{
  "best_global_step": 1230,
  "best_metric": 0.10221888870000839,
  "best_model_checkpoint": "factory_qwen_results1/checkpoint-1230",
  "epoch": 3.0,
  "eval_steps": 30,
  "global_step": 1245,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.024125452352231604,
      "grad_norm": 0.2783676878347129,
      "learning_rate": 0.0002942959550338895,
      "loss": 0.6652,
      "step": 10
    },
    {
      "epoch": 0.04825090470446321,
      "grad_norm": 0.28703240039853717,
      "learning_rate": 0.00038288786510166846,
      "loss": 0.3853,
      "step": 20
    },
    {
      "epoch": 0.07237635705669482,
      "grad_norm": 0.23489702427244202,
      "learning_rate": 0.00043471081035858025,
      "loss": 0.3137,
      "step": 30
    },
    {
      "epoch": 0.07237635705669482,
      "eval_accuracy": 0.926024110161071,
      "eval_loss": 0.30167156457901,
      "eval_runtime": 405.387,
      "eval_samples_per_second": 0.91,
      "eval_steps_per_second": 0.116,
      "step": 30
    },
    {
      "epoch": 0.09650180940892641,
      "grad_norm": 0.1754393640352717,
      "learning_rate": 0.0004714797751694474,
      "loss": 0.2954,
      "step": 40
    },
    {
      "epoch": 0.12062726176115803,
      "grad_norm": 0.15708221934215447,
      "learning_rate": 0.0005,
      "loss": 0.2754,
      "step": 50
    },
    {
      "epoch": 0.14475271411338964,
      "grad_norm": 0.1415910434185202,
      "learning_rate": 0.000496234309623431,
      "loss": 0.2642,
      "step": 60
    },
    {
      "epoch": 0.14475271411338964,
      "eval_accuracy": 0.9343403177634182,
      "eval_loss": 0.2687387764453888,
      "eval_runtime": 403.2157,
      "eval_samples_per_second": 0.915,
      "eval_steps_per_second": 0.117,
      "step": 60
    },
    {
      "epoch": 0.16887816646562123,
      "grad_norm": 0.7233999563282641,
      "learning_rate": 0.0004920502092050209,
      "loss": 0.246,
      "step": 70
    },
    {
      "epoch": 0.19300361881785283,
      "grad_norm": 0.15126714843742456,
      "learning_rate": 0.00048786610878661084,
      "loss": 0.2617,
      "step": 80
    },
    {
      "epoch": 0.21712907117008443,
      "grad_norm": 0.18113362419377327,
      "learning_rate": 0.00048368200836820086,
      "loss": 0.2464,
      "step": 90
    },
    {
      "epoch": 0.21712907117008443,
      "eval_accuracy": 0.9387037062326197,
      "eval_loss": 0.2396717518568039,
      "eval_runtime": 403.9587,
      "eval_samples_per_second": 0.913,
      "eval_steps_per_second": 0.116,
      "step": 90
    },
    {
      "epoch": 0.24125452352231605,
      "grad_norm": 0.16359059325760722,
      "learning_rate": 0.0004794979079497908,
      "loss": 0.2269,
      "step": 100
    },
    {
      "epoch": 0.26537997587454765,
      "grad_norm": 0.21182264228144074,
      "learning_rate": 0.00047531380753138073,
      "loss": 0.2174,
      "step": 110
    },
    {
      "epoch": 0.28950542822677927,
      "grad_norm": 0.15334984385329428,
      "learning_rate": 0.0004711297071129707,
      "loss": 0.2345,
      "step": 120
    },
    {
      "epoch": 0.28950542822677927,
      "eval_accuracy": 0.9462369166670195,
      "eval_loss": 0.21788497269153595,
      "eval_runtime": 403.0123,
      "eval_samples_per_second": 0.916,
      "eval_steps_per_second": 0.117,
      "step": 120
    },
    {
      "epoch": 0.31363088057901084,
      "grad_norm": 0.18749056865185024,
      "learning_rate": 0.00046694560669456066,
      "loss": 0.2062,
      "step": 130
    },
    {
      "epoch": 0.33775633293124246,
      "grad_norm": 0.13840158791299492,
      "learning_rate": 0.0004627615062761507,
      "loss": 0.1912,
      "step": 140
    },
    {
      "epoch": 0.3618817852834741,
      "grad_norm": 0.13775767602582611,
      "learning_rate": 0.0004585774058577406,
      "loss": 0.2104,
      "step": 150
    },
    {
      "epoch": 0.3618817852834741,
      "eval_accuracy": 0.9488062413732385,
      "eval_loss": 0.20275263488292694,
      "eval_runtime": 401.6942,
      "eval_samples_per_second": 0.919,
      "eval_steps_per_second": 0.117,
      "step": 150
    },
    {
      "epoch": 0.38600723763570566,
      "grad_norm": 0.1050841146486849,
      "learning_rate": 0.00045439330543933055,
      "loss": 0.1603,
      "step": 160
    },
    {
      "epoch": 0.4101326899879373,
      "grad_norm": 0.18927635582411037,
      "learning_rate": 0.0004502092050209205,
      "loss": 0.2139,
      "step": 170
    },
    {
      "epoch": 0.43425814234016885,
      "grad_norm": 0.11880806838601764,
      "learning_rate": 0.0004460251046025104,
      "loss": 0.1645,
      "step": 180
    },
    {
      "epoch": 0.43425814234016885,
      "eval_accuracy": 0.9499409173007779,
      "eval_loss": 0.20010298490524292,
      "eval_runtime": 403.852,
      "eval_samples_per_second": 0.914,
      "eval_steps_per_second": 0.116,
      "step": 180
    },
    {
      "epoch": 0.4583835946924005,
      "grad_norm": 0.2014843056895063,
      "learning_rate": 0.00044184100418410044,
      "loss": 0.187,
      "step": 190
    },
    {
      "epoch": 0.4825090470446321,
      "grad_norm": 0.18231897218308146,
      "learning_rate": 0.0004376569037656904,
      "loss": 0.154,
      "step": 200
    },
    {
      "epoch": 0.5066344993968637,
      "grad_norm": 0.13194167945743362,
      "learning_rate": 0.00043347280334728037,
      "loss": 0.1761,
      "step": 210
    },
    {
      "epoch": 0.5066344993968637,
      "eval_accuracy": 0.9543018942665147,
      "eval_loss": 0.1826241910457611,
      "eval_runtime": 404.3782,
      "eval_samples_per_second": 0.913,
      "eval_steps_per_second": 0.116,
      "step": 210
    },
    {
      "epoch": 0.5307599517490953,
      "grad_norm": 0.14917404579787943,
      "learning_rate": 0.0004292887029288703,
      "loss": 0.1842,
      "step": 220
    },
    {
      "epoch": 0.5548854041013269,
      "grad_norm": 0.15237265227811675,
      "learning_rate": 0.00042510460251046024,
      "loss": 0.1958,
      "step": 230
    },
    {
      "epoch": 0.5790108564535585,
      "grad_norm": 0.14049761965942137,
      "learning_rate": 0.00042092050209205026,
      "loss": 0.1668,
      "step": 240
    },
    {
      "epoch": 0.5790108564535585,
      "eval_accuracy": 0.9568256826398248,
      "eval_loss": 0.17410343885421753,
      "eval_runtime": 402.6992,
      "eval_samples_per_second": 0.916,
      "eval_steps_per_second": 0.117,
      "step": 240
    },
    {
      "epoch": 0.6031363088057901,
      "grad_norm": 0.112802778835781,
      "learning_rate": 0.00041673640167364017,
      "loss": 0.1751,
      "step": 250
    },
    {
      "epoch": 0.6272617611580217,
      "grad_norm": 0.11393045286666546,
      "learning_rate": 0.00041255230125523013,
      "loss": 0.2251,
      "step": 260
    },
    {
      "epoch": 0.6513872135102533,
      "grad_norm": 0.08862975282341802,
      "learning_rate": 0.0004083682008368201,
      "loss": 0.156,
      "step": 270
    },
    {
      "epoch": 0.6513872135102533,
      "eval_accuracy": 0.9566394963210819,
      "eval_loss": 0.16717860102653503,
      "eval_runtime": 404.0672,
      "eval_samples_per_second": 0.913,
      "eval_steps_per_second": 0.116,
      "step": 270
    },
    {
      "epoch": 0.6755126658624849,
      "grad_norm": 0.0905056489896768,
      "learning_rate": 0.00040418410041841,
      "loss": 0.1805,
      "step": 280
    },
    {
      "epoch": 0.6996381182147166,
      "grad_norm": 0.1484925416388826,
      "learning_rate": 0.0004,
      "loss": 0.163,
      "step": 290
    },
    {
      "epoch": 0.7237635705669482,
      "grad_norm": 0.09857653563229761,
      "learning_rate": 0.00039581589958159,
      "loss": 0.1416,
      "step": 300
    },
    {
      "epoch": 0.7237635705669482,
      "eval_accuracy": 0.9553425424182775,
      "eval_loss": 0.16858066618442535,
      "eval_runtime": 407.9196,
      "eval_samples_per_second": 0.905,
      "eval_steps_per_second": 0.115,
      "step": 300
    },
    {
      "epoch": 0.7478890229191797,
      "grad_norm": 0.17615701586838117,
      "learning_rate": 0.00039163179916317995,
      "loss": 0.1274,
      "step": 310
    },
    {
      "epoch": 0.7720144752714113,
      "grad_norm": 0.14037034593387046,
      "learning_rate": 0.00038744769874476985,
      "loss": 0.1444,
      "step": 320
    },
    {
      "epoch": 0.7961399276236429,
      "grad_norm": 0.11153876921530137,
      "learning_rate": 0.0003832635983263598,
      "loss": 0.1361,
      "step": 330
    },
    {
      "epoch": 0.7961399276236429,
      "eval_accuracy": 0.9591870578430155,
      "eval_loss": 0.15865620970726013,
      "eval_runtime": 403.3152,
      "eval_samples_per_second": 0.915,
      "eval_steps_per_second": 0.117,
      "step": 330
    },
    {
      "epoch": 0.8202653799758746,
      "grad_norm": 0.08620383733148518,
      "learning_rate": 0.00037907949790794984,
      "loss": 0.1424,
      "step": 340
    },
    {
      "epoch": 0.8443908323281062,
      "grad_norm": 0.17837451819381245,
      "learning_rate": 0.00037489539748953974,
      "loss": 0.1467,
      "step": 350
    },
    {
      "epoch": 0.8685162846803377,
      "grad_norm": 0.11827457610350564,
      "learning_rate": 0.0003707112970711297,
      "loss": 0.162,
      "step": 360
    },
    {
      "epoch": 0.8685162846803377,
      "eval_accuracy": 0.9606840676467204,
      "eval_loss": 0.1538599133491516,
      "eval_runtime": 401.6037,
      "eval_samples_per_second": 0.919,
      "eval_steps_per_second": 0.117,
      "step": 360
    },
    {
      "epoch": 0.8926417370325693,
      "grad_norm": 0.14280393590631585,
      "learning_rate": 0.00036652719665271967,
      "loss": 0.1743,
      "step": 370
    },
    {
      "epoch": 0.916767189384801,
      "grad_norm": 0.1761373590063179,
      "learning_rate": 0.00036234309623430963,
      "loss": 0.2,
      "step": 380
    },
    {
      "epoch": 0.9408926417370326,
      "grad_norm": 0.07884840352407207,
      "learning_rate": 0.0003581589958158996,
      "loss": 0.1177,
      "step": 390
    },
    {
      "epoch": 0.9408926417370326,
      "eval_accuracy": 0.9621479681849544,
      "eval_loss": 0.14951427280902863,
      "eval_runtime": 400.095,
      "eval_samples_per_second": 0.922,
      "eval_steps_per_second": 0.117,
      "step": 390
    },
    {
      "epoch": 0.9650180940892642,
      "grad_norm": 0.10601036641526983,
      "learning_rate": 0.00035397489539748956,
      "loss": 0.1493,
      "step": 400
    },
    {
      "epoch": 0.9891435464414958,
      "grad_norm": 0.13438968986656621,
      "learning_rate": 0.0003497907949790795,
      "loss": 0.1589,
      "step": 410
    },
    {
      "epoch": 1.0120627261761157,
      "grad_norm": 0.11429119825055273,
      "learning_rate": 0.00034560669456066943,
      "loss": 0.1276,
      "step": 420
    },
    {
      "epoch": 1.0120627261761157,
      "eval_accuracy": 0.963990042301987,
      "eval_loss": 0.14495952427387238,
      "eval_runtime": 405.3643,
      "eval_samples_per_second": 0.91,
      "eval_steps_per_second": 0.116,
      "step": 420
    },
    {
      "epoch": 1.0361881785283473,
      "grad_norm": 0.2704841163974856,
      "learning_rate": 0.0003414225941422594,
      "loss": 0.1067,
      "step": 430
    },
    {
      "epoch": 1.060313630880579,
      "grad_norm": 0.09327533568947116,
      "learning_rate": 0.0003372384937238494,
      "loss": 0.099,
      "step": 440
    },
    {
      "epoch": 1.0844390832328106,
      "grad_norm": 0.09863534693006064,
      "learning_rate": 0.0003330543933054394,
      "loss": 0.113,
      "step": 450
    },
    {
      "epoch": 1.0844390832328106,
      "eval_accuracy": 0.9625700639346573,
      "eval_loss": 0.14536267518997192,
      "eval_runtime": 398.3607,
      "eval_samples_per_second": 0.926,
      "eval_steps_per_second": 0.118,
      "step": 450
    },
    {
      "epoch": 1.1085645355850422,
      "grad_norm": 0.11503605061256633,
      "learning_rate": 0.0003288702928870293,
      "loss": 0.0957,
      "step": 460
    },
    {
      "epoch": 1.1326899879372738,
      "grad_norm": 0.17276708958246292,
      "learning_rate": 0.00032468619246861925,
      "loss": 0.1115,
      "step": 470
    },
    {
      "epoch": 1.1568154402895054,
      "grad_norm": 0.10569564290630626,
      "learning_rate": 0.0003205020920502092,
      "loss": 0.0844,
      "step": 480
    },
    {
      "epoch": 1.1568154402895054,
      "eval_accuracy": 0.9642106626092617,
      "eval_loss": 0.13868066668510437,
      "eval_runtime": 400.4906,
      "eval_samples_per_second": 0.921,
      "eval_steps_per_second": 0.117,
      "step": 480
    },
    {
      "epoch": 1.180940892641737,
      "grad_norm": 0.11136672433159371,
      "learning_rate": 0.0003163179916317992,
      "loss": 0.1161,
      "step": 490
    },
    {
      "epoch": 1.2050663449939687,
      "grad_norm": 0.14898870645702791,
      "learning_rate": 0.00031213389121338914,
      "loss": 0.1058,
      "step": 500
    },
    {
      "epoch": 1.2291917973462003,
      "grad_norm": 0.07640023682119597,
      "learning_rate": 0.0003079497907949791,
      "loss": 0.1035,
      "step": 510
    },
    {
      "epoch": 1.2291917973462003,
      "eval_accuracy": 0.9660363242586102,
      "eval_loss": 0.13527461886405945,
      "eval_runtime": 407.5369,
      "eval_samples_per_second": 0.905,
      "eval_steps_per_second": 0.115,
      "step": 510
    },
    {
      "epoch": 1.2533172496984317,
      "grad_norm": 0.09167723018930267,
      "learning_rate": 0.000303765690376569,
      "loss": 0.079,
      "step": 520
    },
    {
      "epoch": 1.2774427020506636,
      "grad_norm": 0.15463659567340096,
      "learning_rate": 0.000299581589958159,
      "loss": 0.0891,
      "step": 530
    },
    {
      "epoch": 1.301568154402895,
      "grad_norm": 0.21450751771826193,
      "learning_rate": 0.000295397489539749,
      "loss": 0.0903,
      "step": 540
    },
    {
      "epoch": 1.301568154402895,
      "eval_accuracy": 0.9660035559061296,
      "eval_loss": 0.13518056273460388,
      "eval_runtime": 404.8834,
      "eval_samples_per_second": 0.911,
      "eval_steps_per_second": 0.116,
      "step": 540
    },
    {
      "epoch": 1.3256936067551266,
      "grad_norm": 0.11254309553850446,
      "learning_rate": 0.00029121338912133896,
      "loss": 0.0924,
      "step": 550
    },
    {
      "epoch": 1.3498190591073582,
      "grad_norm": 0.1174557633067926,
      "learning_rate": 0.00028702928870292887,
      "loss": 0.1022,
      "step": 560
    },
    {
      "epoch": 1.3739445114595898,
      "grad_norm": 0.09842275495023206,
      "learning_rate": 0.00028284518828451883,
      "loss": 0.0927,
      "step": 570
    },
    {
      "epoch": 1.3739445114595898,
      "eval_accuracy": 0.9671573848321734,
      "eval_loss": 0.13157206773757935,
      "eval_runtime": 405.3445,
      "eval_samples_per_second": 0.91,
      "eval_steps_per_second": 0.116,
      "step": 570
    },
    {
      "epoch": 1.3980699638118215,
      "grad_norm": 0.1270496879595898,
      "learning_rate": 0.0002786610878661088,
      "loss": 0.0983,
      "step": 580
    },
    {
      "epoch": 1.422195416164053,
      "grad_norm": 0.11860896076212517,
      "learning_rate": 0.0002744769874476987,
      "loss": 0.0975,
      "step": 590
    },
    {
      "epoch": 1.4463208685162847,
      "grad_norm": 0.0941817384648685,
      "learning_rate": 0.0002702928870292887,
      "loss": 0.1017,
      "step": 600
    },
    {
      "epoch": 1.4463208685162847,
      "eval_accuracy": 0.9694876205658554,
      "eval_loss": 0.125931054353714,
      "eval_runtime": 400.9655,
      "eval_samples_per_second": 0.92,
      "eval_steps_per_second": 0.117,
      "step": 600
    },
    {
      "epoch": 1.4704463208685163,
      "grad_norm": 0.0979546267580835,
      "learning_rate": 0.0002661087866108787,
      "loss": 0.0864,
      "step": 610
    },
    {
      "epoch": 1.494571773220748,
      "grad_norm": 0.04978373728280121,
      "learning_rate": 0.00026192468619246865,
      "loss": 0.0772,
      "step": 620
    },
    {
      "epoch": 1.5186972255729794,
      "grad_norm": 0.11349667484083058,
      "learning_rate": 0.00025774058577405855,
      "loss": 0.0805,
      "step": 630
    },
    {
      "epoch": 1.5186972255729794,
      "eval_accuracy": 0.9690919035684665,
      "eval_loss": 0.12952929735183716,
      "eval_runtime": 407.0582,
      "eval_samples_per_second": 0.907,
      "eval_steps_per_second": 0.115,
      "step": 630
    },
    {
      "epoch": 1.5428226779252112,
      "grad_norm": 0.08738366633979745,
      "learning_rate": 0.0002535564853556485,
      "loss": 0.0947,
      "step": 640
    },
    {
      "epoch": 1.5669481302774426,
      "grad_norm": 0.07525859699508217,
      "learning_rate": 0.0002493723849372385,
      "loss": 0.1102,
      "step": 650
    },
    {
      "epoch": 1.5910735826296745,
      "grad_norm": 0.11035985461288858,
      "learning_rate": 0.00024518828451882844,
      "loss": 0.1307,
      "step": 660
    },
    {
      "epoch": 1.5910735826296745,
      "eval_accuracy": 0.9708699913097606,
      "eval_loss": 0.12111840397119522,
      "eval_runtime": 402.2427,
      "eval_samples_per_second": 0.917,
      "eval_steps_per_second": 0.117,
      "step": 660
    },
    {
      "epoch": 1.6151990349819059,
      "grad_norm": 0.0846832605738454,
      "learning_rate": 0.0002410041841004184,
      "loss": 0.0738,
      "step": 670
    },
    {
      "epoch": 1.6393244873341375,
      "grad_norm": 0.08888153295105361,
      "learning_rate": 0.0002368200836820084,
      "loss": 0.0777,
      "step": 680
    },
    {
      "epoch": 1.6634499396863691,
      "grad_norm": 0.08800292216830932,
      "learning_rate": 0.00023263598326359833,
      "loss": 0.0863,
      "step": 690
    },
    {
      "epoch": 1.6634499396863691,
      "eval_accuracy": 0.9711017910206424,
      "eval_loss": 0.11842313408851624,
      "eval_runtime": 405.1616,
      "eval_samples_per_second": 0.911,
      "eval_steps_per_second": 0.116,
      "step": 690
    },
    {
      "epoch": 1.6875753920386007,
      "grad_norm": 0.11264895254963977,
      "learning_rate": 0.00022845188284518827,
      "loss": 0.0822,
      "step": 700
    },
    {
      "epoch": 1.7117008443908324,
      "grad_norm": 0.053826072481400256,
      "learning_rate": 0.00022426778242677826,
      "loss": 0.0776,
      "step": 710
    },
    {
      "epoch": 1.7358262967430638,
      "grad_norm": 0.10176323860970168,
      "learning_rate": 0.0002200836820083682,
      "loss": 0.065,
      "step": 720
    },
    {
      "epoch": 1.7358262967430638,
      "eval_accuracy": 0.971402633824704,
      "eval_loss": 0.11691941320896149,
      "eval_runtime": 408.0363,
      "eval_samples_per_second": 0.904,
      "eval_steps_per_second": 0.115,
      "step": 720
    },
    {
      "epoch": 1.7599517490952956,
      "grad_norm": 0.14859706282251087,
      "learning_rate": 0.0002158995815899582,
      "loss": 0.0947,
      "step": 730
    },
    {
      "epoch": 1.784077201447527,
      "grad_norm": 0.10065034423169357,
      "learning_rate": 0.00021171548117154812,
      "loss": 0.0963,
      "step": 740
    },
    {
      "epoch": 1.8082026537997589,
      "grad_norm": 0.07196724349941265,
      "learning_rate": 0.00020753138075313806,
      "loss": 0.0899,
      "step": 750
    },
    {
      "epoch": 1.8082026537997589,
      "eval_accuracy": 0.9723688383710148,
      "eval_loss": 0.11116459220647812,
      "eval_runtime": 401.4182,
      "eval_samples_per_second": 0.919,
      "eval_steps_per_second": 0.117,
      "step": 750
    },
    {
      "epoch": 1.8323281061519903,
      "grad_norm": 0.07836756871728362,
      "learning_rate": 0.00020334728033472805,
      "loss": 0.0845,
      "step": 760
    },
    {
      "epoch": 1.8564535585042221,
      "grad_norm": 0.08960999009516811,
      "learning_rate": 0.00019916317991631799,
      "loss": 0.0836,
      "step": 770
    },
    {
      "epoch": 1.8805790108564535,
      "grad_norm": 0.08193539772677307,
      "learning_rate": 0.00019497907949790795,
      "loss": 0.0736,
      "step": 780
    },
    {
      "epoch": 1.8805790108564535,
      "eval_accuracy": 0.9734024748833037,
      "eval_loss": 0.10833366960287094,
      "eval_runtime": 405.6636,
      "eval_samples_per_second": 0.91,
      "eval_steps_per_second": 0.116,
      "step": 780
    },
    {
      "epoch": 1.9047044632086851,
      "grad_norm": 0.06908567074998409,
      "learning_rate": 0.0001907949790794979,
      "loss": 0.0805,
      "step": 790
    },
    {
      "epoch": 1.9288299155609168,
      "grad_norm": 0.11126793777990279,
      "learning_rate": 0.00018661087866108788,
      "loss": 0.0686,
      "step": 800
    },
    {
      "epoch": 1.9529553679131484,
      "grad_norm": 0.07286700968797473,
      "learning_rate": 0.00018242677824267784,
      "loss": 0.0772,
      "step": 810
    },
    {
      "epoch": 1.9529553679131484,
      "eval_accuracy": 0.9727764523869131,
      "eval_loss": 0.1093606948852539,
      "eval_runtime": 400.9048,
      "eval_samples_per_second": 0.92,
      "eval_steps_per_second": 0.117,
      "step": 810
    },
    {
      "epoch": 1.97708082026538,
      "grad_norm": 0.09637695938928079,
      "learning_rate": 0.00017824267782426778,
      "loss": 0.0651,
      "step": 820
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.05902656978962385,
      "learning_rate": 0.00017405857740585774,
      "loss": 0.0809,
      "step": 830
    },
    {
      "epoch": 2.0241254523522314,
      "grad_norm": 0.10392170028551989,
      "learning_rate": 0.0001698744769874477,
      "loss": 0.047,
      "step": 840
    },
    {
      "epoch": 2.0241254523522314,
      "eval_accuracy": 0.973407946955895,
      "eval_loss": 0.11182070523500443,
      "eval_runtime": 402.0193,
      "eval_samples_per_second": 0.918,
      "eval_steps_per_second": 0.117,
      "step": 840
    },
    {
      "epoch": 2.0482509047044632,
      "grad_norm": 0.1272900478377503,
      "learning_rate": 0.00016569037656903767,
      "loss": 0.0566,
      "step": 850
    },
    {
      "epoch": 2.0723763570566947,
      "grad_norm": 0.12975815140383112,
      "learning_rate": 0.00016150627615062763,
      "loss": 0.0524,
      "step": 860
    },
    {
      "epoch": 2.0965018094089265,
      "grad_norm": 0.08241098466991122,
      "learning_rate": 0.00015732217573221756,
      "loss": 0.0389,
      "step": 870
    },
    {
      "epoch": 2.0965018094089265,
      "eval_accuracy": 0.9735341912064753,
      "eval_loss": 0.11427704989910126,
      "eval_runtime": 402.9947,
      "eval_samples_per_second": 0.916,
      "eval_steps_per_second": 0.117,
      "step": 870
    },
    {
      "epoch": 2.120627261761158,
      "grad_norm": 0.07085785092981728,
      "learning_rate": 0.00015313807531380753,
      "loss": 0.046,
      "step": 880
    },
    {
      "epoch": 2.1447527141133897,
      "grad_norm": 0.061480856518437405,
      "learning_rate": 0.0001489539748953975,
      "loss": 0.0451,
      "step": 890
    },
    {
      "epoch": 2.168878166465621,
      "grad_norm": 0.08207170391672733,
      "learning_rate": 0.00014476987447698745,
      "loss": 0.0519,
      "step": 900
    },
    {
      "epoch": 2.168878166465621,
      "eval_accuracy": 0.9741671480335237,
      "eval_loss": 0.11114877462387085,
      "eval_runtime": 398.2764,
      "eval_samples_per_second": 0.926,
      "eval_steps_per_second": 0.118,
      "step": 900
    },
    {
      "epoch": 2.193003618817853,
      "grad_norm": 0.09830708838871588,
      "learning_rate": 0.00014058577405857742,
      "loss": 0.037,
      "step": 910
    },
    {
      "epoch": 2.2171290711700844,
      "grad_norm": 0.11119247754282878,
      "learning_rate": 0.00013640167364016735,
      "loss": 0.069,
      "step": 920
    },
    {
      "epoch": 2.2412545235223162,
      "grad_norm": 0.05909682523831251,
      "learning_rate": 0.00013221757322175732,
      "loss": 0.0417,
      "step": 930
    },
    {
      "epoch": 2.2412545235223162,
      "eval_accuracy": 0.9751163341221841,
      "eval_loss": 0.1100468710064888,
      "eval_runtime": 400.8085,
      "eval_samples_per_second": 0.921,
      "eval_steps_per_second": 0.117,
      "step": 930
    },
    {
      "epoch": 2.2653799758745476,
      "grad_norm": 0.11124083437458757,
      "learning_rate": 0.00012803347280334728,
      "loss": 0.0494,
      "step": 940
    },
    {
      "epoch": 2.2895054282267795,
      "grad_norm": 0.10089762753720484,
      "learning_rate": 0.00012384937238493724,
      "loss": 0.0437,
      "step": 950
    },
    {
      "epoch": 2.313630880579011,
      "grad_norm": 0.10495311174294904,
      "learning_rate": 0.0001196652719665272,
      "loss": 0.0485,
      "step": 960
    },
    {
      "epoch": 2.313630880579011,
      "eval_accuracy": 0.9747861335000979,
      "eval_loss": 0.10849116742610931,
      "eval_runtime": 402.7977,
      "eval_samples_per_second": 0.916,
      "eval_steps_per_second": 0.117,
      "step": 960
    },
    {
      "epoch": 2.3377563329312423,
      "grad_norm": 0.07416800940523141,
      "learning_rate": 0.00011548117154811716,
      "loss": 0.0613,
      "step": 970
    },
    {
      "epoch": 2.361881785283474,
      "grad_norm": 0.08870151624144841,
      "learning_rate": 0.00011129707112970712,
      "loss": 0.0395,
      "step": 980
    },
    {
      "epoch": 2.3860072376357055,
      "grad_norm": 0.11291586342307855,
      "learning_rate": 0.00010711297071129707,
      "loss": 0.0539,
      "step": 990
    },
    {
      "epoch": 2.3860072376357055,
      "eval_accuracy": 0.9757843658817348,
      "eval_loss": 0.10554877668619156,
      "eval_runtime": 401.1352,
      "eval_samples_per_second": 0.92,
      "eval_steps_per_second": 0.117,
      "step": 990
    },
    {
      "epoch": 2.4101326899879374,
      "grad_norm": 0.07638450695123498,
      "learning_rate": 0.00010292887029288703,
      "loss": 0.0401,
      "step": 1000
    },
    {
      "epoch": 2.434258142340169,
      "grad_norm": 0.04155570662195518,
      "learning_rate": 9.874476987447698e-05,
      "loss": 0.0414,
      "step": 1010
    },
    {
      "epoch": 2.4583835946924006,
      "grad_norm": 0.12536376696276827,
      "learning_rate": 9.456066945606695e-05,
      "loss": 0.031,
      "step": 1020
    },
    {
      "epoch": 2.4583835946924006,
      "eval_accuracy": 0.9759608749444663,
      "eval_loss": 0.10678882151842117,
      "eval_runtime": 395.2034,
      "eval_samples_per_second": 0.934,
      "eval_steps_per_second": 0.119,
      "step": 1020
    },
    {
      "epoch": 2.482509047044632,
      "grad_norm": 0.05853715998683253,
      "learning_rate": 9.037656903765691e-05,
      "loss": 0.0434,
      "step": 1030
    },
    {
      "epoch": 2.5066344993968634,
      "grad_norm": 0.15899926330528433,
      "learning_rate": 8.619246861924686e-05,
      "loss": 0.0432,
      "step": 1040
    },
    {
      "epoch": 2.5307599517490953,
      "grad_norm": 0.06334825797446753,
      "learning_rate": 8.200836820083682e-05,
      "loss": 0.0367,
      "step": 1050
    },
    {
      "epoch": 2.5307599517490953,
      "eval_accuracy": 0.9760688047707028,
      "eval_loss": 0.10758133232593536,
      "eval_runtime": 401.3401,
      "eval_samples_per_second": 0.919,
      "eval_steps_per_second": 0.117,
      "step": 1050
    },
    {
      "epoch": 2.554885404101327,
      "grad_norm": 0.09436564895529882,
      "learning_rate": 7.782426778242679e-05,
      "loss": 0.0417,
      "step": 1060
    },
    {
      "epoch": 2.5790108564535585,
      "grad_norm": 0.12307695043858644,
      "learning_rate": 7.364016736401674e-05,
      "loss": 0.0361,
      "step": 1070
    },
    {
      "epoch": 2.60313630880579,
      "grad_norm": 0.07128798743449313,
      "learning_rate": 6.94560669456067e-05,
      "loss": 0.0294,
      "step": 1080
    },
    {
      "epoch": 2.60313630880579,
      "eval_accuracy": 0.9773206678217247,
      "eval_loss": 0.10543525964021683,
      "eval_runtime": 395.6037,
      "eval_samples_per_second": 0.933,
      "eval_steps_per_second": 0.119,
      "step": 1080
    },
    {
      "epoch": 2.627261761158022,
      "grad_norm": 0.07856660485495531,
      "learning_rate": 6.527196652719665e-05,
      "loss": 0.0311,
      "step": 1090
    },
    {
      "epoch": 2.651387213510253,
      "grad_norm": 0.06747659225158464,
      "learning_rate": 6.108786610878661e-05,
      "loss": 0.0388,
      "step": 1100
    },
    {
      "epoch": 2.675512665862485,
      "grad_norm": 0.14828084051336715,
      "learning_rate": 5.690376569037657e-05,
      "loss": 0.0329,
      "step": 1110
    },
    {
      "epoch": 2.675512665862485,
      "eval_accuracy": 0.9770663114662669,
      "eval_loss": 0.10486458986997604,
      "eval_runtime": 402.8414,
      "eval_samples_per_second": 0.916,
      "eval_steps_per_second": 0.117,
      "step": 1110
    },
    {
      "epoch": 2.6996381182147164,
      "grad_norm": 0.10392026207749133,
      "learning_rate": 5.2719665271966525e-05,
      "loss": 0.0328,
      "step": 1120
    },
    {
      "epoch": 2.7237635705669483,
      "grad_norm": 0.07730371410587372,
      "learning_rate": 4.853556485355649e-05,
      "loss": 0.0347,
      "step": 1130
    },
    {
      "epoch": 2.7478890229191797,
      "grad_norm": 0.03784134955733082,
      "learning_rate": 4.4351464435146445e-05,
      "loss": 0.0358,
      "step": 1140
    },
    {
      "epoch": 2.7478890229191797,
      "eval_accuracy": 0.9772854475598465,
      "eval_loss": 0.10271470993757248,
      "eval_runtime": 400.9018,
      "eval_samples_per_second": 0.92,
      "eval_steps_per_second": 0.117,
      "step": 1140
    },
    {
      "epoch": 2.772014475271411,
      "grad_norm": 0.09653218696173602,
      "learning_rate": 4.01673640167364e-05,
      "loss": 0.0301,
      "step": 1150
    },
    {
      "epoch": 2.796139927623643,
      "grad_norm": 0.09744072771477266,
      "learning_rate": 3.5983263598326365e-05,
      "loss": 0.0442,
      "step": 1160
    },
    {
      "epoch": 2.820265379975875,
      "grad_norm": 0.07884255949164899,
      "learning_rate": 3.1799163179916314e-05,
      "loss": 0.0321,
      "step": 1170
    },
    {
      "epoch": 2.820265379975875,
      "eval_accuracy": 0.9776449962552778,
      "eval_loss": 0.10332793742418289,
      "eval_runtime": 403.7568,
      "eval_samples_per_second": 0.914,
      "eval_steps_per_second": 0.116,
      "step": 1170
    },
    {
      "epoch": 2.844390832328106,
      "grad_norm": 0.08277855244189673,
      "learning_rate": 2.7615062761506278e-05,
      "loss": 0.0322,
      "step": 1180
    },
    {
      "epoch": 2.8685162846803376,
      "grad_norm": 0.11858206842827564,
      "learning_rate": 2.3430962343096234e-05,
      "loss": 0.054,
      "step": 1190
    },
    {
      "epoch": 2.8926417370325694,
      "grad_norm": 0.09689243001487692,
      "learning_rate": 1.9246861924686194e-05,
      "loss": 0.0337,
      "step": 1200
    },
    {
      "epoch": 2.8926417370325694,
      "eval_accuracy": 0.9776567900152591,
      "eval_loss": 0.10333515703678131,
      "eval_runtime": 403.7644,
      "eval_samples_per_second": 0.914,
      "eval_steps_per_second": 0.116,
      "step": 1200
    },
    {
      "epoch": 2.916767189384801,
      "grad_norm": 0.04818702941528732,
      "learning_rate": 1.506276150627615e-05,
      "loss": 0.0401,
      "step": 1210
    },
    {
      "epoch": 2.9408926417370327,
      "grad_norm": 0.10805288272206603,
      "learning_rate": 1.0878661087866109e-05,
      "loss": 0.033,
      "step": 1220
    },
    {
      "epoch": 2.965018094089264,
      "grad_norm": 0.07483254719666964,
      "learning_rate": 6.694560669456067e-06,
      "loss": 0.0456,
      "step": 1230
    },
    {
      "epoch": 2.965018094089264,
      "eval_accuracy": 0.9780354933237786,
      "eval_loss": 0.10221888870000839,
      "eval_runtime": 401.8718,
      "eval_samples_per_second": 0.918,
      "eval_steps_per_second": 0.117,
      "step": 1230
    },
    {
      "epoch": 2.989143546441496,
      "grad_norm": 0.06015457354275012,
      "learning_rate": 2.510460251046025e-06,
      "loss": 0.0244,
      "step": 1240
    },
    {
      "epoch": 3.0,
      "step": 1245,
      "total_flos": 84124459565056.0,
      "train_loss": 0.11350005203222176,
      "train_runtime": 83292.593,
      "train_samples_per_second": 0.119,
      "train_steps_per_second": 0.015
    }
  ],
  "logging_steps": 10,
  "max_steps": 1245,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 30,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 84124459565056.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}