{
  "best_global_step": 192,
  "best_metric": 0.69560647,
  "best_model_checkpoint": "/mmu_mllm_hdd_2/yifanzhang/models/tool_final/qwen_tool_all_data_180k_alldata_wogemini_retool2k_mmeu10k_filter_tuihuo_lr1e_6/v0-20250618-155155/checkpoint-192",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 192,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.010478061558611657,
      "grad_norm": 5.3835768699646,
      "learning_rate": 6.666666666666667e-08,
      "loss": 1.0226306915283203,
      "memory(GiB)": 44.4,
      "step": 1,
      "token_acc": 0.7902787219578518,
      "train_speed(iter/s)": 0.023134
    },
    {
      "epoch": 0.05239030779305828,
      "grad_norm": 5.482320785522461,
      "learning_rate": 3.333333333333333e-07,
      "loss": 0.9853244423866272,
      "memory(GiB)": 44.41,
      "step": 5,
      "token_acc": 0.7981000574329044,
      "train_speed(iter/s)": 0.063148
    },
    {
      "epoch": 0.10478061558611657,
      "grad_norm": 5.281559467315674,
      "learning_rate": 6.666666666666666e-07,
      "loss": 0.9907465934753418,
      "memory(GiB)": 44.41,
      "step": 10,
      "token_acc": 0.7960847551280461,
      "train_speed(iter/s)": 0.079762
    },
    {
      "epoch": 0.15717092337917485,
      "grad_norm": 4.199541091918945,
      "learning_rate": 1e-06,
      "loss": 0.9603083610534668,
      "memory(GiB)": 44.41,
      "step": 15,
      "token_acc": 0.7847730042656916,
      "train_speed(iter/s)": 0.086676
    },
    {
      "epoch": 0.20956123117223313,
      "grad_norm": 2.6114139556884766,
      "learning_rate": 9.991540791356342e-07,
      "loss": 0.911494255065918,
      "memory(GiB)": 44.41,
      "step": 20,
      "token_acc": 0.7907820762593932,
      "train_speed(iter/s)": 0.091839
    },
    {
      "epoch": 0.26195153896529144,
      "grad_norm": 1.4692460298538208,
      "learning_rate": 9.966191788709714e-07,
      "loss": 0.86248779296875,
      "memory(GiB)": 44.41,
      "step": 25,
      "token_acc": 0.7835635250284222,
      "train_speed(iter/s)": 0.094886
    },
    {
      "epoch": 0.3143418467583497,
      "grad_norm": 1.2658460140228271,
      "learning_rate": 9.92403876506104e-07,
      "loss": 0.8281326293945312,
      "memory(GiB)": 44.41,
      "step": 30,
      "token_acc": 0.7954776456301068,
      "train_speed(iter/s)": 0.096695
    },
    {
      "epoch": 0.366732154551408,
      "grad_norm": 1.1135852336883545,
      "learning_rate": 9.865224352899118e-07,
      "loss": 0.785943078994751,
      "memory(GiB)": 44.41,
      "step": 35,
      "token_acc": 0.7947409577076513,
      "train_speed(iter/s)": 0.098384
    },
    {
      "epoch": 0.41912246234446626,
      "grad_norm": 0.934022068977356,
      "learning_rate": 9.789947561577443e-07,
      "loss": 0.7608952045440673,
      "memory(GiB)": 44.41,
      "step": 40,
      "token_acc": 0.8277342667909038,
      "train_speed(iter/s)": 0.099693
    },
    {
      "epoch": 0.4715127701375246,
      "grad_norm": 0.933360755443573,
      "learning_rate": 9.698463103929541e-07,
      "loss": 0.7670727252960206,
      "memory(GiB)": 58.74,
      "step": 45,
      "token_acc": 0.7954774159967614,
      "train_speed(iter/s)": 0.100167
    },
    {
      "epoch": 0.5239030779305829,
      "grad_norm": 0.7834925055503845,
      "learning_rate": 9.591080534401371e-07,
      "loss": 0.7326742172241211,
      "memory(GiB)": 58.74,
      "step": 50,
      "token_acc": 0.804578482890403,
      "train_speed(iter/s)": 0.100862
    },
    {
      "epoch": 0.5762933857236411,
      "grad_norm": 0.642005980014801,
      "learning_rate": 9.468163201617061e-07,
      "loss": 0.7098912715911865,
      "memory(GiB)": 58.74,
      "step": 55,
      "token_acc": 0.8023350941984947,
      "train_speed(iter/s)": 0.101457
    },
    {
      "epoch": 0.6286836935166994,
      "grad_norm": 0.5564186573028564,
      "learning_rate": 9.330127018922193e-07,
      "loss": 0.7002639770507812,
      "memory(GiB)": 58.74,
      "step": 60,
      "token_acc": 0.8084159021406728,
      "train_speed(iter/s)": 0.102039
    },
    {
      "epoch": 0.6810740013097577,
      "grad_norm": 0.5595128536224365,
      "learning_rate": 9.177439057064682e-07,
      "loss": 0.7075435161590576,
      "memory(GiB)": 58.74,
      "step": 65,
      "token_acc": 0.8167705488246142,
      "train_speed(iter/s)": 0.102488
    },
    {
      "epoch": 0.733464309102816,
      "grad_norm": 0.5548214912414551,
      "learning_rate": 9.010615963775219e-07,
      "loss": 0.7494614601135254,
      "memory(GiB)": 58.74,
      "step": 70,
      "token_acc": 0.7996152406037366,
      "train_speed(iter/s)": 0.102826
    },
    {
      "epoch": 0.7858546168958742,
      "grad_norm": 0.5225480198860168,
      "learning_rate": 8.83022221559489e-07,
      "loss": 0.7125186920166016,
      "memory(GiB)": 58.74,
      "step": 75,
      "token_acc": 0.8273133275875578,
      "train_speed(iter/s)": 0.103148
    },
    {
      "epoch": 0.8382449246889325,
      "grad_norm": 0.5410817265510559,
      "learning_rate": 8.636868207865243e-07,
      "loss": 0.7034761905670166,
      "memory(GiB)": 58.74,
      "step": 80,
      "token_acc": 0.8277207704682812,
      "train_speed(iter/s)": 0.103508
    },
    {
      "epoch": 0.8906352324819908,
      "grad_norm": 0.51569002866745,
      "learning_rate": 8.431208189343669e-07,
      "loss": 0.6858974933624268,
      "memory(GiB)": 58.74,
      "step": 85,
      "token_acc": 0.8288310690146258,
      "train_speed(iter/s)": 0.103837
    },
    {
      "epoch": 0.9430255402750491,
      "grad_norm": 0.49336692690849304,
      "learning_rate": 8.213938048432696e-07,
      "loss": 0.6981162071228028,
      "memory(GiB)": 58.74,
      "step": 90,
      "token_acc": 0.8163834897478095,
      "train_speed(iter/s)": 0.103802
    },
    {
      "epoch": 0.9954158480681073,
      "grad_norm": 0.4902209937572479,
      "learning_rate": 7.985792958513931e-07,
      "loss": 0.6680845260620117,
      "memory(GiB)": 58.74,
      "step": 95,
      "token_acc": 0.8077775923320856,
      "train_speed(iter/s)": 0.10403
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.7270359396934509,
      "eval_runtime": 3.812,
      "eval_samples_per_second": 32.004,
      "eval_steps_per_second": 4.197,
      "eval_token_acc": 0.8126405885265615,
      "step": 96
    },
    {
      "epoch": 1.0419122462344466,
      "grad_norm": 0.5008202791213989,
      "learning_rate": 7.74754489035403e-07,
      "loss": 0.6931419372558594,
      "memory(GiB)": 69.7,
      "step": 100,
      "token_acc": 0.8228939404602716,
      "train_speed(iter/s)": 0.097155
    },
    {
      "epoch": 1.0943025540275049,
      "grad_norm": 0.4630364179611206,
      "learning_rate": 7.5e-07,
      "loss": 0.6557597160339356,
      "memory(GiB)": 69.7,
      "step": 105,
      "token_acc": 0.8266370917514396,
      "train_speed(iter/s)": 0.097434
    },
    {
      "epoch": 1.1466928618205632,
      "grad_norm": 0.5154017806053162,
      "learning_rate": 7.243995901002311e-07,
      "loss": 0.655954122543335,
      "memory(GiB)": 69.7,
      "step": 110,
      "token_acc": 0.8196414769752425,
      "train_speed(iter/s)": 0.097691
    },
    {
      "epoch": 1.1990831696136215,
      "grad_norm": 0.48749059438705444,
      "learning_rate": 6.980398830195784e-07,
      "loss": 0.6581990242004394,
      "memory(GiB)": 69.7,
      "step": 115,
      "token_acc": 0.828079317895458,
      "train_speed(iter/s)": 0.098246
    },
    {
      "epoch": 1.2514734774066798,
      "grad_norm": 0.4727058410644531,
      "learning_rate": 6.710100716628344e-07,
      "loss": 0.6710573196411133,
      "memory(GiB)": 69.7,
      "step": 120,
      "token_acc": 0.82493355998755,
      "train_speed(iter/s)": 0.098554
    },
    {
      "epoch": 1.3038637851997381,
      "grad_norm": 0.4838932454586029,
      "learning_rate": 6.434016163555451e-07,
      "loss": 0.6274348258972168,
      "memory(GiB)": 72.64,
      "step": 125,
      "token_acc": 0.8356775725442703,
      "train_speed(iter/s)": 0.09868
    },
    {
      "epoch": 1.3562540929927964,
      "grad_norm": 0.5088004469871521,
      "learning_rate": 6.153079353712201e-07,
      "loss": 0.6917521476745605,
      "memory(GiB)": 72.64,
      "step": 130,
      "token_acc": 0.8195788933110797,
      "train_speed(iter/s)": 0.098996
    },
    {
      "epoch": 1.4086444007858545,
      "grad_norm": 0.43357038497924805,
      "learning_rate": 5.868240888334652e-07,
      "loss": 0.6407961845397949,
      "memory(GiB)": 72.64,
      "step": 135,
      "token_acc": 0.8241987694284514,
      "train_speed(iter/s)": 0.099344
    },
    {
      "epoch": 1.4610347085789128,
      "grad_norm": 0.46167683601379395,
      "learning_rate": 5.580464570626151e-07,
      "loss": 0.6401975631713868,
      "memory(GiB)": 72.64,
      "step": 140,
      "token_acc": 0.8140715160475303,
      "train_speed(iter/s)": 0.099505
    },
    {
      "epoch": 1.5134250163719711,
      "grad_norm": 0.48245716094970703,
      "learning_rate": 5.290724144552379e-07,
      "loss": 0.6458414554595947,
      "memory(GiB)": 72.64,
      "step": 145,
      "token_acc": 0.829897042257211,
      "train_speed(iter/s)": 0.099836
    },
    {
      "epoch": 1.5658153241650294,
      "grad_norm": 0.4646553695201874,
      "learning_rate": 5e-07,
      "loss": 0.6638086318969727,
      "memory(GiB)": 72.64,
      "step": 150,
      "token_acc": 0.8305597185117612,
      "train_speed(iter/s)": 0.100123
    },
    {
      "epoch": 1.6182056319580878,
      "grad_norm": 0.4738137423992157,
      "learning_rate": 4.7092758554476206e-07,
      "loss": 0.65560941696167,
      "memory(GiB)": 72.64,
      "step": 155,
      "token_acc": 0.835213900041534,
      "train_speed(iter/s)": 0.100234
    },
    {
      "epoch": 1.670595939751146,
      "grad_norm": 0.4571834206581116,
      "learning_rate": 4.419535429373848e-07,
      "loss": 0.6428651809692383,
      "memory(GiB)": 72.64,
      "step": 160,
      "token_acc": 0.8090443928554508,
      "train_speed(iter/s)": 0.100516
    },
    {
      "epoch": 1.7229862475442044,
      "grad_norm": 0.4719304144382477,
      "learning_rate": 4.131759111665348e-07,
      "loss": 0.6572139739990235,
      "memory(GiB)": 72.64,
      "step": 165,
      "token_acc": 0.8091730066675509,
      "train_speed(iter/s)": 0.100744
    },
    {
      "epoch": 1.7753765553372625,
      "grad_norm": 0.4634987711906433,
      "learning_rate": 3.846920646287799e-07,
      "loss": 0.6705328464508057,
      "memory(GiB)": 72.64,
      "step": 170,
      "token_acc": 0.8068579338115012,
      "train_speed(iter/s)": 0.100944
    },
    {
      "epoch": 1.827766863130321,
      "grad_norm": 0.4913578927516937,
      "learning_rate": 3.56598383644455e-07,
      "loss": 0.6417989730834961,
      "memory(GiB)": 72.64,
      "step": 175,
      "token_acc": 0.8260864515829327,
      "train_speed(iter/s)": 0.10128
    },
    {
      "epoch": 1.880157170923379,
      "grad_norm": 0.44139742851257324,
      "learning_rate": 3.2898992833716563e-07,
      "loss": 0.6202288150787354,
      "memory(GiB)": 72.64,
      "step": 180,
      "token_acc": 0.8227838827838828,
      "train_speed(iter/s)": 0.101534
    },
    {
      "epoch": 1.9325474787164376,
      "grad_norm": 0.5277097821235657,
      "learning_rate": 3.0196011698042156e-07,
      "loss": 0.69473876953125,
      "memory(GiB)": 72.64,
      "step": 185,
      "token_acc": 0.8144716112785779,
      "train_speed(iter/s)": 0.101837
    },
    {
      "epoch": 1.9849377865094957,
      "grad_norm": 0.469308465719223,
      "learning_rate": 2.756004098997689e-07,
      "loss": 0.6366086006164551,
      "memory(GiB)": 72.64,
      "step": 190,
      "token_acc": 0.8070857224599759,
      "train_speed(iter/s)": 0.102045
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.6956064701080322,
      "eval_runtime": 3.8275,
      "eval_samples_per_second": 31.874,
      "eval_steps_per_second": 4.18,
      "eval_token_acc": 0.8187026950555281,
      "step": 192
    }
  ],
  "logging_steps": 5,
  "max_steps": 285,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 1.9820351920086712e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}