{ "best_global_step": 1500, "best_metric": 1.14039981, "best_model_checkpoint": "/NHNHOME/data/sanghyeok/qwen-cua/runs/probe2000_idm_code_full/v1-20260621-231848/checkpoint-1500", "epoch": 0.34594594594594597, "eval_steps": 500, "global_step": 1500, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00023063063063063062, "grad_norm": 1592.4940185546875, "learning_rate": 1.0000000000000001e-07, "loss": 79.31526184082031, "loss_idm": 4.953125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0621337890625, "loss_idm_text": 1.8660085201263428, "loss_idm_type": 3.003211200237274, "loss_sft": 0.013178251218050718, "step": 1, "token_acc": 0.855064946292334 }, { "epoch": 0.0023063063063063064, "grad_norm": 1613.28076171875, "learning_rate": 1.0000000000000002e-06, "loss": 76.10415310329861, "loss_idm": 4.284722222222222, "loss_idm_anl": 0.0, "loss_idm_coord": 0.05967881944444445, "loss_idm_text": 1.7714627848731146, "loss_idm_type": 2.89741759830051, "loss_sft": 0.013162603808773888, "step": 10, "token_acc": 0.8557915694044088 }, { "epoch": 0.004612612612612613, "grad_norm": 348.3831481933594, "learning_rate": 2.0000000000000003e-06, "loss": 50.42054443359375, "loss_idm": 2.9625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0595703125, "loss_idm_text": 1.6456905126571655, "loss_idm_type": 1.4186124086380005, "loss_sft": 0.013041298277676106, "step": 20, "token_acc": 0.8575046262325297 }, { "epoch": 0.006918918918918919, "grad_norm": 126.2293930053711, "learning_rate": 3e-06, "loss": 44.00212707519531, "loss_idm": 2.5375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.052490234375, "loss_idm_text": 1.6745069742202758, "loss_idm_type": 0.9961585998535156, "loss_sft": 0.013043427839875221, "step": 30, "token_acc": 0.8575169162329329 }, { "epoch": 0.009225225225225226, "grad_norm": 161.60565185546875, "learning_rate": 4.000000000000001e-06, "loss": 40.35985412597656, "loss_idm": 2.340625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0490234375, "loss_idm_text": 1.5050048112869263, "loss_idm_type": 0.9409058809280395, "loss_sft": 0.013070817664265632, "step": 40, "token_acc": 0.8573629929949759 }, { "epoch": 0.011531531531531532, "grad_norm": 366.02923583984375, "learning_rate": 5e-06, "loss": 36.85423583984375, "loss_idm": 2.15625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.045751953125, "loss_idm_text": 1.454877495765686, "loss_idm_type": 0.7732545018196106, "loss_sft": 0.014006125554442406, "step": 50, "token_acc": 0.8500105439434349 }, { "epoch": 0.013837837837837838, "grad_norm": 334.4949951171875, "learning_rate": 6e-06, "loss": 27.9380615234375, "loss_idm": 1.6390625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.04501953125, "loss_idm_text": 1.0883261203765868, "loss_idm_type": 0.5830961108207703, "loss_sft": 0.014523161575198174, "step": 60, "token_acc": 0.8460712220110717 }, { "epoch": 0.016144144144144144, "grad_norm": 287.8810729980469, "learning_rate": 7e-06, "loss": 28.939056396484375, "loss_idm": 1.696875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.04052734375, "loss_idm_text": 1.0840964794158936, "loss_idm_type": 0.6544631838798523, "loss_sft": 0.01408173404633999, "step": 70, "token_acc": 0.8500501916749074 }, { "epoch": 0.01845045045045045, "grad_norm": 164.0933380126953, "learning_rate": 8.000000000000001e-06, "loss": 24.88965606689453, "loss_idm": 1.4421875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0427734375, "loss_idm_text": 0.8765177607536316, "loss_idm_type": 0.607001006603241, "loss_sft": 0.013972016051411629, "step": 80, "token_acc": 0.850499639034801 }, { "epoch": 0.020756756756756756, "grad_norm": 271.42523193359375, "learning_rate": 9e-06, "loss": 24.212254333496094, "loss_idm": 1.409375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.03955078125, "loss_idm_text": 0.8325042963027954, "loss_idm_type": 0.6114315390586853, "loss_sft": 0.014018873870372772, "step": 90, "token_acc": 0.8487399360383294 }, { "epoch": 0.023063063063063063, "grad_norm": 229.35255432128906, "learning_rate": 1e-05, "loss": 25.997027587890624, "loss_idm": 1.5015625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.039697265625, "loss_idm_text": 0.9402544498443604, "loss_idm_type": 0.6157279968261719, "loss_sft": 0.014289245754480363, "step": 100, "token_acc": 0.8460278937791673 }, { "epoch": 0.025369369369369368, "grad_norm": 294.6478576660156, "learning_rate": 9.999316524962347e-06, "loss": 25.035136413574218, "loss_idm": 1.45625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.034033203125, "loss_idm_text": 0.9034677743911743, "loss_idm_type": 0.5977115869522095, "loss_sft": 0.014173812046647073, "step": 110, "token_acc": 0.8471177587505733 }, { "epoch": 0.027675675675675675, "grad_norm": 233.29827880859375, "learning_rate": 9.99726628670463e-06, "loss": 26.405145263671876, "loss_idm": 1.553125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.035498046875, "loss_idm_text": 0.968865442276001, "loss_idm_type": 0.6166337251663208, "loss_sft": 0.013977998867630959, "step": 120, "token_acc": 0.8493716829225565 }, { "epoch": 0.029981981981981983, "grad_norm": 314.5971374511719, "learning_rate": 9.993849845741525e-06, "loss": 25.430509948730467, "loss_idm": 1.471875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.03779296875, "loss_idm_text": 0.9332091808319092, "loss_idm_type": 0.5891867876052856, "loss_sft": 0.014243610203266144, "step": 130, "token_acc": 0.8470191198367073 }, { "epoch": 0.03228828828828829, "grad_norm": 199.4022674560547, "learning_rate": 9.989068136093873e-06, "loss": 23.130665588378907, "loss_idm": 1.3390625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.03154296875, "loss_idm_text": 0.7847968459129333, "loss_idm_type": 0.6001523256301879, "loss_sft": 0.014024525508284568, "step": 140, "token_acc": 0.8488714086911894 }, { "epoch": 0.034594594594594595, "grad_norm": 250.71766662597656, "learning_rate": 9.98292246503335e-06, "loss": 26.04180908203125, "loss_idm": 1.5328125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0306640625, "loss_idm_text": 0.9394279241561889, "loss_idm_type": 0.6282362818717957, "loss_sft": 0.014096884429454804, "step": 150, "token_acc": 0.8478837080550887 }, { "epoch": 0.0369009009009009, "grad_norm": 166.97149658203125, "learning_rate": 9.975414512725058e-06, "loss": 21.466415405273438, "loss_idm": 1.246875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.032177734375, "loss_idm_text": 0.7001996755599975, "loss_idm_type": 0.5806768774986267, "loss_sft": 0.013779866695404052, "step": 160, "token_acc": 0.8516435390217498 }, { "epoch": 0.03920720720720721, "grad_norm": 166.47470092773438, "learning_rate": 9.966546331768192e-06, "loss": 21.78998565673828, "loss_idm": 1.2703125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0283203125, "loss_idm_text": 0.6905548334121704, "loss_idm_type": 0.6141234040260315, "loss_sft": 0.014114163070917129, "step": 170, "token_acc": 0.8484753365377703 }, { "epoch": 0.04151351351351351, "grad_norm": 218.82046508789062, "learning_rate": 9.956320346634877e-06, "loss": 22.165342712402342, "loss_idm": 1.2859375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.02763671875, "loss_idm_text": 0.7303984642028809, "loss_idm_type": 0.5983686685562134, "loss_sft": 0.013757656514644622, "step": 180, "token_acc": 0.8512506835204802 }, { "epoch": 0.04381981981981982, "grad_norm": 219.92880249023438, "learning_rate": 9.944739353007344e-06, "loss": 21.063986206054686, "loss_idm": 1.209375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.02802734375, "loss_idm_text": 0.6704511880874634, "loss_idm_type": 0.589215099811554, "loss_sft": 0.014041738212108612, "step": 190, "token_acc": 0.8489764234034295 }, { "epoch": 0.04612612612612613, "grad_norm": 160.0710906982422, "learning_rate": 9.931806517013612e-06, "loss": 22.00762023925781, "loss_idm": 1.26875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.028125, "loss_idm_text": 0.649952232837677, "loss_idm_type": 0.6678654789924622, "loss_sft": 0.014099276438355447, "step": 200, "token_acc": 0.8480138468195586 }, { "epoch": 0.048432432432432435, "grad_norm": 231.97708129882812, "learning_rate": 9.917525374361913e-06, "loss": 22.615927124023436, "loss_idm": 1.3203125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0266357421875, "loss_idm_text": 0.7614493370056152, "loss_idm_type": 0.5959304571151733, "loss_sft": 0.014167248830199241, "step": 210, "token_acc": 0.8475955923181496 }, { "epoch": 0.050738738738738735, "grad_norm": 199.168212890625, "learning_rate": 9.901899829374048e-06, "loss": 20.7582763671875, "loss_idm": 1.1828125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.030615234375, "loss_idm_text": 0.6515186548233032, "loss_idm_type": 0.5868193507194519, "loss_sft": 0.013925193250179291, "step": 220, "token_acc": 0.8496626268588034 }, { "epoch": 0.05304504504504504, "grad_norm": 160.00375366210938, "learning_rate": 9.884934153917998e-06, "loss": 22.84142608642578, "loss_idm": 1.34375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.028173828125, "loss_idm_text": 0.7582223653793335, "loss_idm_type": 0.6115516662597656, "loss_sft": 0.014063675329089165, "step": 230, "token_acc": 0.8490615719208343 }, { "epoch": 0.05535135135135135, "grad_norm": 185.79376220703125, "learning_rate": 9.86663298624003e-06, "loss": 21.26628875732422, "loss_idm": 1.225, "loss_idm_anl": 0.0, "loss_idm_coord": 0.02509765625, "loss_idm_text": 0.7225316286087036, "loss_idm_type": 0.5519452333450318, "loss_sft": 0.014292314276099205, "step": 240, "token_acc": 0.8462079420660857 }, { "epoch": 0.05765765765765766, "grad_norm": 165.556884765625, "learning_rate": 9.847001329696653e-06, "loss": 20.54603271484375, "loss_idm": 1.1765625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.02568359375, "loss_idm_text": 0.6419167995452881, "loss_idm_type": 0.586991822719574, "loss_sft": 0.014299248531460761, "step": 250, "token_acc": 0.846426558505606 }, { "epoch": 0.059963963963963966, "grad_norm": 181.50538635253906, "learning_rate": 9.826044551386743e-06, "loss": 21.09702911376953, "loss_idm": 1.2109375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.023583984375, "loss_idm_text": 0.6792068004608154, "loss_idm_type": 0.5862257599830627, "loss_sft": 0.014360376074910164, "step": 260, "token_acc": 0.8455009827053707 }, { "epoch": 0.06227027027027027, "grad_norm": 159.72378540039062, "learning_rate": 9.803768380684242e-06, "loss": 20.137149047851562, "loss_idm": 1.1671875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.02783203125, "loss_idm_text": 0.6360970735549927, "loss_idm_type": 0.5649659037590027, "loss_sft": 0.014287288114428521, "step": 270, "token_acc": 0.8466014645893744 }, { "epoch": 0.06457657657657657, "grad_norm": 142.44703674316406, "learning_rate": 9.780178907671788e-06, "loss": 21.38773193359375, "loss_idm": 1.2390625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.027734375, "loss_idm_text": 0.679478108882904, "loss_idm_type": 0.5993240475654602, "loss_sft": 0.014427094906568527, "step": 280, "token_acc": 0.845819646334507 }, { "epoch": 0.06688288288288288, "grad_norm": 167.3060302734375, "learning_rate": 9.755282581475769e-06, "loss": 18.871426391601563, "loss_idm": 1.0859375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.026220703125, "loss_idm_text": 0.5865870356559754, "loss_idm_type": 0.536228609085083, "loss_sft": 0.014549480751156807, "step": 290, "token_acc": 0.8449012068675806 }, { "epoch": 0.06918918918918919, "grad_norm": 138.20761108398438, "learning_rate": 9.729086208503174e-06, "loss": 20.88959197998047, "loss_idm": 1.221875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.026416015625, "loss_idm_text": 0.7052467823028564, "loss_idm_type": 0.5426920533180237, "loss_sft": 0.014955276995897293, "step": 300, "token_acc": 0.8407353983815368 }, { "epoch": 0.0714954954954955, "grad_norm": 183.0985565185547, "learning_rate": 9.701596950580807e-06, "loss": 20.191453552246095, "loss_idm": 1.153125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0275390625, "loss_idm_text": 0.6399867296218872, "loss_idm_type": 0.5646127104759217, "loss_sft": 0.01429852657020092, "step": 310, "token_acc": 0.8468957501589651 }, { "epoch": 0.0738018018018018, "grad_norm": 139.37664794921875, "learning_rate": 9.672822322997305e-06, "loss": 18.14356384277344, "loss_idm": 1.053125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.024609375, "loss_idm_text": 0.5447785139083863, "loss_idm_type": 0.534551739692688, "loss_sft": 0.01457935906946659, "step": 320, "token_acc": 0.8441828406914472 }, { "epoch": 0.07610810810810811, "grad_norm": 165.80722045898438, "learning_rate": 9.642770192448537e-06, "loss": 21.30792236328125, "loss_idm": 1.2375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0251953125, "loss_idm_text": 0.7330950379371644, "loss_idm_type": 0.5434134244918823, "loss_sft": 0.014491663500666618, "step": 330, "token_acc": 0.844535425337065 }, { "epoch": 0.07841441441441442, "grad_norm": 135.9119415283203, "learning_rate": 9.611448774886925e-06, "loss": 21.87939910888672, "loss_idm": 1.271875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.02431640625, "loss_idm_text": 0.7017751932144165, "loss_idm_type": 0.6117419362068176, "loss_sft": 0.01456201747059822, "step": 340, "token_acc": 0.8440950452721977 }, { "epoch": 0.08072072072072072, "grad_norm": 168.16905212402344, "learning_rate": 9.578866633275289e-06, "loss": 20.681498718261718, "loss_idm": 1.18125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.023046875, "loss_idm_text": 0.6695239424705506, "loss_idm_type": 0.5692394852638245, "loss_sft": 0.014816844463348388, "step": 350, "token_acc": 0.8415710439552528 }, { "epoch": 0.08302702702702702, "grad_norm": 166.8290252685547, "learning_rate": 9.545032675245814e-06, "loss": 18.921286010742186, "loss_idm": 1.1046875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0237548828125, "loss_idm_text": 0.6103208661079407, "loss_idm_type": 0.519065773487091, "loss_sft": 0.014447029307484626, "step": 360, "token_acc": 0.8450666103171796 }, { "epoch": 0.08533333333333333, "grad_norm": 126.34339904785156, "learning_rate": 9.509956150664796e-06, "loss": 18.619956970214844, "loss_idm": 1.084375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.02314453125, "loss_idm_text": 0.6043006479740143, "loss_idm_type": 0.5059460878372193, "loss_sft": 0.014802501350641251, "step": 370, "token_acc": 0.8417970844912486 }, { "epoch": 0.08763963963963964, "grad_norm": 196.3380584716797, "learning_rate": 9.473646649103819e-06, "loss": 20.677947998046875, "loss_idm": 1.1921875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0225830078125, "loss_idm_text": 0.6596546173095703, "loss_idm_type": 0.5795348882675171, "loss_sft": 0.014789283275604248, "step": 380, "token_acc": 0.8423334299504732 }, { "epoch": 0.08994594594594595, "grad_norm": 134.393798828125, "learning_rate": 9.43611409721806e-06, "loss": 18.679299926757814, "loss_idm": 1.0765625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.023193359375, "loss_idm_text": 0.5959227085113525, "loss_idm_type": 0.5186289191246033, "loss_sft": 0.01440901979804039, "step": 390, "token_acc": 0.845818279940636 }, { "epoch": 0.09225225225225225, "grad_norm": 152.67922973632812, "learning_rate": 9.397368756032445e-06, "loss": 19.257763671875, "loss_idm": 1.109375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0220703125, "loss_idm_text": 0.628661286830902, "loss_idm_type": 0.5230772495269775, "loss_sft": 0.014480241760611533, "step": 400, "token_acc": 0.8448503815357494 }, { "epoch": 0.09455855855855856, "grad_norm": 122.16210174560547, "learning_rate": 9.357421218136387e-06, "loss": 19.842500305175783, "loss_idm": 1.14375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.023193359375, "loss_idm_text": 0.6230831980705261, "loss_idm_type": 0.5645471215248108, "loss_sft": 0.014090249687433243, "step": 410, "token_acc": 0.8484338211188269 }, { "epoch": 0.09686486486486487, "grad_norm": 167.9059295654297, "learning_rate": 9.31628240478787e-06, "loss": 20.216983032226562, "loss_idm": 1.1921875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0225830078125, "loss_idm_text": 0.6476395845413208, "loss_idm_type": 0.5624619722366333, "loss_sft": 0.01480226032435894, "step": 420, "token_acc": 0.8419641120404591 }, { "epoch": 0.09917117117117118, "grad_norm": 168.45751953125, "learning_rate": 9.273963562927695e-06, "loss": 19.274252319335936, "loss_idm": 1.1203125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0229736328125, "loss_idm_text": 0.5932227373123169, "loss_idm_type": 0.5581031084060669, "loss_sft": 0.014671726524829865, "step": 430, "token_acc": 0.8432967886740331 }, { "epoch": 0.10147747747747747, "grad_norm": 185.98695373535156, "learning_rate": 9.230476262104678e-06, "loss": 20.714811706542967, "loss_idm": 1.2, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0226806640625, "loss_idm_text": 0.697117555141449, "loss_idm_type": 0.5450912117958069, "loss_sft": 0.014781413972377777, "step": 440, "token_acc": 0.841088699598131 }, { "epoch": 0.10378378378378378, "grad_norm": 86.4629135131836, "learning_rate": 9.185832391312644e-06, "loss": 20.035540771484374, "loss_idm": 1.1671875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0203857421875, "loss_idm_text": 0.7159459352493286, "loss_idm_type": 0.48527697324752805, "loss_sft": 0.014686542749404907, "step": 450, "token_acc": 0.8427634581544566 }, { "epoch": 0.10609009009009009, "grad_norm": 187.40481567382812, "learning_rate": 9.140044155740102e-06, "loss": 19.649851989746093, "loss_idm": 1.1328125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0243408203125, "loss_idm_text": 0.6601190686225891, "loss_idm_type": 0.5130853056907654, "loss_sft": 0.014707070216536522, "step": 460, "token_acc": 0.8420556576404618 }, { "epoch": 0.1083963963963964, "grad_norm": 182.9649200439453, "learning_rate": 9.093124073433464e-06, "loss": 20.768836975097656, "loss_idm": 1.2015625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0226806640625, "loss_idm_text": 0.7002580642700196, "loss_idm_type": 0.5446828842163086, "loss_sft": 0.014763540402054787, "step": 470, "token_acc": 0.8420311073517537 }, { "epoch": 0.1107027027027027, "grad_norm": 149.92550659179688, "learning_rate": 9.045084971874738e-06, "loss": 18.808259582519533, "loss_idm": 1.084375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0229736328125, "loss_idm_text": 0.602880334854126, "loss_idm_type": 0.5190148711204529, "loss_sft": 0.014968127384781838, "step": 480, "token_acc": 0.8410174126994425 }, { "epoch": 0.11300900900900901, "grad_norm": 137.2638702392578, "learning_rate": 8.995939984474624e-06, "loss": 20.157275390625, "loss_idm": 1.1765625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0243408203125, "loss_idm_text": 0.6711506247520447, "loss_idm_type": 0.5331731677055359, "loss_sft": 0.014955581724643707, "step": 490, "token_acc": 0.840715306484563 }, { "epoch": 0.11531531531531532, "grad_norm": 142.88665771484375, "learning_rate": 8.94570254698197e-06, "loss": 19.9982666015625, "loss_idm": 1.1578125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0236328125, "loss_idm_text": 0.6669049620628357, "loss_idm_type": 0.5289612531661987, "loss_sft": 0.014832272753119468, "step": 500, "token_acc": 0.8413044197474939 }, { "epoch": 0.11531531531531532, "eval_loss": 1.343860387802124, "eval_runtime": 169.0406, "eval_samples_per_second": 10.323, "eval_steps_per_second": 2.585, "eval_token_acc": 0.8296746896271974, "step": 500 }, { "epoch": 0.11762162162162162, "grad_norm": 121.90780639648438, "learning_rate": 8.894386393810563e-06, "loss": 19.183363342285155, "loss_idm": 1.10703125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.02314453125, "loss_idm_text": 0.6232369840145111, "loss_idm_type": 0.5219475150108337, "loss_sft": 0.014694250002503396, "step": 510, "token_acc": 0.8422317950328234 }, { "epoch": 0.11992792792792793, "grad_norm": 130.7283477783203, "learning_rate": 8.842005554284296e-06, "loss": 21.542631530761717, "loss_idm": 1.275, "loss_idm_anl": 0.0, "loss_idm_coord": 0.023486328125, "loss_idm_text": 0.7642043590545654, "loss_idm_type": 0.527753472328186, "loss_sft": 0.014815109595656396, "step": 520, "token_acc": 0.8421332755567278 }, { "epoch": 0.12223423423423424, "grad_norm": 139.00326538085938, "learning_rate": 8.788574348801676e-06, "loss": 19.373654174804688, "loss_idm": 1.1203125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0209228515625, "loss_idm_text": 0.632088041305542, "loss_idm_type": 0.5272146463394165, "loss_sft": 0.014772192761301994, "step": 530, "token_acc": 0.8428789749734085 }, { "epoch": 0.12454054054054053, "grad_norm": 139.85838317871094, "learning_rate": 8.734107384920771e-06, "loss": 19.69678649902344, "loss_idm": 1.1359375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0214111328125, "loss_idm_text": 0.6562704801559448, "loss_idm_type": 0.5232262134552002, "loss_sft": 0.014777998626232147, "step": 540, "token_acc": 0.8425625030223899 }, { "epoch": 0.12684684684684686, "grad_norm": 148.52243041992188, "learning_rate": 8.67861955336566e-06, "loss": 18.976077270507812, "loss_idm": 1.0984375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.021044921875, "loss_idm_text": 0.6217063784599304, "loss_idm_type": 0.5128381371498107, "loss_sft": 0.014830715954303741, "step": 550, "token_acc": 0.8408637973317883 }, { "epoch": 0.12915315315315315, "grad_norm": 146.99465942382812, "learning_rate": 8.622126023955446e-06, "loss": 17.574090576171876, "loss_idm": 1.021875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0211181640625, "loss_idm_text": 0.5682262539863586, "loss_idm_type": 0.4789894104003906, "loss_sft": 0.014643288403749465, "step": 560, "token_acc": 0.8428118655908181 }, { "epoch": 0.13145945945945947, "grad_norm": 152.93589782714844, "learning_rate": 8.564642241456986e-06, "loss": 19.4894775390625, "loss_idm": 1.1359375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0201904296875, "loss_idm_text": 0.6324522256851196, "loss_idm_type": 0.5348370909690857, "loss_sft": 0.014856469631195069, "step": 570, "token_acc": 0.8417346370130199 }, { "epoch": 0.13376576576576577, "grad_norm": 123.00813293457031, "learning_rate": 8.506183921362443e-06, "loss": 18.439350891113282, "loss_idm": 1.046875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0206298828125, "loss_idm_text": 0.5719715774059295, "loss_idm_type": 0.5290378570556641, "loss_sft": 0.01490309163928032, "step": 580, "token_acc": 0.8404926826479667 }, { "epoch": 0.13607207207207206, "grad_norm": 123.43096923828125, "learning_rate": 8.446767045592829e-06, "loss": 19.559785461425783, "loss_idm": 1.1203125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.022509765625, "loss_idm_text": 0.6619894862174988, "loss_idm_type": 0.5079665780067444, "loss_sft": 0.014454486221075058, "step": 590, "token_acc": 0.8448515319710282 }, { "epoch": 0.13837837837837838, "grad_norm": 132.0702667236328, "learning_rate": 8.386407858128707e-06, "loss": 17.04791717529297, "loss_idm": 0.971875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0239501953125, "loss_idm_text": 0.5366034507751465, "loss_idm_type": 0.4742018222808838, "loss_sft": 0.014516119658946992, "step": 600, "token_acc": 0.8445947126557285 }, { "epoch": 0.14068468468468467, "grad_norm": 115.69120025634766, "learning_rate": 8.325122860569241e-06, "loss": 16.37723388671875, "loss_idm": 0.9421875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0227783203125, "loss_idm_text": 0.500444620847702, "loss_idm_type": 0.4702631950378418, "loss_sft": 0.014540355652570724, "step": 610, "token_acc": 0.8434274898223862 }, { "epoch": 0.142990990990991, "grad_norm": 130.52622985839844, "learning_rate": 8.262928807620843e-06, "loss": 17.805007934570312, "loss_idm": 1.028125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.02314453125, "loss_idm_text": 0.5241958260536194, "loss_idm_type": 0.5350467205047608, "loss_sft": 0.014786705374717712, "step": 620, "token_acc": 0.8413365725890637 }, { "epoch": 0.1452972972972973, "grad_norm": 164.35609436035156, "learning_rate": 8.199842702516584e-06, "loss": 17.97591552734375, "loss_idm": 1.0296875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0201904296875, "loss_idm_text": 0.5606541872024536, "loss_idm_type": 0.511587917804718, "loss_sft": 0.01503448113799095, "step": 630, "token_acc": 0.8400106109751245 }, { "epoch": 0.1476036036036036, "grad_norm": 156.69053649902344, "learning_rate": 8.135881792367686e-06, "loss": 19.665069580078125, "loss_idm": 1.153125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0185546875, "loss_idm_text": 0.6834821462631225, "loss_idm_type": 0.4970504879951477, "loss_sft": 0.014596361666917801, "step": 640, "token_acc": 0.843500234929801 }, { "epoch": 0.1499099099099099, "grad_norm": 135.3104705810547, "learning_rate": 8.071063563448341e-06, "loss": 18.526170349121095, "loss_idm": 1.0796875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.020263671875, "loss_idm_text": 0.6125914216041565, "loss_idm_type": 0.4950984835624695, "loss_sft": 0.014367320761084556, "step": 650, "token_acc": 0.8459901142882337 }, { "epoch": 0.15221621621621623, "grad_norm": 123.56324005126953, "learning_rate": 8.005405736415127e-06, "loss": 17.6597412109375, "loss_idm": 1.015625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0221435546875, "loss_idm_text": 0.5704326272010803, "loss_idm_type": 0.4809408187866211, "loss_sft": 0.014742322266101837, "step": 660, "token_acc": 0.8423076448143437 }, { "epoch": 0.15452252252252252, "grad_norm": 127.19869995117188, "learning_rate": 7.938926261462366e-06, "loss": 16.650668334960937, "loss_idm": 0.9515625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.019775390625, "loss_idm_text": 0.5215044260025025, "loss_idm_type": 0.46839267015457153, "loss_sft": 0.014879125729203224, "step": 670, "token_acc": 0.8405719579176686 }, { "epoch": 0.15682882882882884, "grad_norm": 165.220458984375, "learning_rate": 7.871643313414718e-06, "loss": 17.50257110595703, "loss_idm": 1.0015625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.022314453125, "loss_idm_text": 0.5456109881401062, "loss_idm_type": 0.4950627326965332, "loss_sft": 0.01498027853667736, "step": 680, "token_acc": 0.8399575879479895 }, { "epoch": 0.15913513513513514, "grad_norm": 161.6404266357422, "learning_rate": 7.803575286758365e-06, "loss": 17.011473083496092, "loss_idm": 0.9921875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0203857421875, "loss_idm_text": 0.5609789371490479, "loss_idm_type": 0.45157656669616697, "loss_sft": 0.014758187159895898, "step": 690, "token_acc": 0.8414021280985645 }, { "epoch": 0.16144144144144143, "grad_norm": 120.51724243164062, "learning_rate": 7.734740790612137e-06, "loss": 17.647825622558592, "loss_idm": 1.01875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0201904296875, "loss_idm_text": 0.5838465094566345, "loss_idm_type": 0.46862274408340454, "loss_sft": 0.014633599296212196, "step": 700, "token_acc": 0.8431996616079811 }, { "epoch": 0.16374774774774775, "grad_norm": 170.06082153320312, "learning_rate": 7.66515864363997e-06, "loss": 19.361347961425782, "loss_idm": 1.1140625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0193115234375, "loss_idm_text": 0.682057511806488, "loss_idm_type": 0.4779606580734253, "loss_sft": 0.014754829555749893, "step": 710, "token_acc": 0.8426819048755804 }, { "epoch": 0.16605405405405405, "grad_norm": 140.20143127441406, "learning_rate": 7.594847868906076e-06, "loss": 18.635955810546875, "loss_idm": 1.0734375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0216064453125, "loss_idm_text": 0.6446984887123108, "loss_idm_type": 0.46787831783294676, "loss_sft": 0.014802397787570953, "step": 720, "token_acc": 0.8416590932420016 }, { "epoch": 0.16836036036036037, "grad_norm": 130.18458557128906, "learning_rate": 7.52382768867422e-06, "loss": 18.703033447265625, "loss_idm": 1.078125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0206298828125, "loss_idm_text": 0.6053381562232971, "loss_idm_type": 0.5120768785476685, "loss_sft": 0.015017282590270043, "step": 730, "token_acc": 0.8400633789067389 }, { "epoch": 0.17066666666666666, "grad_norm": 141.50216674804688, "learning_rate": 7.452117519152542e-06, "loss": 17.169430541992188, "loss_idm": 0.9890625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.02119140625, "loss_idm_text": 0.5741037368774414, "loss_idm_type": 0.44670088291168214, "loss_sft": 0.014911935850977897, "step": 740, "token_acc": 0.8413064777939722 }, { "epoch": 0.17297297297297298, "grad_norm": 381.0265197753906, "learning_rate": 7.379736965185369e-06, "loss": 17.39481201171875, "loss_idm": 1.0078125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0203125, "loss_idm_text": 0.5736348688602447, "loss_idm_type": 0.4622136473655701, "loss_sft": 0.014967407658696174, "step": 750, "token_acc": 0.8404096603119668 }, { "epoch": 0.17527927927927928, "grad_norm": 99.33075714111328, "learning_rate": 7.30670581489344e-06, "loss": 17.591326904296874, "loss_idm": 1.015625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.01943359375, "loss_idm_text": 0.5623904466629028, "loss_idm_type": 0.48672530651092527, "loss_sft": 0.014822157844901086, "step": 760, "token_acc": 0.8409873327517785 }, { "epoch": 0.1775855855855856, "grad_norm": 167.3463897705078, "learning_rate": 7.233044034264034e-06, "loss": 17.634791564941406, "loss_idm": 1.0203125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.019189453125, "loss_idm_text": 0.5990640759468079, "loss_idm_type": 0.4531928539276123, "loss_sft": 0.01482548825442791, "step": 770, "token_acc": 0.842217710493471 }, { "epoch": 0.1798918918918919, "grad_norm": 180.26498413085938, "learning_rate": 7.158771761692464e-06, "loss": 16.831851196289062, "loss_idm": 0.9734375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0206787109375, "loss_idm_text": 0.5070213794708252, "loss_idm_type": 0.49378237724304197, "loss_sft": 0.0149398323148489, "step": 780, "token_acc": 0.8399789433151118 }, { "epoch": 0.18219819819819819, "grad_norm": 146.3783416748047, "learning_rate": 7.083909302476453e-06, "loss": 20.254476928710936, "loss_idm": 1.1703125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0190673828125, "loss_idm_text": 0.6638812899589539, "loss_idm_type": 0.5520286917686462, "loss_sft": 0.014968634769320487, "step": 790, "token_acc": 0.8404702738415722 }, { "epoch": 0.1845045045045045, "grad_norm": 131.1219482421875, "learning_rate": 7.008477123264849e-06, "loss": 17.661578369140624, "loss_idm": 1.028125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0191650390625, "loss_idm_text": 0.6021960020065308, "loss_idm_type": 0.45235241651535035, "loss_sft": 0.014716891944408417, "step": 800, "token_acc": 0.8420775256235287 }, { "epoch": 0.1868108108108108, "grad_norm": 103.17658233642578, "learning_rate": 6.932495846462262e-06, "loss": 20.432151794433594, "loss_idm": 1.184375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0184814453125, "loss_idm_text": 0.733825957775116, "loss_idm_type": 0.49434599876403806, "loss_sft": 0.014786238595843316, "step": 810, "token_acc": 0.8414228227929852 }, { "epoch": 0.18911711711711712, "grad_norm": 165.00164794921875, "learning_rate": 6.855986244591104e-06, "loss": 16.781680297851562, "loss_idm": 0.96875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0200927734375, "loss_idm_text": 0.574801790714264, "loss_idm_type": 0.4236711263656616, "loss_sft": 0.014908399060368538, "step": 820, "token_acc": 0.8403310689620388 }, { "epoch": 0.19142342342342342, "grad_norm": 148.9649200439453, "learning_rate": 6.778969234612583e-06, "loss": 18.193667602539062, "loss_idm": 1.0484375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.01845703125, "loss_idm_text": 0.6122974157333374, "loss_idm_type": 0.47533333897590635, "loss_sft": 0.014953377842903137, "step": 830, "token_acc": 0.8400144278851671 }, { "epoch": 0.19372972972972974, "grad_norm": 182.86627197265625, "learning_rate": 6.701465872208216e-06, "loss": 17.96783905029297, "loss_idm": 1.040625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.020166015625, "loss_idm_text": 0.6057761192321778, "loss_idm_type": 0.46598526239395144, "loss_sft": 0.015001512691378593, "step": 840, "token_acc": 0.8398306158449229 }, { "epoch": 0.19603603603603603, "grad_norm": 162.88619995117188, "learning_rate": 6.6234973460234184e-06, "loss": 14.121282958984375, "loss_idm": 0.809375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0208984375, "loss_idm_text": 0.45296228528022764, "loss_idm_type": 0.3782169342041016, "loss_sft": 0.01484096460044384, "step": 850, "token_acc": 0.8408548948297047 }, { "epoch": 0.19834234234234235, "grad_norm": 138.98036193847656, "learning_rate": 6.545084971874738e-06, "loss": 19.859603881835938, "loss_idm": 1.1203125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0204345703125, "loss_idm_text": 0.6227663040161133, "loss_idm_type": 0.5677083492279053, "loss_sft": 0.014872971922159195, "step": 860, "token_acc": 0.8419059541128763 }, { "epoch": 0.20064864864864865, "grad_norm": 114.0866470336914, "learning_rate": 6.466250186922325e-06, "loss": 17.332069396972656, "loss_idm": 1.0171875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.019140625, "loss_idm_text": 0.5944822430610657, "loss_idm_type": 0.4399122357368469, "loss_sft": 0.01457468681037426, "step": 870, "token_acc": 0.8431677371323297 }, { "epoch": 0.20295495495495494, "grad_norm": 120.35962677001953, "learning_rate": 6.387014543809224e-06, "loss": 18.15101318359375, "loss_idm": 1.04375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.020361328125, "loss_idm_text": 0.6401098132133484, "loss_idm_type": 0.4436539769172668, "loss_sft": 0.0147173710167408, "step": 880, "token_acc": 0.8424536391896238 }, { "epoch": 0.20526126126126126, "grad_norm": 174.99981689453125, "learning_rate": 6.3073997047691e-06, "loss": 16.543356323242186, "loss_idm": 0.9515625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.019189453125, "loss_idm_text": 0.5327290534973145, "loss_idm_type": 0.4514281630516052, "loss_sft": 0.014785801619291305, "step": 890, "token_acc": 0.8419114203024312 }, { "epoch": 0.20756756756756756, "grad_norm": 151.06468200683594, "learning_rate": 6.227427435703997e-06, "loss": 16.995733642578124, "loss_idm": 0.9875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.018994140625, "loss_idm_text": 0.5460891246795654, "loss_idm_type": 0.4672634840011597, "loss_sft": 0.014472068846225738, "step": 900, "token_acc": 0.8442095221030661 }, { "epoch": 0.20987387387387388, "grad_norm": 133.1666259765625, "learning_rate": 6.147119600233758e-06, "loss": 17.9807861328125, "loss_idm": 1.0484375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.020166015625, "loss_idm_text": 0.5863479137420654, "loss_idm_type": 0.48693227767944336, "loss_sft": 0.014701117947697639, "step": 910, "token_acc": 0.8430407477028895 }, { "epoch": 0.21218018018018017, "grad_norm": 184.44479370117188, "learning_rate": 6.066498153718735e-06, "loss": 16.594337463378906, "loss_idm": 0.9578125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0224365234375, "loss_idm_text": 0.551909339427948, "loss_idm_type": 0.4326519906520844, "loss_sft": 0.014737406745553017, "step": 920, "token_acc": 0.8421604346747479 }, { "epoch": 0.2144864864864865, "grad_norm": 169.61996459960938, "learning_rate": 5.985585137257401e-06, "loss": 17.667451477050783, "loss_idm": 1.0171875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0206298828125, "loss_idm_text": 0.6201748490333557, "loss_idm_type": 0.43307361006736755, "loss_sft": 0.014662803709506988, "step": 930, "token_acc": 0.8420754680619393 }, { "epoch": 0.2167927927927928, "grad_norm": 137.0713653564453, "learning_rate": 5.904402671660551e-06, "loss": 17.701638793945314, "loss_idm": 1.0296875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0200439453125, "loss_idm_text": 0.6406922459602356, "loss_idm_type": 0.41507343053817747, "loss_sft": 0.014637062698602677, "step": 940, "token_acc": 0.8429834188591938 }, { "epoch": 0.2190990990990991, "grad_norm": 122.10049438476562, "learning_rate": 5.82297295140367e-06, "loss": 15.998197937011719, "loss_idm": 0.92890625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0203857421875, "loss_idm_text": 0.5469140887260437, "loss_idm_type": 0.4025842845439911, "loss_sft": 0.014529398083686829, "step": 950, "token_acc": 0.8437657098716236 }, { "epoch": 0.2214054054054054, "grad_norm": 149.300537109375, "learning_rate": 5.74131823855921e-06, "loss": 18.87696990966797, "loss_idm": 1.103125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.01923828125, "loss_idm_text": 0.6745636343955994, "loss_idm_type": 0.456106698513031, "loss_sft": 0.014411784335970878, "step": 960, "token_acc": 0.8451125206797825 }, { "epoch": 0.22371171171171172, "grad_norm": 124.75553131103516, "learning_rate": 5.659460856710346e-06, "loss": 19.086871337890624, "loss_idm": 1.1125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0179443359375, "loss_idm_text": 0.7104646563529968, "loss_idm_type": 0.43523850440979006, "loss_sft": 0.014291927590966225, "step": 970, "token_acc": 0.8462941753717098 }, { "epoch": 0.22601801801801802, "grad_norm": 144.1509246826172, "learning_rate": 5.577423184847932e-06, "loss": 19.0544921875, "loss_idm": 1.103125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0209716796875, "loss_idm_text": 0.6754696369171143, "loss_idm_type": 0.4647406220436096, "loss_sft": 0.014486004412174226, "step": 980, "token_acc": 0.8444574023769905 }, { "epoch": 0.2283243243243243, "grad_norm": 148.3787841796875, "learning_rate": 5.495227651252315e-06, "loss": 17.188890075683595, "loss_idm": 0.990625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0214111328125, "loss_idm_text": 0.5891213119029999, "loss_idm_type": 0.4334494173526764, "loss_sft": 0.01478007510304451, "step": 990, "token_acc": 0.8416261870377664 }, { "epoch": 0.23063063063063063, "grad_norm": 134.9239044189453, "learning_rate": 5.412896727361663e-06, "loss": 15.146394348144531, "loss_idm": 0.87578125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0206298828125, "loss_idm_text": 0.5002685964107514, "loss_idm_type": 0.3955659091472626, "loss_sft": 0.014745334535837174, "step": 1000, "token_acc": 0.8416804747692777 }, { "epoch": 0.23063063063063063, "eval_loss": 1.2327880859375, "eval_runtime": 166.9771, "eval_samples_per_second": 10.451, "eval_steps_per_second": 2.617, "eval_token_acc": 0.8295414506468012, "step": 1000 }, { "epoch": 0.23293693693693693, "grad_norm": 124.12373352050781, "learning_rate": 5.3304529216284974e-06, "loss": 17.167878723144533, "loss_idm": 0.98125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0199951171875, "loss_idm_text": 0.5918982982635498, "loss_idm_type": 0.43085501790046693, "loss_sft": 0.014523088186979293, "step": 1010, "token_acc": 0.8444175480975747 }, { "epoch": 0.23524324324324325, "grad_norm": 131.99546813964844, "learning_rate": 5.247918773366112e-06, "loss": 14.159613037109375, "loss_idm": 0.81484375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.01875, "loss_idm_text": 0.4083675146102905, "loss_idm_type": 0.4275529682636261, "loss_sft": 0.0146888118237257, "step": 1020, "token_acc": 0.8415360933638757 }, { "epoch": 0.23754954954954954, "grad_norm": 116.93572235107422, "learning_rate": 5.165316846586541e-06, "loss": 15.603080749511719, "loss_idm": 0.90625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0185791015625, "loss_idm_text": 0.52706378698349, "loss_idm_type": 0.3998319745063782, "loss_sft": 0.01447979360818863, "step": 1030, "token_acc": 0.8434217713051085 }, { "epoch": 0.23985585585585587, "grad_norm": 186.3677215576172, "learning_rate": 5.082669723831793e-06, "loss": 17.206391906738283, "loss_idm": 0.9890625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0183837890625, "loss_idm_text": 0.5997193098068238, "loss_idm_type": 0.4266417622566223, "loss_sft": 0.014923496916890145, "step": 1040, "token_acc": 0.8400955159770253 }, { "epoch": 0.24216216216216216, "grad_norm": 134.91090393066406, "learning_rate": 5e-06, "loss": 14.17005615234375, "loss_idm": 0.81796875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.019482421875, "loss_idm_text": 0.45228556394577024, "loss_idm_type": 0.3835588574409485, "loss_sft": 0.014760823175311088, "step": 1050, "token_acc": 0.8408674793872597 }, { "epoch": 0.24446846846846848, "grad_norm": 156.60739135742188, "learning_rate": 4.917330276168208e-06, "loss": 15.966705322265625, "loss_idm": 0.9203125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0190185546875, "loss_idm_text": 0.5112396121025086, "loss_idm_type": 0.43749035596847535, "loss_sft": 0.014795352146029473, "step": 1060, "token_acc": 0.8414434859351285 }, { "epoch": 0.24677477477477477, "grad_norm": 109.64134979248047, "learning_rate": 4.8346831534134595e-06, "loss": 17.60172119140625, "loss_idm": 1.02109375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0205810546875, "loss_idm_text": 0.5611821413040161, "loss_idm_type": 0.4880287885665894, "loss_sft": 0.014694352447986603, "step": 1070, "token_acc": 0.8420225056411745 }, { "epoch": 0.24908108108108107, "grad_norm": 144.90110778808594, "learning_rate": 4.752081226633888e-06, "loss": 16.909609985351562, "loss_idm": 0.9734375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0178955078125, "loss_idm_text": 0.5911481857299805, "loss_idm_type": 0.4173463821411133, "loss_sft": 0.01456589177250862, "step": 1080, "token_acc": 0.8422988826546816 }, { "epoch": 0.25138738738738736, "grad_norm": 156.7343292236328, "learning_rate": 4.669547078371503e-06, "loss": 17.835157775878905, "loss_idm": 1.03125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.018505859375, "loss_idm_text": 0.602037787437439, "loss_idm_type": 0.4644427180290222, "loss_sft": 0.014457069337368011, "step": 1090, "token_acc": 0.8447727688647594 }, { "epoch": 0.2536936936936937, "grad_norm": 151.3989715576172, "learning_rate": 4.587103272638339e-06, "loss": 15.797918701171875, "loss_idm": 0.9109375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0177490234375, "loss_idm_text": 0.5759642958641052, "loss_idm_type": 0.3644200384616852, "loss_sft": 0.014347700774669648, "step": 1100, "token_acc": 0.8458367433542481 }, { "epoch": 0.256, "grad_norm": 154.04307556152344, "learning_rate": 4.504772348747687e-06, "loss": 16.6408203125, "loss_idm": 0.9640625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.017138671875, "loss_idm_text": 0.5653828263282776, "loss_idm_type": 0.42803622484207154, "loss_sft": 0.014483000338077544, "step": 1110, "token_acc": 0.8437569514130214 }, { "epoch": 0.2583063063063063, "grad_norm": 136.6959228515625, "learning_rate": 4.42257681515207e-06, "loss": 17.77167510986328, "loss_idm": 1.0453125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0193115234375, "loss_idm_text": 0.6610914945602417, "loss_idm_type": 0.400143039226532, "loss_sft": 0.01451752446591854, "step": 1120, "token_acc": 0.8445543010340107 }, { "epoch": 0.2606126126126126, "grad_norm": 135.94491577148438, "learning_rate": 4.340539143289655e-06, "loss": 17.628460693359376, "loss_idm": 1.0203125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.01865234375, "loss_idm_text": 0.6353346109390259, "loss_idm_type": 0.41818714141845703, "loss_sft": 0.014454802125692367, "step": 1130, "token_acc": 0.8443907936562738 }, { "epoch": 0.26291891891891894, "grad_norm": 160.8830108642578, "learning_rate": 4.25868176144079e-06, "loss": 16.29990234375, "loss_idm": 0.94375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0177001953125, "loss_idm_text": 0.541166114807129, "loss_idm_type": 0.4297985553741455, "loss_sft": 0.014698965847492218, "step": 1140, "token_acc": 0.842613082682653 }, { "epoch": 0.26522522522522524, "grad_norm": 135.12367248535156, "learning_rate": 4.17702704859633e-06, "loss": 16.195712280273437, "loss_idm": 0.9375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.020068359375, "loss_idm_text": 0.5394660949707031, "loss_idm_type": 0.42227545380592346, "loss_sft": 0.0147840466350317, "step": 1150, "token_acc": 0.8406737136218165 }, { "epoch": 0.26753153153153153, "grad_norm": 216.79522705078125, "learning_rate": 4.0955973283394525e-06, "loss": 15.386363220214843, "loss_idm": 0.88125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.01923828125, "loss_idm_text": 0.5185725450515747, "loss_idm_type": 0.3934124231338501, "loss_sft": 0.014546865224838256, "step": 1160, "token_acc": 0.8429609969438807 }, { "epoch": 0.2698378378378378, "grad_norm": 192.94996643066406, "learning_rate": 4.0144148627426e-06, "loss": 15.84566650390625, "loss_idm": 0.9125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.018603515625, "loss_idm_text": 0.5233206033706665, "loss_idm_type": 0.41824811697006226, "loss_sft": 0.014757906273007392, "step": 1170, "token_acc": 0.8413835079300053 }, { "epoch": 0.2721441441441441, "grad_norm": 136.8788604736328, "learning_rate": 3.9335018462812664e-06, "loss": 15.686166381835937, "loss_idm": 0.903125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0183349609375, "loss_idm_text": 0.45666194558143614, "loss_idm_type": 0.47452101707458494, "loss_sft": 0.014861632883548737, "step": 1180, "token_acc": 0.8401439821611449 }, { "epoch": 0.27445045045045047, "grad_norm": 144.01904296875, "learning_rate": 3.852880399766243e-06, "loss": 17.8814697265625, "loss_idm": 1.03125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.018603515625, "loss_idm_text": 0.6181171536445618, "loss_idm_type": 0.45043269991874696, "loss_sft": 0.014725609496235848, "step": 1190, "token_acc": 0.8415764879488441 }, { "epoch": 0.27675675675675676, "grad_norm": 142.98623657226562, "learning_rate": 3.7725725642960047e-06, "loss": 15.80457000732422, "loss_idm": 0.91484375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0194580078125, "loss_idm_text": 0.5567480087280273, "loss_idm_type": 0.38129279017448425, "loss_sft": 0.014737585186958313, "step": 1200, "token_acc": 0.8415397949338522 }, { "epoch": 0.27906306306306305, "grad_norm": 114.70137023925781, "learning_rate": 3.6926002952309015e-06, "loss": 15.290737915039063, "loss_idm": 0.88671875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0188232421875, "loss_idm_text": 0.515410315990448, "loss_idm_type": 0.3913223624229431, "loss_sft": 0.014572849497199059, "step": 1210, "token_acc": 0.8432653274893042 }, { "epoch": 0.28136936936936935, "grad_norm": 144.1276092529297, "learning_rate": 3.6129854561907786e-06, "loss": 12.707015991210938, "loss_idm": 0.73203125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0183349609375, "loss_idm_text": 0.3691279649734497, "loss_idm_type": 0.3762137830257416, "loss_sft": 0.014775990322232247, "step": 1220, "token_acc": 0.8407985566363351 }, { "epoch": 0.2836756756756757, "grad_norm": 162.1075897216797, "learning_rate": 3.533749813077677e-06, "loss": 17.085980224609376, "loss_idm": 0.98515625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0185546875, "loss_idm_text": 0.6291376233100892, "loss_idm_type": 0.390378075838089, "loss_sft": 0.014458349719643594, "step": 1230, "token_acc": 0.8444264451798582 }, { "epoch": 0.285981981981982, "grad_norm": 210.1119842529297, "learning_rate": 3.4549150281252635e-06, "loss": 15.181900024414062, "loss_idm": 0.8828125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0185791015625, "loss_idm_text": 0.5205450057983398, "loss_idm_type": 0.3800136804580688, "loss_sft": 0.014402059838175774, "step": 1240, "token_acc": 0.8441134821036184 }, { "epoch": 0.2882882882882883, "grad_norm": 138.51524353027344, "learning_rate": 3.3765026539765832e-06, "loss": 16.950105285644533, "loss_idm": 0.9875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0195556640625, "loss_idm_text": 0.6216296911239624, "loss_idm_type": 0.3894159376621246, "loss_sft": 0.014125967025756836, "step": 1250, "token_acc": 0.847919528627195 }, { "epoch": 0.2905945945945946, "grad_norm": 154.9911346435547, "learning_rate": 3.298534127791785e-06, "loss": 16.449783325195312, "loss_idm": 0.959375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0198974609375, "loss_idm_text": 0.557077956199646, "loss_idm_type": 0.42165766954421996, "loss_sft": 0.014437104761600494, "step": 1260, "token_acc": 0.8440432417820339 }, { "epoch": 0.2929009009009009, "grad_norm": 126.49798583984375, "learning_rate": 3.2210307653874175e-06, "loss": 14.849807739257812, "loss_idm": 0.85703125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.016943359375, "loss_idm_text": 0.5337203025817872, "loss_idm_type": 0.3474113941192627, "loss_sft": 0.014366917312145233, "step": 1270, "token_acc": 0.8457429629704389 }, { "epoch": 0.2952072072072072, "grad_norm": 162.35890197753906, "learning_rate": 3.1440137554088957e-06, "loss": 15.314067077636718, "loss_idm": 0.88125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0182373046875, "loss_idm_text": 0.5182104647159577, "loss_idm_type": 0.39087520241737367, "loss_sft": 0.014491885155439376, "step": 1280, "token_acc": 0.84312589885166 }, { "epoch": 0.2975135135135135, "grad_norm": 156.34283447265625, "learning_rate": 3.06750415353774e-06, "loss": 15.587838745117187, "loss_idm": 0.903125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.017919921875, "loss_idm_text": 0.5396368145942688, "loss_idm_type": 0.3871441423892975, "loss_sft": 0.014515075832605362, "step": 1290, "token_acc": 0.8436580424785217 }, { "epoch": 0.2998198198198198, "grad_norm": 169.38795471191406, "learning_rate": 2.991522876735154e-06, "loss": 15.711705017089844, "loss_idm": 0.909375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.018310546875, "loss_idm_text": 0.5583156704902649, "loss_idm_type": 0.3756734192371368, "loss_sft": 0.014504583552479744, "step": 1300, "token_acc": 0.8438711647805225 }, { "epoch": 0.3021261261261261, "grad_norm": 165.150146484375, "learning_rate": 2.9160906975235493e-06, "loss": 13.628689575195313, "loss_idm": 0.78671875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0202880859375, "loss_idm_text": 0.4506654143333435, "loss_idm_type": 0.35060195326805116, "loss_sft": 0.014570535719394683, "step": 1310, "token_acc": 0.8431417376648077 }, { "epoch": 0.30443243243243245, "grad_norm": 199.06549072265625, "learning_rate": 2.8412282383075362e-06, "loss": 16.37169647216797, "loss_idm": 0.9515625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0171875, "loss_idm_text": 0.6052956461906434, "loss_idm_type": 0.37099899649620055, "loss_sft": 0.01451174020767212, "step": 1320, "token_acc": 0.8432796170522042 }, { "epoch": 0.30673873873873875, "grad_norm": 162.7277374267578, "learning_rate": 2.766955965735968e-06, "loss": 14.637554931640626, "loss_idm": 0.85078125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0188232421875, "loss_idm_text": 0.5082914412021637, "loss_idm_type": 0.3572277069091797, "loss_sft": 0.014705703407526017, "step": 1330, "token_acc": 0.8422121601269894 }, { "epoch": 0.30904504504504504, "grad_norm": 143.47894287109375, "learning_rate": 2.693294185106562e-06, "loss": 13.652301025390624, "loss_idm": 0.7828125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0198486328125, "loss_idm_text": 0.4210363507270813, "loss_idm_type": 0.3825237274169922, "loss_sft": 0.014496737718582153, "step": 1340, "token_acc": 0.8437972852010349 }, { "epoch": 0.31135135135135134, "grad_norm": 132.78858947753906, "learning_rate": 2.6202630348146323e-06, "loss": 15.083595275878906, "loss_idm": 0.8796875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0176025390625, "loss_idm_text": 0.5048262238502502, "loss_idm_type": 0.39085535407066346, "loss_sft": 0.01434645541012287, "step": 1350, "token_acc": 0.8459678060808526 }, { "epoch": 0.3136576576576577, "grad_norm": 165.5104217529297, "learning_rate": 2.5478824808474613e-06, "loss": 15.311534118652343, "loss_idm": 0.87421875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0175537109375, "loss_idm_text": 0.5322446703910828, "loss_idm_type": 0.37764355540275574, "loss_sft": 0.014268488809466361, "step": 1360, "token_acc": 0.8452123559853058 }, { "epoch": 0.315963963963964, "grad_norm": 141.0616455078125, "learning_rate": 2.476172311325783e-06, "loss": 15.717620849609375, "loss_idm": 0.915625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0179931640625, "loss_idm_text": 0.5490303754806518, "loss_idm_type": 0.38535616993904115, "loss_sft": 0.014547385275363922, "step": 1370, "token_acc": 0.8435834472048124 }, { "epoch": 0.3182702702702703, "grad_norm": 202.39620971679688, "learning_rate": 2.4051521310939258e-06, "loss": 16.194735717773437, "loss_idm": 0.9453125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.016943359375, "loss_idm_text": 0.5614462852478027, "loss_idm_type": 0.40362446904182436, "loss_sft": 0.014665166661143304, "step": 1380, "token_acc": 0.8426359432769256 }, { "epoch": 0.32057657657657657, "grad_norm": 186.8247833251953, "learning_rate": 2.3348413563600324e-06, "loss": 14.730172729492187, "loss_idm": 0.8421875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0188720703125, "loss_idm_text": 0.4475443482398987, "loss_idm_type": 0.42451398968696596, "loss_sft": 0.014449363574385643, "step": 1390, "token_acc": 0.8445057561070892 }, { "epoch": 0.32288288288288286, "grad_norm": 114.01448822021484, "learning_rate": 2.265259209387867e-06, "loss": 14.568830871582032, "loss_idm": 0.8296875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0177490234375, "loss_idm_text": 0.4782296776771545, "loss_idm_type": 0.38517374396324155, "loss_sft": 0.014272304996848107, "step": 1400, "token_acc": 0.8455116168894466 }, { "epoch": 0.3251891891891892, "grad_norm": 152.14208984375, "learning_rate": 2.1964247132416373e-06, "loss": 15.284608459472656, "loss_idm": 0.8859375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0169677734375, "loss_idm_text": 0.5515658617019653, "loss_idm_type": 0.35711957812309264, "loss_sft": 0.014497144892811776, "step": 1410, "token_acc": 0.8435979398066908 }, { "epoch": 0.3274954954954955, "grad_norm": 137.8850555419922, "learning_rate": 2.1283566865852824e-06, "loss": 14.024693298339844, "loss_idm": 0.80703125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.018408203125, "loss_idm_text": 0.48794840574264525, "loss_idm_type": 0.34030611515045167, "loss_sft": 0.014438638836145401, "step": 1420, "token_acc": 0.8444949198905822 }, { "epoch": 0.3298018018018018, "grad_norm": 193.08645629882812, "learning_rate": 2.061073738537635e-06, "loss": 14.654190063476562, "loss_idm": 0.84921875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.02021484375, "loss_idm_text": 0.5063535213470459, "loss_idm_type": 0.35956308245658875, "loss_sft": 0.014374737441539765, "step": 1430, "token_acc": 0.8457109989803655 }, { "epoch": 0.3321081081081081, "grad_norm": 179.3224639892578, "learning_rate": 1.9945942635848745e-06, "loss": 15.624932861328125, "loss_idm": 0.90234375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.019384765625, "loss_idm_text": 0.5428369641304016, "loss_idm_type": 0.38439915180206297, "loss_sft": 0.014417222514748574, "step": 1440, "token_acc": 0.845333890181186 }, { "epoch": 0.33441441441441444, "grad_norm": 154.86846923828125, "learning_rate": 1.928936436551661e-06, "loss": 17.012544250488283, "loss_idm": 0.9953125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.018603515625, "loss_idm_text": 0.63501957654953, "loss_idm_type": 0.38010231852531434, "loss_sft": 0.014432965219020844, "step": 1450, "token_acc": 0.8441891544960063 }, { "epoch": 0.33672072072072073, "grad_norm": 135.02999877929688, "learning_rate": 1.864118207632315e-06, "loss": 14.688017272949219, "loss_idm": 0.8484375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0168701171875, "loss_idm_text": 0.5101109147071838, "loss_idm_type": 0.36122512221336367, "loss_sft": 0.014461452886462212, "step": 1460, "token_acc": 0.8441797287309384 }, { "epoch": 0.33902702702702703, "grad_norm": 161.4497833251953, "learning_rate": 1.8001572974834169e-06, "loss": 14.854217529296875, "loss_idm": 0.8578125, "loss_idm_anl": 0.0, "loss_idm_coord": 0.019482421875, "loss_idm_text": 0.5380716323852539, "loss_idm_type": 0.341589891910553, "loss_sft": 0.014205808565020562, "step": 1470, "token_acc": 0.8471874874954983 }, { "epoch": 0.3413333333333333, "grad_norm": 133.96571350097656, "learning_rate": 1.7370711923791567e-06, "loss": 13.000572204589844, "loss_idm": 0.7375, "loss_idm_anl": 0.0, "loss_idm_coord": 0.019384765625, "loss_idm_text": 0.4172104477882385, "loss_idm_type": 0.34592602252960203, "loss_sft": 0.014315107464790344, "step": 1480, "token_acc": 0.8460156840421251 }, { "epoch": 0.3436396396396396, "grad_norm": 120.62774658203125, "learning_rate": 1.6748771394307584e-06, "loss": 14.395199584960938, "loss_idm": 0.83671875, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0180419921875, "loss_idm_text": 0.4689497768878937, "loss_idm_type": 0.3827894330024719, "loss_sft": 0.014493611082434655, "step": 1490, "token_acc": 0.8434915816796925 }, { "epoch": 0.34594594594594597, "grad_norm": 160.82275390625, "learning_rate": 1.6135921418712959e-06, "loss": 14.267446899414063, "loss_idm": 0.8265625, "loss_idm_anl": 0.0, "loss_idm_coord": 0.0183837890625, "loss_idm_text": 0.486546927690506, "loss_idm_type": 0.357412725687027, "loss_sft": 0.014240951836109161, "step": 1500, "token_acc": 0.8465638485857505 }, { "epoch": 0.34594594594594597, "eval_loss": 1.1403998136520386, "eval_runtime": 168.4218, "eval_samples_per_second": 10.361, "eval_steps_per_second": 2.595, "eval_token_acc": 0.8319841652873965, "step": 1500 } ], "logging_steps": 10, "max_steps": 2000, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 1.8759920794589987e+19, "train_batch_size": 1, "trial_name": null, "trial_params": null }