,
      "learning_rate": 9.989780116764115e-05,
      "loss": 1.1042,
      "step": 940
    },
    {
      "epoch": 0.5530434305518089,
      "grad_norm": 0.0712890625,
      "learning_rate": 9.887583632728845e-05,
      "loss": 1.1062,
      "step": 945
    },
    {
      "epoch": 0.5559695862690143,
      "grad_norm": 0.07080078125,
      "learning_rate": 9.785398890036867e-05,
      "loss": 1.1092,
      "step": 950
    },
    {
      "epoch": 0.5588957419862196,
      "grad_norm": 0.0693359375,
      "learning_rate": 9.683236561388e-05,
      "loss": 1.1173,
      "step": 955
    },
    {
      "epoch": 0.5618218977034249,
      "grad_norm": 0.0703125,
      "learning_rate": 9.581107317141026e-05,
      "loss": 1.1407,
      "step": 960
    },
    {
      "epoch": 0.5647480534206303,
      "grad_norm": 0.07666015625,
      "learning_rate": 9.479021824199229e-05,
      "loss": 1.1365,
      "step": 965
    },
    {
      "epoch": 0.5676742091378356,
      "grad_norm": 0.07763671875,
      "learning_rate": 9.376990744896276e-05,
      "loss": 1.1031,
      "step": 970
    },
    {
      "epoch": 0.570600364855041,
      "grad_norm": 0.0703125,
      "learning_rate": 9.275024735882588e-05,
      "loss": 1.0896,
      "step": 975
    },
    {
      "epoch": 0.5735265205722463,
      "grad_norm": 0.07275390625,
      "learning_rate": 9.173134447012322e-05,
      "loss": 1.1094,
      "step": 980
    },
    {
      "epoch": 0.5764526762894516,
      "grad_norm": 0.0751953125,
      "learning_rate": 9.071330520231033e-05,
      "loss": 1.1127,
      "step": 985
    },
    {
      "epoch": 0.579378832006657,
      "grad_norm": 0.0693359375,
      "learning_rate": 8.969623588464163e-05,
      "loss": 1.1176,
      "step": 990
    },
    {
      "epoch": 0.5823049877238623,
      "grad_norm": 0.0703125,
      "learning_rate": 8.868024274506505e-05,
      "loss": 1.112,
      "step": 995
    },
    {
      "epoch": 0.5852311434410676,
      "grad_norm": 0.0712890625,
      "learning_rate": 8.766543189912705e-05,
      "loss": 1.0846,
      "step": 1000
    },
    {
      "epoch": 0.588157299158273,
      "grad_norm": 0.07177734375,
      "learning_rate": 8.665190933888904e-05,
      "loss": 1.0961,
      "step": 1005
    },
    {
      "epoch": 0.5910834548754783,
      "grad_norm": 0.0712890625,
      "learning_rate": 8.56397809218574e-05,
      "loss": 1.1146,
      "step": 1010
    },
    {
      "epoch": 0.5940096105926836,
      "grad_norm": 0.07568359375,
      "learning_rate": 8.4629152359927e-05,
      "loss": 1.1066,
      "step": 1015
    },
    {
      "epoch": 0.596935766309889,
      "grad_norm": 0.07080078125,
      "learning_rate": 8.362012920834014e-05,
      "loss": 1.1253,
      "step": 1020
    },
    {
      "epoch": 0.5998619220270943,
      "grad_norm": 0.0693359375,
      "learning_rate": 8.261281685466177e-05,
      "loss": 1.1072,
      "step": 1025
    },
    {
      "epoch": 0.6027880777442998,
      "grad_norm": 0.07373046875,
      "learning_rate": 8.160732050777235e-05,
      "loss": 1.1147,
      "step": 1030
    },
    {
      "epoch": 0.6057142334615051,
      "grad_norm": 0.0712890625,
      "learning_rate": 8.060374518687926e-05,
      "loss": 1.11,
      "step": 1035
    },
    {
      "epoch": 0.6086403891787104,
      "grad_norm": 0.07080078125,
      "learning_rate": 7.960219571054799e-05,
      "loss": 1.123,
      "step": 1040
    },
    {
      "epoch": 0.6115665448959158,
      "grad_norm": 0.06787109375,
      "learning_rate": 7.860277668575449e-05,
      "loss": 1.1035,
      "step": 1045
    },
    {
      "epoch": 0.6144927006131211,
      "grad_norm": 0.0712890625,
      "learning_rate": 7.76055924969594e-05,
      "loss": 1.09,
      "step": 1050
    },
    {
      "epoch": 0.6174188563303264,
      "grad_norm": 0.0712890625,
      "learning_rate": 7.661074729520548e-05,
      "loss": 1.1279,
      "step": 1055
    },
    {
      "epoch": 0.6203450120475318,
      "grad_norm": 0.072265625,
      "learning_rate": 7.561834498723974e-05,
      "loss": 1.1141,
      "step": 1060
    },
    {
      "epoch": 0.6232711677647371,
      "grad_norm": 0.068359375,
      "learning_rate": 7.462848922466092e-05,
      "loss": 1.1102,
      "step": 1065
    },
    {
      "epoch": 0.6261973234819425,
      "grad_norm": 0.0673828125,
      "learning_rate": 7.364128339309326e-05,
      "loss": 1.1128,
      "step": 1070
    },
    {
      "epoch": 0.6291234791991478,
      "grad_norm": 0.07275390625,
      "learning_rate": 7.265683060138868e-05,
      "loss": 1.1054,
      "step": 1075
    },
    {
      "epoch": 0.6320496349163531,
      "grad_norm": 0.07666015625,
      "learning_rate": 7.167523367085749e-05,
      "loss": 1.1097,
      "step": 1080
    },
    {
      "epoch": 0.6349757906335585,
      "grad_norm": 0.0673828125,
      "learning_rate": 7.069659512452918e-05,
      "loss": 1.1148,
      "step": 1085
    },
    {
      "epoch": 0.6379019463507638,
      "grad_norm": 0.06982421875,
      "learning_rate": 6.972101717644429e-05,
      "loss": 1.0997,
      "step": 1090
    },
    {
      "epoch": 0.6408281020679691,
      "grad_norm": 0.07080078125,
      "learning_rate": 6.874860172097883e-05,
      "loss": 1.097,
      "step": 1095
    },
    {
      "epoch": 0.6437542577851745,
      "grad_norm": 0.07177734375,
      "learning_rate": 6.777945032220187e-05,
      "loss": 1.1006,
      "step": 1100
    },
    {
      "epoch": 0.6466804135023798,
      "grad_norm": 0.0693359375,
      "learning_rate": 6.681366420326747e-05,
      "loss": 1.1191,
      "step": 1105
    },
    {
      "epoch": 0.6496065692195852,
      "grad_norm": 0.162109375,
      "learning_rate": 6.58513442358427e-05,
      "loss": 1.0901,
      "step": 1110
    },
    {
      "epoch": 0.6525327249367905,
      "grad_norm": 0.0712890625,
      "learning_rate": 6.489259092957193e-05,
      "loss": 1.1113,
      "step": 1115
    },
    {
      "epoch": 0.6554588806539958,
      "grad_norm": 0.0751953125,
      "learning_rate": 6.3937504421579e-05,
      "loss": 1.0945,
      "step": 1120
    },
    {
      "epoch": 0.6583850363712012,
      "grad_norm": 0.06884765625,
      "learning_rate": 6.298618446600856e-05,
      "loss": 1.1073,
      "step": 1125
    },
    {
      "epoch": 0.6613111920884065,
      "grad_norm": 0.0693359375,
      "learning_rate": 6.203873042360722e-05,
      "loss": 1.1178,
      "step": 1130
    },
    {
      "epoch": 0.6642373478056118,
      "grad_norm": 0.072265625,
      "learning_rate": 6.109524125134571e-05,
      "loss": 1.1291,
      "step": 1135
    },
    {
      "epoch": 0.6671635035228172,
      "grad_norm": 0.07080078125,
      "learning_rate": 6.015581549208322e-05,
      "loss": 1.0985,
      "step": 1140
    },
    {
      "epoch": 0.6700896592400225,
      "grad_norm": 0.06884765625,
      "learning_rate": 5.9220551264275356e-05,
      "loss": 1.1294,
      "step": 1145
    },
    {
      "epoch": 0.6730158149572278,
      "grad_norm": 0.06884765625,
      "learning_rate": 5.828954625172597e-05,
      "loss": 1.115,
      "step": 1150
    },
    {
      "epoch": 0.6759419706744332,
      "grad_norm": 0.06689453125,
      "learning_rate": 5.736289769338441e-05,
      "loss": 1.1024,
      "step": 1155
    },
    {
      "epoch": 0.6788681263916385,
      "grad_norm": 0.0703125,
      "learning_rate": 5.644070237318977e-05,
      "loss": 1.0993,
      "step": 1160
    },
    {
      "epoch": 0.6817942821088439,
      "grad_norm": 0.0693359375,
      "learning_rate": 5.552305660996202e-05,
      "loss": 1.1172,
      "step": 1165
    },
    {
      "epoch": 0.6847204378260492,
      "grad_norm": 0.0693359375,
      "learning_rate": 5.4610056247341814e-05,
      "loss": 1.0988,
      "step": 1170
    },
    {
      "epoch": 0.6876465935432545,
      "grad_norm": 0.0712890625,
      "learning_rate": 5.3701796643780524e-05,
      "loss": 1.1142,
      "step": 1175
    },
    {
      "epoch": 0.6905727492604599,
      "grad_norm": 0.0693359375,
      "learning_rate": 5.279837266258016e-05,
      "loss": 1.1271,
      "step": 1180
    },
    {
      "epoch": 0.6934989049776652,
      "grad_norm": 0.06884765625,
      "learning_rate": 5.189987866198548e-05,
      "loss": 1.1055,
      "step": 1185
    },
    {
      "epoch": 0.6964250606948705,
      "grad_norm": 0.06982421875,
      "learning_rate": 5.100640848532878e-05,
      "loss": 1.1277,
      "step": 1190
    },
    {
      "epoch": 0.6993512164120759,
      "grad_norm": 0.06787109375,
      "learning_rate": 5.011805545122826e-05,
      "loss": 1.1093,
      "step": 1195
    },
    {
      "epoch": 0.7022773721292812,
      "grad_norm": 0.06884765625,
      "learning_rate": 4.923491234384158e-05,
      "loss": 1.1055,
      "step": 1200
    },
    {
      "epoch": 0.7052035278464865,
      "grad_norm": 0.0693359375,
      "learning_rate": 4.8357071403174746e-05,
      "loss": 1.098,
      "step": 1205
    },
    {
      "epoch": 0.7081296835636919,
      "grad_norm": 0.06689453125,
      "learning_rate": 4.748462431544826e-05,
      "loss": 1.114,
      "step": 1210
    },
    {
      "epoch": 0.7110558392808972,
      "grad_norm": 0.06884765625,
      "learning_rate": 4.661766220352097e-05,
      "loss": 1.1073,
      "step": 1215
    },
    {
      "epoch": 0.7139819949981026,
      "grad_norm": 0.06982421875,
      "learning_rate": 4.5756275617372465e-05,
      "loss": 1.1121,
      "step": 1220
    },
    {
      "epoch": 0.7169081507153079,
      "grad_norm": 0.0693359375,
      "learning_rate": 4.490055452464594e-05,
      "loss": 1.12,
      "step": 1225
    },
    {
      "epoch": 0.7198343064325132,
      "grad_norm": 0.06787109375,
      "learning_rate": 4.405058830125137e-05,
      "loss": 1.092,
      "step": 1230
    },
    {
      "epoch": 0.7227604621497186,
      "grad_norm": 0.06640625,
      "learning_rate": 4.320646572203033e-05,
      "loss": 1.0998,
      "step": 1235
    },
    {
      "epoch": 0.7256866178669239,
      "grad_norm": 0.06982421875,
      "learning_rate": 4.236827495148443e-05,
      "loss": 1.0993,
      "step": 1240
    },
    {
      "epoch": 0.7286127735841292,
      "grad_norm": 0.068359375,
      "learning_rate": 4.153610353456654e-05,
      "loss": 1.1323,
      "step": 1245
    },
    {
      "epoch": 0.7315389293013346,
      "grad_norm": 0.06982421875,
      "learning_rate": 4.071003838753737e-05,
      "loss": 1.1264,
      "step": 1250
    },
    {
      "epoch": 0.7344650850185399,
      "grad_norm": 0.0693359375,
      "learning_rate": 3.9890165788887365e-05,
      "loss": 1.1057,
      "step": 1255
    },
    {
      "epoch": 0.7373912407357452,
      "grad_norm": 0.06884765625,
      "learning_rate": 3.9076571370325364e-05,
      "loss": 1.1119,
      "step": 1260
    },
    {
      "epoch": 0.7403173964529506,
      "grad_norm": 0.0712890625,
      "learning_rate": 3.82693401078349e-05,
      "loss": 1.0996,
      "step": 1265
    },
    {
      "epoch": 0.7432435521701559,
      "grad_norm": 0.06787109375,
      "learning_rate": 3.7468556312798685e-05,
      "loss": 1.1051,
      "step": 1270
    },
    {
      "epoch": 0.7461697078873613,
      "grad_norm": 0.06787109375,
      "learning_rate": 3.667430362319277e-05,
      "loss": 1.0959,
      "step": 1275
    },
    {
      "epoch": 0.7490958636045666,
      "grad_norm": 0.068359375,
      "learning_rate": 3.588666499485115e-05,
      "loss": 1.1129,
      "step": 1280
    },
    {
      "epoch": 0.7520220193217719,
      "grad_norm": 0.0693359375,
      "learning_rate": 3.510572269280097e-05,
      "loss": 1.1184,
      "step": 1285
    },
    {
      "epoch": 0.7549481750389773,
      "grad_norm": 0.06787109375,
      "learning_rate": 3.433155828267089e-05,
      "loss": 1.1003,
      "step": 1290
    },
    {
      "epoch": 0.7578743307561826,
      "grad_norm": 0.06787109375,
      "learning_rate": 3.356425262217164e-05,
      "loss": 1.106,
      "step": 1295
    },
    {
      "epoch": 0.7608004864733879,
      "grad_norm": 0.06787109375,
      "learning_rate": 3.280388585265075e-05,
      "loss": 1.1066,
      "step": 1300
    },
    {
      "epoch": 0.7637266421905933,
      "grad_norm": 0.06689453125,
      "learning_rate": 3.205053739072248e-05,
      "loss": 1.1026,
      "step": 1305
    },
    {
      "epoch": 0.7666527979077986,
      "grad_norm": 0.06787109375,
      "learning_rate": 3.130428591997282e-05,
      "loss": 1.1256,
      "step": 1310
    },
    {
      "epoch": 0.769578953625004,
      "grad_norm": 0.0673828125,
      "learning_rate": 3.0565209382741664e-05,
      "loss": 1.1018,
      "step": 1315
    },
    {
      "epoch": 0.7725051093422093,
      "grad_norm": 0.068359375,
      "learning_rate": 2.9833384971981838e-05,
      "loss": 1.1099,
      "step": 1320
    },
    {
      "epoch": 0.7754312650594147,
      "grad_norm": 0.0673828125,
      "learning_rate": 2.9108889123196824e-05,
      "loss": 1.0995,
      "step": 1325
    },
    {
      "epoch": 0.7783574207766201,
      "grad_norm": 0.0673828125,
      "learning_rate": 2.839179750645752e-05,
      "loss": 1.1194,
      "step": 1330
    },
    {
      "epoch": 0.7812835764938254,
      "grad_norm": 0.0693359375,
      "learning_rate": 2.768218501849862e-05,
      "loss": 1.0955,
      "step": 1335
    },
    {
      "epoch": 0.7842097322110307,
      "grad_norm": 0.06591796875,
      "learning_rate": 2.6980125774896238e-05,
      "loss": 1.0712,
      "step": 1340
    },
    {
      "epoch": 0.7871358879282361,
      "grad_norm": 0.0693359375,
      "learning_rate": 2.6285693102326868e-05,
      "loss": 1.1019,
      "step": 1345
    },
    {
      "epoch": 0.7900620436454414,
      "grad_norm": 0.0673828125,
      "learning_rate": 2.559895953090856e-05,
      "loss": 1.1022,
      "step": 1350
    },
    {
      "epoch": 0.7929881993626468,
      "grad_norm": 0.06787109375,
      "learning_rate": 2.491999678662582e-05,
      "loss": 1.1027,
      "step": 1355
    },
    {
      "epoch": 0.7959143550798521,
      "grad_norm": 0.06787109375,
      "learning_rate": 2.4248875783837987e-05,
      "loss": 1.1292,
      "step": 1360
    },
    {
      "epoch": 0.7988405107970574,
      "grad_norm": 0.0693359375,
      "learning_rate": 2.358566661787257e-05,
      "loss": 1.1117,
      "step": 1365
    },
    {
      "epoch": 0.8017666665142628,
      "grad_norm": 0.06787109375,
      "learning_rate": 2.293043855770416e-05,
      "loss": 1.1176,
      "step": 1370
    },
    {
      "epoch": 0.8046928222314681,
      "grad_norm": 0.06982421875,
      "learning_rate": 2.2283260038719646e-05,
      "loss": 1.1074,
      "step": 1375
    },
    {
      "epoch": 0.8076189779486734,
      "grad_norm": 0.06689453125,
      "learning_rate": 2.1644198655570504e-05,
      "loss": 1.1123,
      "step": 1380
    },
    {
      "epoch": 0.8105451336658788,
      "grad_norm": 0.068359375,
      "learning_rate": 2.1013321155112754e-05,
      "loss": 1.0979,
      "step": 1385
    },
    {
      "epoch": 0.8134712893830841,
      "grad_norm": 0.068359375,
      "learning_rate": 2.0390693429435627e-05,
      "loss": 1.1102,
      "step": 1390
    },
    {
      "epoch": 0.8163974451002894,
      "grad_norm": 0.06884765625,
      "learning_rate": 1.977638050897954e-05,
      "loss": 1.1133,
      "step": 1395
    },
    {
      "epoch": 0.8193236008174948,
      "grad_norm": 0.0673828125,
      "learning_rate": 1.917044655574387e-05,
      "loss": 1.1045,
      "step": 1400
    },
    {
      "epoch": 0.8222497565347001,
      "grad_norm": 0.0673828125,
      "learning_rate": 1.8572954856585535e-05,
      "loss": 1.0967,
      "step": 1405
    },
    {
      "epoch": 0.8251759122519055,
      "grad_norm": 0.0673828125,
      "learning_rate": 1.798396781660914e-05,
      "loss": 1.1199,
      "step": 1410
    },
    {
      "epoch": 0.8281020679691108,
      "grad_norm": 0.06884765625,
      "learning_rate": 1.7403546952648885e-05,
      "loss": 1.1039,
      "step": 1415
    },
    {
      "epoch": 0.8310282236863161,
      "grad_norm": 0.06787109375,
      "learning_rate": 1.6831752886843512e-05,
      "loss": 1.1106,
      "step": 1420
    },
    {
      "epoch": 0.8339543794035215,
      "grad_norm": 0.06982421875,
      "learning_rate": 1.626864534030469e-05,
      "loss": 1.106,
      "step": 1425
    },
    {
      "epoch": 0.8368805351207268,
      "grad_norm": 0.0673828125,
      "learning_rate": 1.571428312687928e-05,
      "loss": 1.1004,
      "step": 1430
    },
    {
      "epoch": 0.8398066908379321,
      "grad_norm": 0.0693359375,
      "learning_rate": 1.5168724147006652e-05,
      "loss": 1.1244,
      "step": 1435
    },
    {
      "epoch": 0.8427328465551375,
      "grad_norm": 0.0673828125,
      "learning_rate": 1.4632025381671133e-05,
      "loss": 1.1227,
      "step": 1440
    },
    {
      "epoch": 0.8456590022723428,
      "grad_norm": 0.0673828125,
      "learning_rate": 1.4104242886450824e-05,
      "loss": 1.1073,
      "step": 1445
    },
    {
      "epoch": 0.8485851579895481,
      "grad_norm": 0.07568359375,
      "learning_rate": 1.3585431785662627e-05,
      "loss": 1.0903,
      "step": 1450
    },
    {
      "epoch": 0.8515113137067535,
      "grad_norm": 0.06884765625,
      "learning_rate": 1.3075646266604913e-05,
      "loss": 1.1129,
      "step": 1455
    },
    {
      "epoch": 0.8544374694239588,
      "grad_norm": 0.06982421875,
      "learning_rate": 1.257493957389796e-05,
      "loss": 1.1293,
      "step": 1460
    },
    {
      "epoch": 0.8573636251411642,
      "grad_norm": 0.0693359375,
      "learning_rate": 1.208336400392268e-05,
      "loss": 1.0987,
      "step": 1465
    },
    {
      "epoch": 0.8602897808583695,
      "grad_norm": 0.06689453125,
      "learning_rate": 1.1600970899358588e-05,
      "loss": 1.1044,
      "step": 1470
    },
    {
      "epoch": 0.8632159365755748,
      "grad_norm": 0.0693359375,
      "learning_rate": 1.1127810643821401e-05,
      "loss": 1.1182,
      "step": 1475
    },
    {
      "epoch": 0.8661420922927802,
      "grad_norm": 0.0673828125,
      "learning_rate": 1.0663932656600505e-05,
      "loss": 1.0957,
      "step": 1480
    },
    {
      "epoch": 0.8690682480099855,
      "grad_norm": 0.06689453125,
      "learning_rate": 1.0209385387497517e-05,
      "loss": 1.1238,
      "step": 1485
    },
    {
      "epoch": 0.8719944037271908,
      "grad_norm": 0.068359375,
      "learning_rate": 9.764216311765905e-06,
      "loss": 1.1209,
      "step": 1490
    },
    {
      "epoch": 0.8749205594443962,
      "grad_norm": 0.06591796875,
      "learning_rate": 9.328471925152381e-06,
      "loss": 1.1046,
      "step": 1495
    },
    {
      "epoch": 0.8778467151616015,
      "grad_norm": 0.06787109375,
      "learning_rate": 8.902197739040708e-06,
      "loss": 1.1205,
      "step": 1500
    },
    {
      "epoch": 0.8807728708788068,
      "grad_norm": 0.06787109375,
      "learning_rate": 8.485438275698154e-06,
      "loss": 1.1202,
      "step": 1505
    },
    {
      "epoch": 0.8836990265960122,
      "grad_norm": 0.07080078125,
      "learning_rate": 8.078237063625538e-06,
      "loss": 1.1177,
      "step": 1510
    },
    {
      "epoch": 0.8866251823132175,
      "grad_norm": 0.0673828125,
      "learning_rate": 7.680636633010695e-06,
      "loss": 1.1116,
      "step": 1515
    },
    {
      "epoch": 0.8895513380304229,
      "grad_norm": 0.0712890625,
      "learning_rate": 7.292678511286522e-06,
      "loss": 1.1067,
      "step": 1520
    },
    {
      "epoch": 0.8924774937476282,
      "grad_norm": 0.06884765625,
      "learning_rate": 6.914403218793608e-06,
      "loss": 1.12,
      "step": 1525
    },
    {
      "epoch": 0.8954036494648335,
      "grad_norm": 0.0693359375,
      "learning_rate": 6.5458502645480924e-06,
      "loss": 1.1298,
      "step": 1530
    },
    {
      "epoch": 0.8983298051820389,
      "grad_norm": 0.0673828125,
      "learning_rate": 6.187058142115077e-06,
      "loss": 1.1069,
      "step": 1535
    },
    {
      "epoch": 0.9012559608992442,
      "grad_norm": 0.06640625,
      "learning_rate": 5.838064325588288e-06,
      "loss": 1.0941,
      "step": 1540
    },
    {
      "epoch": 0.9041821166164495,
      "grad_norm": 0.06640625,
      "learning_rate": 5.498905265675958e-06,
      "loss": 1.0976,
      "step": 1545
    },
    {
      "epoch": 0.9071082723336549,
      "grad_norm": 0.06787109375,
      "learning_rate": 5.169616385893794e-06,
      "loss": 1.101,
      "step": 1550
    },
    {
      "epoch": 0.9100344280508602,
      "grad_norm": 0.0654296875,
      "learning_rate": 4.850232078865169e-06,
      "loss": 1.121,
      "step": 1555
    },
    {
      "epoch": 0.9129605837680655,
      "grad_norm": 0.068359375,
      "learning_rate": 4.5407857027289555e-06,
      "loss": 1.1013,
      "step": 1560
    },
    {
      "epoch": 0.9158867394852709,
      "grad_norm": 0.0693359375,
      "learning_rate": 4.241309577655406e-06,
      "loss": 1.1464,
      "step": 1565
    },
    {
      "epoch": 0.9188128952024762,
      "grad_norm": 0.0654296875,
      "learning_rate": 3.951834982470526e-06,
      "loss": 1.1111,
      "step": 1570
    },
    {
      "epoch": 0.9217390509196816,
      "grad_norm": 0.06640625,
      "learning_rate": 3.672392151389137e-06,
      "loss": 1.1078,
      "step": 1575
    },
    {
      "epoch": 0.9246652066368869,
      "grad_norm": 0.07080078125,
      "learning_rate": 3.4030102708570212e-06,
      "loss": 1.1195,
      "step": 1580
    },
    {
      "epoch": 0.9275913623540922,
      "grad_norm": 0.0703125,
      "learning_rate": 3.143717476502572e-06,
      "loss": 1.1216,
      "step": 1585
    },
    {
      "epoch": 0.9305175180712976,
      "grad_norm": 0.0673828125,
      "learning_rate": 2.8945408501981906e-06,
      "loss": 1.0936,
      "step": 1590
    },
    {
      "epoch": 0.9334436737885029,
      "grad_norm": 0.06787109375,
      "learning_rate": 2.6555064172316234e-06,
      "loss": 1.106,
      "step": 1595
    },
    {
      "epoch": 0.9363698295057082,
      "grad_norm": 0.068359375,
      "learning_rate": 2.4266391435878387e-06,
      "loss": 1.1061,
      "step": 1600
    },
    {
      "epoch": 0.9392959852229136,
      "grad_norm": 0.068359375,
      "learning_rate": 2.2079629333414453e-06,
      "loss": 1.1119,
      "step": 1605
    },
    {
      "epoch": 0.9422221409401189,
      "grad_norm": 0.06787109375,
      "learning_rate": 1.999500626159967e-06,
      "loss": 1.1094,
      "step": 1610
    },
    {
      "epoch": 0.9451482966573242,
      "grad_norm": 0.06884765625,
      "learning_rate": 1.8012739949183844e-06,
      "loss": 1.1141,
      "step": 1615
    },
    {
      "epoch": 0.9480744523745297,
      "grad_norm": 0.06591796875,
      "learning_rate": 1.6133037434250985e-06,
      "loss": 1.1084,
      "step": 1620
    },
    {
      "epoch": 0.951000608091735,
      "grad_norm": 0.068359375,
      "learning_rate": 1.4356095042594386e-06,
      "loss": 1.1208,
      "step": 1625
    },
    {
      "epoch": 0.9539267638089404,
      "grad_norm": 0.0693359375,
      "learning_rate": 1.2682098367212237e-06,
      "loss": 1.1116,
      "step": 1630
    },
    {
      "epoch": 0.9568529195261457,
      "grad_norm": 0.06640625,
      "learning_rate": 1.1111222248922471e-06,
      "loss": 1.1047,
      "step": 1635
    },
    {
      "epoch": 0.959779075243351,
      "grad_norm": 0.0673828125,
      "learning_rate": 9.643630758102484e-07,
      "loss": 1.0998,
      "step": 1640
    },
    {
      "epoch": 0.9627052309605564,
      "grad_norm": 0.0654296875,
      "learning_rate": 8.279477177551842e-07,
      "loss": 1.1073,
      "step": 1645
    },
    {
      "epoch": 0.9656313866777617,
      "grad_norm": 0.0673828125,
      "learning_rate": 7.018903986483083e-07,
      "loss": 1.1124,
      "step": 1650
    },
    {
      "epoch": 0.968557542394967,
      "grad_norm": 0.0654296875,
      "learning_rate": 5.862042845640403e-07,
      "loss": 1.1023,
      "step": 1655
    },
    {
      "epoch": 0.9714836981121724,
      "grad_norm": 0.06396484375,
      "learning_rate": 4.809014583548432e-07,
      "loss": 1.1234,
      "step": 1660
    },
    {
      "epoch": 0.9744098538293777,
      "grad_norm": 0.06982421875,
      "learning_rate": 3.859929183892108e-07,
      "loss": 1.111,
      "step": 1665
    },
    {
      "epoch": 0.9773360095465831,
      "grad_norm": 0.06787109375,
      "learning_rate": 3.014885774029419e-07,
      "loss": 1.1179,
      "step": 1670
    },
    {
      "epoch": 0.9802621652637884,
      "grad_norm": 0.064453125,
      "learning_rate": 2.2739726146381311e-07,
      "loss": 1.1082,
      "step": 1675
    },
    {
      "epoch": 0.9831883209809937,
      "grad_norm": 0.06982421875,
      "learning_rate": 1.6372670904974963e-07,
      "loss": 1.14,
      "step": 1680
    },
    {
      "epoch": 0.9861144766981991,
      "grad_norm": 0.06591796875,
      "learning_rate": 1.1048357024054934e-07,
      "loss": 1.1126,
      "step": 1685
    },
    {
      "epoch": 0.9890406324154044,
      "grad_norm": 0.06689453125,
      "learning_rate": 6.76734060233275e-08,
      "loss": 1.1087,
      "step": 1690
    },
    {
      "epoch": 0.9919667881326097,
      "grad_norm": 0.06591796875,
      "learning_rate": 3.5300687711703475e-08,
      "loss": 1.1235,
      "step": 1695
    },
    {
      "epoch": 0.9948929438498151,
      "grad_norm": 0.0654296875,
      "learning_rate": 1.3368796478807621e-08,
      "loss": 1.1019,
      "step": 1700
    },
    {
      "epoch": 0.9978190995670204,
      "grad_norm": 0.06787109375,
      "learning_rate": 1.8800230040860733e-09,
      "loss": 1.1143,
      "step": 1705
    },
    {
      "epoch": 0.9995747929973436,
      "eval_loss": 1.278271198272705,
      "eval_runtime": 1249.6014,
      "eval_samples_per_second": 12.937,
      "eval_steps_per_second": 12.937,
      "step": 1708
    },
    {
      "epoch": 0.9995747929973436,
      "step": 1708,
      "total_flos": 2.8130589802160456e+18,
      "train_loss": 0.978924176871637,
      "train_runtime": 53778.1468,
      "train_samples_per_second": 4.067,
      "train_steps_per_second": 0.032
    }
  ],
  "logging_steps": 5,
  "max_steps": 1708,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 20,
  "total_flos": 2.8130589802160456e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}