{
  "best_metric": 2.176323175430298,
  "best_model_checkpoint": "../../saves/Llama3-8B-Chinese-Chat/lora/sft/checkpoint-2400",
  "epoch": 9.481481481481481,
  "eval_steps": 400,
  "global_step": 2400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04,
      "grad_norm": 0.6577214002609253,
      "learning_rate": 2.5e-05,
      "loss": 2.8395,
      "step": 10
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.9053373336791992,
      "learning_rate": 5e-05,
      "loss": 2.7212,
      "step": 20
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.8605895042419434,
      "learning_rate": 4.9998041801805566e-05,
      "loss": 2.6919,
      "step": 30
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.9140876531600952,
      "learning_rate": 4.999293114538139e-05,
      "loss": 2.615,
      "step": 40
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.8557806015014648,
      "learning_rate": 4.998353314622318e-05,
      "loss": 2.6138,
      "step": 50
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.9230502843856812,
      "learning_rate": 4.997022133030516e-05,
      "loss": 2.5346,
      "step": 60
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.0639443397521973,
      "learning_rate": 4.9952997783001254e-05,
      "loss": 2.4718,
      "step": 70
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.0025442838668823,
      "learning_rate": 4.9931865202480996e-05,
      "loss": 2.5162,
      "step": 80
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.0188956260681152,
      "learning_rate": 4.990682689928687e-05,
      "loss": 2.4471,
      "step": 90
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.122934341430664,
      "learning_rate": 4.9877886795815685e-05,
      "loss": 2.5208,
      "step": 100
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.2770124673843384,
      "learning_rate": 4.98450494257041e-05,
      "loss": 2.4132,
      "step": 110
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5353667736053467,
      "learning_rate": 4.980831993311844e-05,
      "loss": 2.4551,
      "step": 120
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.3800839185714722,
      "learning_rate": 4.976770407194877e-05,
      "loss": 2.4097,
      "step": 130
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.532073736190796,
      "learning_rate": 4.972320820490759e-05,
      "loss": 2.4367,
      "step": 140
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.7717530727386475,
      "learning_rate": 4.967483930253302e-05,
      "loss": 2.4128,
      "step": 150
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.92471444606781,
      "learning_rate": 4.962260494209683e-05,
      "loss": 2.4274,
      "step": 160
    },
    {
      "epoch": 0.67,
      "grad_norm": 2.059626340866089,
      "learning_rate": 4.9566513306417444e-05,
      "loss": 2.4485,
      "step": 170
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.4040436744689941,
      "learning_rate": 4.950657318257805e-05,
      "loss": 2.4313,
      "step": 180
    },
    {
      "epoch": 0.75,
      "grad_norm": 6.483478546142578,
      "learning_rate": 4.944279396055003e-05,
      "loss": 2.4243,
      "step": 190
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.516174554824829,
      "learning_rate": 4.937518563172196e-05,
      "loss": 2.3765,
      "step": 200
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.4100685119628906,
      "learning_rate": 4.930375878733445e-05,
      "loss": 2.4035,
      "step": 210
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.6589595079421997,
      "learning_rate": 4.922852461682093e-05,
      "loss": 2.328,
      "step": 220
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.4953701496124268,
      "learning_rate": 4.9149494906054716e-05,
      "loss": 2.3684,
      "step": 230
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.4060596227645874,
      "learning_rate": 4.906668203550279e-05,
      "loss": 2.3472,
      "step": 240
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.474617838859558,
      "learning_rate": 4.8980098978286215e-05,
      "loss": 2.3998,
      "step": 250
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.42411208152771,
      "learning_rate": 4.888975929814792e-05,
      "loss": 2.3639,
      "step": 260
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.62990140914917,
      "learning_rate": 4.880525335812728e-05,
      "loss": 2.3325,
      "step": 270
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.5688358545303345,
      "learning_rate": 4.8707815568466236e-05,
      "loss": 2.3001,
      "step": 280
    },
    {
      "epoch": 1.15,
      "grad_norm": 3.7384893894195557,
      "learning_rate": 4.860666381067398e-05,
      "loss": 2.3522,
      "step": 290
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.44722318649292,
      "learning_rate": 4.850181393076568e-05,
      "loss": 2.3296,
      "step": 300
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.6258642673492432,
      "learning_rate": 4.839328235408895e-05,
      "loss": 2.3827,
      "step": 310
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.72386634349823,
      "learning_rate": 4.8281086082750825e-05,
      "loss": 2.2641,
      "step": 320
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.783467411994934,
      "learning_rate": 4.816524269295416e-05,
      "loss": 2.278,
      "step": 330
    },
    {
      "epoch": 1.34,
      "grad_norm": 4.662048816680908,
      "learning_rate": 4.804577033224429e-05,
      "loss": 2.264,
      "step": 340
    },
    {
      "epoch": 1.38,
      "grad_norm": 3.12662410736084,
      "learning_rate": 4.7922687716666105e-05,
      "loss": 2.2712,
      "step": 350
    },
    {
      "epoch": 1.42,
      "grad_norm": 2.0605392456054688,
      "learning_rate": 4.779601412783206e-05,
      "loss": 2.2865,
      "step": 360
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.6488828659057617,
      "learning_rate": 4.766576940990157e-05,
      "loss": 2.334,
      "step": 370
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.8102229833602905,
      "learning_rate": 4.7531973966472376e-05,
      "loss": 2.3421,
      "step": 380
    },
    {
      "epoch": 1.54,
      "grad_norm": 2.01405668258667,
      "learning_rate": 4.7394648757384125e-05,
      "loss": 2.3526,
      "step": 390
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.5997238159179688,
      "learning_rate": 4.7253815295434934e-05,
      "loss": 2.2611,
      "step": 400
    },
    {
      "epoch": 1.58,
      "eval_loss": 2.21992826461792,
      "eval_runtime": 142.6668,
      "eval_samples_per_second": 6.308,
      "eval_steps_per_second": 3.154,
      "step": 400
    },
    {
      "epoch": 1.62,
      "grad_norm": 2.2114603519439697,
      "learning_rate": 4.710949564301128e-05,
      "loss": 2.4253,
      "step": 410
    },
    {
      "epoch": 1.66,
      "grad_norm": 2.8670096397399902,
      "learning_rate": 4.6961712408631795e-05,
      "loss": 2.3267,
      "step": 420
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.991642951965332,
      "learning_rate": 4.681048874340548e-05,
      "loss": 2.3256,
      "step": 430
    },
    {
      "epoch": 1.74,
      "grad_norm": 4.117753505706787,
      "learning_rate": 4.6655848337405e-05,
      "loss": 2.2686,
      "step": 440
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.650153636932373,
      "learning_rate": 4.649781541595547e-05,
      "loss": 2.2634,
      "step": 450
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.9907420873641968,
      "learning_rate": 4.633641473583939e-05,
      "loss": 2.3402,
      "step": 460
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.8381316661834717,
      "learning_rate": 4.6171671581418395e-05,
      "loss": 2.381,
      "step": 470
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.8253530263900757,
      "learning_rate": 4.600361176067229e-05,
      "loss": 2.2451,
      "step": 480
    },
    {
      "epoch": 1.94,
      "grad_norm": 2.263423442840576,
      "learning_rate": 4.583226160115608e-05,
      "loss": 2.32,
      "step": 490
    },
    {
      "epoch": 1.98,
      "grad_norm": 3.391731023788452,
      "learning_rate": 4.56576479458756e-05,
      "loss": 2.2602,
      "step": 500
    },
    {
      "epoch": 2.01,
      "grad_norm": 2.6076712608337402,
      "learning_rate": 4.5479798149082406e-05,
      "loss": 2.242,
      "step": 510
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.9152929782867432,
      "learning_rate": 4.529874007198859e-05,
      "loss": 2.259,
      "step": 520
    },
    {
      "epoch": 2.09,
      "grad_norm": 2.9343013763427734,
      "learning_rate": 4.5114502078402103e-05,
      "loss": 2.2679,
      "step": 530
    },
    {
      "epoch": 2.13,
      "grad_norm": 2.068617105484009,
      "learning_rate": 4.4927113030283475e-05,
      "loss": 2.2661,
      "step": 540
    },
    {
      "epoch": 2.17,
      "grad_norm": 7.2981953620910645,
      "learning_rate": 4.473660228322434e-05,
      "loss": 2.3253,
      "step": 550
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.60201096534729,
      "learning_rate": 4.454299968184879e-05,
      "loss": 2.3142,
      "step": 560
    },
    {
      "epoch": 2.25,
      "grad_norm": 3.731210231781006,
      "learning_rate": 4.4346335555137975e-05,
      "loss": 2.3323,
      "step": 570
    },
    {
      "epoch": 2.29,
      "grad_norm": 3.097869634628296,
      "learning_rate": 4.414664071167892e-05,
      "loss": 2.2346,
      "step": 580
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.9533767700195312,
      "learning_rate": 4.394394643483818e-05,
      "loss": 2.3276,
      "step": 590
    },
    {
      "epoch": 2.37,
      "grad_norm": 4.110910892486572,
      "learning_rate": 4.373828447786111e-05,
      "loss": 2.2844,
      "step": 600
    },
    {
      "epoch": 2.41,
      "grad_norm": 2.3068130016326904,
      "learning_rate": 4.352968705889753e-05,
      "loss": 2.3205,
      "step": 610
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.739564061164856,
      "learning_rate": 4.331818685595458e-05,
      "loss": 2.2218,
      "step": 620
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.999505877494812,
      "learning_rate": 4.3103817001777494e-05,
      "loss": 2.2117,
      "step": 630
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.6982237100601196,
      "learning_rate": 4.2886611078659187e-05,
      "loss": 2.174,
      "step": 640
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.7287157773971558,
      "learning_rate": 4.266660311317938e-05,
      "loss": 2.2478,
      "step": 650
    },
    {
      "epoch": 2.61,
      "grad_norm": 2.1112897396087646,
      "learning_rate": 4.244382757087413e-05,
      "loss": 2.2961,
      "step": 660
    },
    {
      "epoch": 2.65,
      "grad_norm": 1.901487946510315,
      "learning_rate": 4.221831935083662e-05,
      "loss": 2.1506,
      "step": 670
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.8849854469299316,
      "learning_rate": 4.199011378024998e-05,
      "loss": 2.3471,
      "step": 680
    },
    {
      "epoch": 2.73,
      "grad_norm": 1.8044487237930298,
      "learning_rate": 4.175924660885314e-05,
      "loss": 2.2516,
      "step": 690
    },
    {
      "epoch": 2.77,
      "grad_norm": 6.139608383178711,
      "learning_rate": 4.152575400334032e-05,
      "loss": 2.2227,
      "step": 700
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.8950961828231812,
      "learning_rate": 4.1289672541695434e-05,
      "loss": 2.2474,
      "step": 710
    },
    {
      "epoch": 2.84,
      "grad_norm": 3.646942615509033,
      "learning_rate": 4.105103920746182e-05,
      "loss": 2.2239,
      "step": 720
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.5586037635803223,
      "learning_rate": 4.0809891383948616e-05,
      "loss": 2.3488,
      "step": 730
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.7886706590652466,
      "learning_rate": 4.056626684837443e-05,
      "loss": 2.2088,
      "step": 740
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.7146466970443726,
      "learning_rate": 4.0320203765949325e-05,
      "loss": 2.2672,
      "step": 750
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.8052889108657837,
      "learning_rate": 4.007174068389599e-05,
      "loss": 2.2723,
      "step": 760
    },
    {
      "epoch": 3.04,
      "grad_norm": 3.6868667602539062,
      "learning_rate": 3.982091652541111e-05,
      "loss": 2.261,
      "step": 770
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.8512885570526123,
      "learning_rate": 3.956777058356785e-05,
      "loss": 2.1709,
      "step": 780
    },
    {
      "epoch": 3.12,
      "grad_norm": 2.124035596847534,
      "learning_rate": 3.931234251516029e-05,
      "loss": 2.1759,
      "step": 790
    },
    {
      "epoch": 3.16,
      "grad_norm": 2.077178478240967,
      "learning_rate": 3.905467233449103e-05,
      "loss": 2.3188,
      "step": 800
    },
    {
      "epoch": 3.16,
      "eval_loss": 2.1877169609069824,
      "eval_runtime": 142.5615,
      "eval_samples_per_second": 6.313,
      "eval_steps_per_second": 3.157,
      "step": 800
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.8644123077392578,
      "learning_rate": 3.879480040710266e-05,
      "loss": 2.2546,
      "step": 810
    },
    {
      "epoch": 3.24,
      "grad_norm": 2.230581521987915,
      "learning_rate": 3.8532767443454316e-05,
      "loss": 2.3114,
      "step": 820
    },
    {
      "epoch": 3.28,
      "grad_norm": 2.7963311672210693,
      "learning_rate": 3.826861449254409e-05,
      "loss": 2.1624,
      "step": 830
    },
    {
      "epoch": 3.32,
      "grad_norm": 2.0492663383483887,
      "learning_rate": 3.8002382935478506e-05,
      "loss": 2.2222,
      "step": 840
    },
    {
      "epoch": 3.36,
      "grad_norm": 2.477332592010498,
      "learning_rate": 3.773411447898992e-05,
      "loss": 2.2765,
      "step": 850
    },
    {
      "epoch": 3.4,
      "grad_norm": 2.9290037155151367,
      "learning_rate": 3.7463851148902895e-05,
      "loss": 2.1336,
      "step": 860
    },
    {
      "epoch": 3.44,
      "grad_norm": 2.215397357940674,
      "learning_rate": 3.7191635283550636e-05,
      "loss": 2.2759,
      "step": 870
    },
    {
      "epoch": 3.48,
      "grad_norm": 2.4170403480529785,
      "learning_rate": 3.691750952714242e-05,
      "loss": 2.2636,
      "step": 880
    },
    {
      "epoch": 3.52,
      "grad_norm": 3.395965337753296,
      "learning_rate": 3.664151682308315e-05,
      "loss": 2.2049,
      "step": 890
    },
    {
      "epoch": 3.56,
      "grad_norm": 15.315768241882324,
      "learning_rate": 3.636370040724599e-05,
      "loss": 2.1917,
      "step": 900
    },
    {
      "epoch": 3.6,
      "grad_norm": 2.1479156017303467,
      "learning_rate": 3.608410380119927e-05,
      "loss": 2.2352,
      "step": 910
    },
    {
      "epoch": 3.63,
      "grad_norm": 1.9286900758743286,
      "learning_rate": 3.580277080538853e-05,
      "loss": 2.2011,
      "step": 920
    },
    {
      "epoch": 3.67,
      "grad_norm": 2.303708791732788,
      "learning_rate": 3.55197454922749e-05,
      "loss": 2.1426,
      "step": 930
    },
    {
      "epoch": 3.71,
      "grad_norm": 2.3887522220611572,
      "learning_rate": 3.523507219943096e-05,
      "loss": 2.158,
      "step": 940
    },
    {
      "epoch": 3.75,
      "grad_norm": 6.511980056762695,
      "learning_rate": 3.494879552259496e-05,
      "loss": 2.305,
      "step": 950
    },
    {
      "epoch": 3.79,
      "grad_norm": 2.579639434814453,
      "learning_rate": 3.466096030868462e-05,
      "loss": 2.3205,
      "step": 960
    },
    {
      "epoch": 3.83,
      "grad_norm": 1.789423942565918,
      "learning_rate": 3.437161164877165e-05,
      "loss": 2.1844,
      "step": 970
    },
    {
      "epoch": 3.87,
      "grad_norm": 1.7820065021514893,
      "learning_rate": 3.408079487101791e-05,
      "loss": 2.2388,
      "step": 980
    },
    {
      "epoch": 3.91,
      "grad_norm": 2.107856512069702,
      "learning_rate": 3.3788555533574535e-05,
      "loss": 2.266,
      "step": 990
    },
    {
      "epoch": 3.95,
      "grad_norm": 2.2462568283081055,
      "learning_rate": 3.349493941744496e-05,
      "loss": 2.2603,
      "step": 1000
    },
    {
      "epoch": 3.99,
      "grad_norm": 1.9653023481369019,
      "learning_rate": 3.319999251931305e-05,
      "loss": 2.2055,
      "step": 1010
    },
    {
      "epoch": 4.03,
      "grad_norm": 2.0476274490356445,
      "learning_rate": 3.29037610443375e-05,
      "loss": 2.2843,
      "step": 1020
    },
    {
      "epoch": 4.07,
      "grad_norm": 2.2384870052337646,
      "learning_rate": 3.260629139891344e-05,
      "loss": 2.1942,
      "step": 1030
    },
    {
      "epoch": 4.11,
      "grad_norm": 1.923712134361267,
      "learning_rate": 3.230763018340269e-05,
      "loss": 2.2639,
      "step": 1040
    },
    {
      "epoch": 4.15,
      "grad_norm": 2.2221322059631348,
      "learning_rate": 3.200782418483348e-05,
      "loss": 2.1726,
      "step": 1050
    },
    {
      "epoch": 4.19,
      "grad_norm": 2.22153377532959,
      "learning_rate": 3.1706920369571006e-05,
      "loss": 2.2761,
      "step": 1060
    },
    {
      "epoch": 4.23,
      "grad_norm": 4.4850616455078125,
      "learning_rate": 3.1404965875959914e-05,
      "loss": 2.1937,
      "step": 1070
    },
    {
      "epoch": 4.27,
      "grad_norm": 2.2232580184936523,
      "learning_rate": 3.110200800693972e-05,
      "loss": 2.2333,
      "step": 1080
    },
    {
      "epoch": 4.31,
      "grad_norm": 2.1202492713928223,
      "learning_rate": 3.0798094222634597e-05,
      "loss": 2.2239,
      "step": 1090
    },
    {
      "epoch": 4.35,
      "grad_norm": 2.625195264816284,
      "learning_rate": 3.0493272132918445e-05,
      "loss": 2.1311,
      "step": 1100
    },
    {
      "epoch": 4.39,
      "grad_norm": 2.0367236137390137,
      "learning_rate": 3.0187589489956503e-05,
      "loss": 2.2407,
      "step": 1110
    },
    {
      "epoch": 4.42,
      "grad_norm": 2.249363422393799,
      "learning_rate": 2.9881094180724757e-05,
      "loss": 2.1846,
      "step": 1120
    },
    {
      "epoch": 4.46,
      "grad_norm": 2.561105489730835,
      "learning_rate": 2.9573834219508084e-05,
      "loss": 2.0067,
      "step": 1130
    },
    {
      "epoch": 4.5,
      "grad_norm": 2.33273983001709,
      "learning_rate": 2.9265857740378576e-05,
      "loss": 2.2424,
      "step": 1140
    },
    {
      "epoch": 4.54,
      "grad_norm": 2.4719011783599854,
      "learning_rate": 2.8957212989655092e-05,
      "loss": 2.1779,
      "step": 1150
    },
    {
      "epoch": 4.58,
      "grad_norm": 2.366910457611084,
      "learning_rate": 2.8647948318345097e-05,
      "loss": 2.1714,
      "step": 1160
    },
    {
      "epoch": 4.62,
      "grad_norm": 2.1899473667144775,
      "learning_rate": 2.8338112174570263e-05,
      "loss": 2.1499,
      "step": 1170
    },
    {
      "epoch": 4.66,
      "grad_norm": 2.3744349479675293,
      "learning_rate": 2.8027753095976793e-05,
      "loss": 2.2543,
      "step": 1180
    },
    {
      "epoch": 4.7,
      "grad_norm": 2.3728389739990234,
      "learning_rate": 2.7716919702131662e-05,
      "loss": 2.2539,
      "step": 1190
    },
    {
      "epoch": 4.74,
      "grad_norm": 2.3791847229003906,
      "learning_rate": 2.740566068690612e-05,
      "loss": 2.1576,
      "step": 1200
    },
    {
      "epoch": 4.74,
      "eval_loss": 2.1767776012420654,
      "eval_runtime": 143.8644,
      "eval_samples_per_second": 6.256,
      "eval_steps_per_second": 3.128,
      "step": 1200
    },
    {
      "epoch": 4.78,
      "grad_norm": 1.933295726776123,
      "learning_rate": 2.70940248108475e-05,
      "loss": 2.1457,
      "step": 1210
    },
    {
      "epoch": 4.82,
      "grad_norm": 1.918013095855713,
      "learning_rate": 2.6782060893540578e-05,
      "loss": 2.1832,
      "step": 1220
    },
    {
      "epoch": 4.86,
      "grad_norm": 2.6071360111236572,
      "learning_rate": 2.646981780595973e-05,
      "loss": 2.2199,
      "step": 1230
    },
    {
      "epoch": 4.9,
      "grad_norm": 2.3134567737579346,
      "learning_rate": 2.6157344462812976e-05,
      "loss": 2.2627,
      "step": 1240
    },
    {
      "epoch": 4.94,
      "grad_norm": 2.3015966415405273,
      "learning_rate": 2.5844689814879234e-05,
      "loss": 2.2163,
      "step": 1250
    },
    {
      "epoch": 4.98,
      "grad_norm": 2.3684730529785156,
      "learning_rate": 2.5531902841339867e-05,
      "loss": 2.1699,
      "step": 1260
    },
    {
      "epoch": 5.02,
      "grad_norm": 1.8125172853469849,
      "learning_rate": 2.5219032542105807e-05,
      "loss": 2.125,
      "step": 1270
    },
    {
      "epoch": 5.06,
      "grad_norm": 1.954872965812683,
      "learning_rate": 2.4906127930141478e-05,
      "loss": 2.2746,
      "step": 1280
    },
    {
      "epoch": 5.1,
      "grad_norm": 2.147181749343872,
      "learning_rate": 2.459323802378656e-05,
      "loss": 2.2056,
      "step": 1290
    },
    {
      "epoch": 5.14,
      "grad_norm": 2.522922992706299,
      "learning_rate": 2.428041183907704e-05,
      "loss": 2.2883,
      "step": 1300
    },
    {
      "epoch": 5.18,
      "grad_norm": 2.107692003250122,
      "learning_rate": 2.396769838206651e-05,
      "loss": 2.1534,
      "step": 1310
    },
    {
      "epoch": 5.21,
      "grad_norm": 2.043048620223999,
      "learning_rate": 2.3655146641149127e-05,
      "loss": 2.1946,
      "step": 1320
    },
    {
      "epoch": 5.25,
      "grad_norm": 3.606001615524292,
      "learning_rate": 2.3342805579385278e-05,
      "loss": 2.2133,
      "step": 1330
    },
    {
      "epoch": 5.29,
      "grad_norm": 1.9504977464675903,
      "learning_rate": 2.3030724126831202e-05,
      "loss": 2.2014,
      "step": 1340
    },
    {
      "epoch": 5.33,
      "grad_norm": 2.8629162311553955,
      "learning_rate": 2.2718951172873868e-05,
      "loss": 2.1473,
      "step": 1350
    },
    {
      "epoch": 5.37,
      "grad_norm": 2.781360387802124,
      "learning_rate": 2.24075355585721e-05,
      "loss": 2.2206,
      "step": 1360
    },
    {
      "epoch": 5.41,
      "grad_norm": 2.6987454891204834,
      "learning_rate": 2.209652606900541e-05,
      "loss": 2.1345,
      "step": 1370
    },
    {
      "epoch": 5.45,
      "grad_norm": 3.2475154399871826,
      "learning_rate": 2.1785971425631446e-05,
      "loss": 2.1578,
      "step": 1380
    },
    {
      "epoch": 5.49,
      "grad_norm": 2.428025007247925,
      "learning_rate": 2.1475920278653592e-05,
      "loss": 2.0979,
      "step": 1390
    },
    {
      "epoch": 5.53,
      "grad_norm": 2.3817710876464844,
      "learning_rate": 2.116642119939952e-05,
      "loss": 2.1287,
      "step": 1400
    },
    {
      "epoch": 5.57,
      "grad_norm": 3.956568479537964,
      "learning_rate": 2.0857522672712298e-05,
      "loss": 2.2171,
      "step": 1410
    },
    {
      "epoch": 5.61,
      "grad_norm": 2.3348941802978516,
      "learning_rate": 2.054927308935489e-05,
      "loss": 2.307,
      "step": 1420
    },
    {
      "epoch": 5.65,
      "grad_norm": 2.4065921306610107,
      "learning_rate": 2.024172073842952e-05,
      "loss": 2.174,
      "step": 1430
    },
    {
      "epoch": 5.69,
      "grad_norm": 2.090249538421631,
      "learning_rate": 1.993491379981284e-05,
      "loss": 2.2008,
      "step": 1440
    },
    {
      "epoch": 5.73,
      "grad_norm": 4.224242687225342,
      "learning_rate": 1.96289003366083e-05,
      "loss": 2.2197,
      "step": 1450
    },
    {
      "epoch": 5.77,
      "grad_norm": 2.2841567993164062,
      "learning_rate": 1.9323728287616806e-05,
      "loss": 2.2154,
      "step": 1460
    },
    {
      "epoch": 5.81,
      "grad_norm": 2.359734296798706,
      "learning_rate": 1.901944545982677e-05,
      "loss": 2.1683,
      "step": 1470
    },
    {
      "epoch": 5.85,
      "grad_norm": 2.155123233795166,
      "learning_rate": 1.8716099520924923e-05,
      "loss": 2.2639,
      "step": 1480
    },
    {
      "epoch": 5.89,
      "grad_norm": 2.440293550491333,
      "learning_rate": 1.8413737991828843e-05,
      "loss": 2.2023,
      "step": 1490
    },
    {
      "epoch": 5.93,
      "grad_norm": 2.8041791915893555,
      "learning_rate": 1.811240823924256e-05,
      "loss": 2.1118,
      "step": 1500
    },
    {
      "epoch": 5.97,
      "grad_norm": 2.284633159637451,
      "learning_rate": 1.7812157468236274e-05,
      "loss": 2.1564,
      "step": 1510
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.427250623703003,
      "learning_rate": 1.7513032714851386e-05,
      "loss": 2.1653,
      "step": 1520
    },
    {
      "epoch": 6.04,
      "grad_norm": 2.3097784519195557,
      "learning_rate": 1.721508083873207e-05,
      "loss": 2.2382,
      "step": 1530
    },
    {
      "epoch": 6.08,
      "grad_norm": 2.3014960289001465,
      "learning_rate": 1.6918348515784387e-05,
      "loss": 2.2634,
      "step": 1540
    },
    {
      "epoch": 6.12,
      "grad_norm": 2.0459794998168945,
      "learning_rate": 1.6622882230864255e-05,
      "loss": 2.1997,
      "step": 1550
    },
    {
      "epoch": 6.16,
      "grad_norm": 1.938412070274353,
      "learning_rate": 1.632872827049533e-05,
      "loss": 2.2056,
      "step": 1560
    },
    {
      "epoch": 6.2,
      "grad_norm": 2.71195650100708,
      "learning_rate": 1.603593271561794e-05,
      "loss": 2.141,
      "step": 1570
    },
    {
      "epoch": 6.24,
      "grad_norm": 2.2068488597869873,
      "learning_rate": 1.5744541434370235e-05,
      "loss": 2.1678,
      "step": 1580
    },
    {
      "epoch": 6.28,
      "grad_norm": 2.839308023452759,
      "learning_rate": 1.545460007490268e-05,
      "loss": 2.1978,
      "step": 1590
    },
    {
      "epoch": 6.32,
      "grad_norm": 2.6756937503814697,
      "learning_rate": 1.5166154058227003e-05,
      "loss": 2.1811,
      "step": 1600
    },
    {
      "epoch": 6.32,
      "eval_loss": 2.1764707565307617,
      "eval_runtime": 143.5803,
      "eval_samples_per_second": 6.268,
      "eval_steps_per_second": 3.134,
      "step": 1600
    },
    {
      "epoch": 6.36,
      "grad_norm": 2.5529654026031494,
      "learning_rate": 1.4879248571100729e-05,
      "loss": 2.1933,
      "step": 1610
    },
    {
      "epoch": 6.4,
      "grad_norm": 2.277837038040161,
      "learning_rate": 1.4593928558948403e-05,
      "loss": 2.1276,
      "step": 1620
    },
    {
      "epoch": 6.44,
      "grad_norm": 2.087127208709717,
      "learning_rate": 1.4310238718820638e-05,
      "loss": 2.1158,
      "step": 1630
    },
    {
      "epoch": 6.48,
      "grad_norm": 2.717027187347412,
      "learning_rate": 1.4028223492392062e-05,
      "loss": 2.1657,
      "step": 1640
    },
    {
      "epoch": 6.52,
      "grad_norm": 5.161482334136963,
      "learning_rate": 1.3747927058999228e-05,
      "loss": 2.1421,
      "step": 1650
    },
    {
      "epoch": 6.56,
      "grad_norm": 3.2789719104766846,
      "learning_rate": 1.3469393328719732e-05,
      "loss": 2.1197,
      "step": 1660
    },
    {
      "epoch": 6.6,
      "grad_norm": 2.523096799850464,
      "learning_rate": 1.3192665935493401e-05,
      "loss": 2.1886,
      "step": 1670
    },
    {
      "epoch": 6.64,
      "grad_norm": 2.650481939315796,
      "learning_rate": 1.2917788230286753e-05,
      "loss": 2.1621,
      "step": 1680
    },
    {
      "epoch": 6.68,
      "grad_norm": 2.2410335540771484,
      "learning_rate": 1.2644803274301879e-05,
      "loss": 2.1615,
      "step": 1690
    },
    {
      "epoch": 6.72,
      "grad_norm": 2.0118298530578613,
      "learning_rate": 1.2373753832230631e-05,
      "loss": 2.1762,
      "step": 1700
    },
    {
      "epoch": 6.76,
      "grad_norm": 2.5051441192626953,
      "learning_rate": 1.210468236555523e-05,
      "loss": 2.1252,
      "step": 1710
    },
    {
      "epoch": 6.8,
      "grad_norm": 1.9113433361053467,
      "learning_rate": 1.1837631025896506e-05,
      "loss": 2.0933,
      "step": 1720
    },
    {
      "epoch": 6.83,
      "grad_norm": 2.689561367034912,
      "learning_rate": 1.1572641648410554e-05,
      "loss": 2.1595,
      "step": 1730
    },
    {
      "epoch": 6.87,
      "grad_norm": 2.5091662406921387,
      "learning_rate": 1.1309755745235005e-05,
      "loss": 2.1441,
      "step": 1740
    },
    {
      "epoch": 6.91,
      "grad_norm": 3.028869152069092,
      "learning_rate": 1.1049014498985932e-05,
      "loss": 2.186,
      "step": 1750
    },
    {
      "epoch": 6.95,
      "grad_norm": 2.106482744216919,
      "learning_rate": 1.0790458756306357e-05,
      "loss": 2.1507,
      "step": 1760
    },
    {
      "epoch": 6.99,
      "grad_norm": 3.192108154296875,
      "learning_rate": 1.0534129021467363e-05,
      "loss": 2.187,
      "step": 1770
    },
    {
      "epoch": 7.03,
      "grad_norm": 2.394237518310547,
      "learning_rate": 1.0280065450022844e-05,
      "loss": 2.1725,
      "step": 1780
    },
    {
      "epoch": 7.07,
      "grad_norm": 2.58278226852417,
      "learning_rate": 1.002830784251896e-05,
      "loss": 2.1623,
      "step": 1790
    },
    {
      "epoch": 7.11,
      "grad_norm": 2.267213821411133,
      "learning_rate": 9.77889563825912e-06,
      "loss": 2.1668,
      "step": 1800
    },
    {
      "epoch": 7.15,
      "grad_norm": 2.4960286617279053,
      "learning_rate": 9.531867909125544e-06,
      "loss": 2.2438,
      "step": 1810
    },
    {
      "epoch": 7.19,
      "grad_norm": 2.337390899658203,
      "learning_rate": 9.287263353458501e-06,
      "loss": 2.2111,
      "step": 1820
    },
    {
      "epoch": 7.23,
      "grad_norm": 3.2264187335968018,
      "learning_rate": 9.045120289993936e-06,
      "loss": 2.1189,
      "step": 1830
    },
    {
      "epoch": 7.27,
      "grad_norm": 2.872370719909668,
      "learning_rate": 8.805476651860607e-06,
      "loss": 2.1599,
      "step": 1840
    },
    {
      "epoch": 7.31,
      "grad_norm": 2.6654868125915527,
      "learning_rate": 8.568369980637678e-06,
      "loss": 2.1976,
      "step": 1850
    },
    {
      "epoch": 7.35,
      "grad_norm": 2.957482099533081,
      "learning_rate": 8.333837420473592e-06,
      "loss": 2.1056,
      "step": 1860
    },
    {
      "epoch": 7.39,
      "grad_norm": 2.7675461769104004,
      "learning_rate": 8.101915712267228e-06,
      "loss": 2.1528,
      "step": 1870
    },
    {
      "epoch": 7.43,
      "grad_norm": 2.2371811866760254,
      "learning_rate": 7.872641187912183e-06,
      "loss": 2.1607,
      "step": 1880
    },
    {
      "epoch": 7.47,
      "grad_norm": 2.760326862335205,
      "learning_rate": 7.64604976460524e-06,
      "loss": 2.0584,
      "step": 1890
    },
    {
      "epoch": 7.51,
      "grad_norm": 2.8428192138671875,
      "learning_rate": 7.422176939219682e-06,
      "loss": 2.1451,
      "step": 1900
    },
    {
      "epoch": 7.55,
      "grad_norm": 2.6749818325042725,
      "learning_rate": 7.2010577827444945e-06,
      "loss": 2.1713,
      "step": 1910
    },
    {
      "epoch": 7.59,
      "grad_norm": 2.2493717670440674,
      "learning_rate": 6.9827269347903204e-06,
      "loss": 2.195,
      "step": 1920
    },
    {
      "epoch": 7.62,
      "grad_norm": 2.8834621906280518,
      "learning_rate": 6.767218598162947e-06,
      "loss": 2.1771,
      "step": 1930
    },
    {
      "epoch": 7.66,
      "grad_norm": 2.361966371536255,
      "learning_rate": 6.554566533505213e-06,
      "loss": 2.2726,
      "step": 1940
    },
    {
      "epoch": 7.7,
      "grad_norm": 2.7109503746032715,
      "learning_rate": 6.344804054008252e-06,
      "loss": 2.1248,
      "step": 1950
    },
    {
      "epoch": 7.74,
      "grad_norm": 2.7265536785125732,
      "learning_rate": 6.137964020192749e-06,
      "loss": 2.1833,
      "step": 1960
    },
    {
      "epoch": 7.78,
      "grad_norm": 2.4823639392852783,
      "learning_rate": 5.934078834761176e-06,
      "loss": 2.1631,
      "step": 1970
    },
    {
      "epoch": 7.82,
      "grad_norm": 2.896096706390381,
      "learning_rate": 5.733180437521676e-06,
      "loss": 2.1357,
      "step": 1980
    },
    {
      "epoch": 7.86,
      "grad_norm": 3.116410732269287,
      "learning_rate": 5.535300300384552e-06,
      "loss": 2.1195,
      "step": 1990
    },
    {
      "epoch": 7.9,
      "grad_norm": 2.8223111629486084,
      "learning_rate": 5.34046942243199e-06,
      "loss": 2.1649,
      "step": 2000
    },
    {
      "epoch": 7.9,
      "eval_loss": 2.1766035556793213,
      "eval_runtime": 142.6761,
      "eval_samples_per_second": 6.308,
      "eval_steps_per_second": 3.154,
      "step": 2000
    },
    {
      "epoch": 7.94,
      "grad_norm": 2.767136812210083,
      "learning_rate": 5.148718325061858e-06,
      "loss": 2.1394,
      "step": 2010
    },
    {
      "epoch": 7.98,
      "grad_norm": 2.9371230602264404,
      "learning_rate": 4.960077047206374e-06,
      "loss": 2.1314,
      "step": 2020
    },
    {
      "epoch": 8.02,
      "grad_norm": 2.5136184692382812,
      "learning_rate": 4.7745751406263165e-06,
      "loss": 2.127,
      "step": 2030
    },
    {
      "epoch": 8.06,
      "grad_norm": 2.1684181690216064,
      "learning_rate": 4.592241665281555e-06,
      "loss": 2.0934,
      "step": 2040
    },
    {
      "epoch": 8.1,
      "grad_norm": 2.427680730819702,
      "learning_rate": 4.4131051847786735e-06,
      "loss": 2.1284,
      "step": 2050
    },
    {
      "epoch": 8.14,
      "grad_norm": 2.8133230209350586,
      "learning_rate": 4.237193761896291e-06,
      "loss": 2.2072,
      "step": 2060
    },
    {
      "epoch": 8.18,
      "grad_norm": 2.575359344482422,
      "learning_rate": 4.06453495418885e-06,
      "loss": 2.1865,
      "step": 2070
    },
    {
      "epoch": 8.22,
      "grad_norm": 2.287165403366089,
      "learning_rate": 3.895155809669599e-06,
      "loss": 2.1804,
      "step": 2080
    },
    {
      "epoch": 8.26,
      "grad_norm": 2.5095739364624023,
      "learning_rate": 3.72908286257333e-06,
      "loss": 2.0919,
      "step": 2090
    },
    {
      "epoch": 8.3,
      "grad_norm": 2.8196492195129395,
      "learning_rate": 3.566342129199665e-06,
      "loss": 2.0937,
      "step": 2100
    },
    {
      "epoch": 8.34,
      "grad_norm": 2.5309383869171143,
      "learning_rate": 3.406959103837412e-06,
      "loss": 2.2098,
      "step": 2110
    },
    {
      "epoch": 8.38,
      "grad_norm": 2.6270852088928223,
      "learning_rate": 3.250958754770775e-06,
      "loss": 2.1873,
      "step": 2120
    },
    {
      "epoch": 8.41,
      "grad_norm": 2.9431707859039307,
      "learning_rate": 3.09836552036791e-06,
      "loss": 2.126,
      "step": 2130
    },
    {
      "epoch": 8.45,
      "grad_norm": 2.5240731239318848,
      "learning_rate": 2.949203305252496e-06,
      "loss": 2.0116,
      "step": 2140
    },
    {
      "epoch": 8.49,
      "grad_norm": 2.583218812942505,
      "learning_rate": 2.8034954765589566e-06,
      "loss": 2.0481,
      "step": 2150
    },
    {
      "epoch": 8.53,
      "grad_norm": 2.5131754875183105,
      "learning_rate": 2.66126486027187e-06,
      "loss": 2.1438,
      "step": 2160
    },
    {
      "epoch": 8.57,
      "grad_norm": 3.5266616344451904,
      "learning_rate": 2.5225337376501015e-06,
      "loss": 2.1933,
      "step": 2170
    },
    {
      "epoch": 8.61,
      "grad_norm": 2.9746336936950684,
      "learning_rate": 2.3873238417363634e-06,
      "loss": 2.1236,
      "step": 2180
    },
    {
      "epoch": 8.65,
      "grad_norm": 2.598071813583374,
      "learning_rate": 2.2556563539525815e-06,
      "loss": 2.0947,
      "step": 2190
    },
    {
      "epoch": 8.69,
      "grad_norm": 3.164978265762329,
      "learning_rate": 2.1275519007817044e-06,
      "loss": 2.1925,
      "step": 2200
    },
    {
      "epoch": 8.73,
      "grad_norm": 2.4741063117980957,
      "learning_rate": 2.003030550536439e-06,
      "loss": 2.1531,
      "step": 2210
    },
    {
      "epoch": 8.77,
      "grad_norm": 2.4534201622009277,
      "learning_rate": 1.882111810215445e-06,
      "loss": 2.1345,
      "step": 2220
    },
    {
      "epoch": 8.81,
      "grad_norm": 2.8546104431152344,
      "learning_rate": 1.7648146224474444e-06,
      "loss": 2.145,
      "step": 2230
    },
    {
      "epoch": 8.85,
      "grad_norm": 2.3898508548736572,
      "learning_rate": 1.651157362523731e-06,
      "loss": 2.1689,
      "step": 2240
    },
    {
      "epoch": 8.89,
      "grad_norm": 2.7107295989990234,
      "learning_rate": 1.541157835519605e-06,
      "loss": 2.232,
      "step": 2250
    },
    {
      "epoch": 8.93,
      "grad_norm": 2.8263099193573,
      "learning_rate": 1.434833273505079e-06,
      "loss": 2.1338,
      "step": 2260
    },
    {
      "epoch": 8.97,
      "grad_norm": 2.6851024627685547,
      "learning_rate": 1.3322003328453874e-06,
      "loss": 2.2601,
      "step": 2270
    },
    {
      "epoch": 9.01,
      "grad_norm": 2.2520766258239746,
      "learning_rate": 1.2332750915916458e-06,
      "loss": 2.1185,
      "step": 2280
    },
    {
      "epoch": 9.05,
      "grad_norm": 2.532531499862671,
      "learning_rate": 1.1380730469621653e-06,
      "loss": 2.0739,
      "step": 2290
    },
    {
      "epoch": 9.09,
      "grad_norm": 2.496253490447998,
      "learning_rate": 1.046609112914701e-06,
      "loss": 2.1072,
      "step": 2300
    },
    {
      "epoch": 9.13,
      "grad_norm": 2.455324649810791,
      "learning_rate": 9.588976178100866e-07,
      "loss": 2.1398,
      "step": 2310
    },
    {
      "epoch": 9.17,
      "grad_norm": 2.132638692855835,
      "learning_rate": 8.749523021676309e-07,
      "loss": 2.1913,
      "step": 2320
    },
    {
      "epoch": 9.2,
      "grad_norm": 3.008934497833252,
      "learning_rate": 7.94786316512583e-07,
      "loss": 2.1616,
      "step": 2330
    },
    {
      "epoch": 9.24,
      "grad_norm": 3.200282096862793,
      "learning_rate": 7.184122193160032e-07,
      "loss": 2.1086,
      "step": 2340
    },
    {
      "epoch": 9.28,
      "grad_norm": 2.8957929611206055,
      "learning_rate": 6.458419750274336e-07,
      "loss": 2.26,
      "step": 2350
    },
    {
      "epoch": 9.32,
      "grad_norm": 2.499394178390503,
      "learning_rate": 5.770869522005839e-07,
      "loss": 2.052,
      "step": 2360
    },
    {
      "epoch": 9.36,
      "grad_norm": 2.7533364295959473,
      "learning_rate": 5.121579217123751e-07,
      "loss": 2.0698,
      "step": 2370
    },
    {
      "epoch": 9.4,
      "grad_norm": 2.28757643699646,
      "learning_rate": 4.510650550756279e-07,
      "loss": 2.1057,
      "step": 2380
    },
    {
      "epoch": 9.44,
      "grad_norm": 2.910390853881836,
      "learning_rate": 3.938179228456379e-07,
      "loss": 2.1283,
      "step": 2390
    },
    {
      "epoch": 9.48,
      "grad_norm": 2.4329330921173096,
      "learning_rate": 3.4042549312087726e-07,
      "loss": 2.1422,
      "step": 2400
    },
    {
      "epoch": 9.48,
      "eval_loss": 2.176323175430298,
      "eval_runtime": 142.2743,
      "eval_samples_per_second": 6.326,
      "eval_steps_per_second": 3.163,
      "step": 2400
    }
  ],
  "logging_steps": 10,
  "max_steps": 2530,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 400,
  "total_flos": 3.0804047954968576e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}