{
  "args": {
    "model": "andreasmartin/apertus-v1.5-8b-text",
    "data_dir": "data/full3",
    "dataset_repo": "None",
    "out": "runs/decider3",
    "epochs": "1.0",
    "lr": "0.0001",
    "batch": "8",
    "accumulation": "4",
    "max_length": "768",
    "lora_r": "32",
    "lora_alpha": "32",
    "lora_dropout": "0.05",
    "seed": "0",
    "template_kwargs": "{}",
    "hub_adapter": "blackfish021/apertus-lightpath-decider-v3-e2",
    "hub_merged": "None",
    "dry_run": "False",
    "no_4bit": "False",
    "limit_train": "0",
    "limit_heldout": "0",
    "max_steps": "-1",
    "skip_merge": "True",
    "save_merged": "False",
    "eval_records": "200",
    "save_steps": "100",
    "resume": "True",
    "resume_dir": "None",
    "die_at_step": "0",
    "predict_batch": "32",
    "max_new_tokens": "16",
    "init_adapter": "blackfish021/apertus-lightpath-decider-v3"
  },
  "resumed_from": null,
  "metrics": {
    "train_runtime": 3817.0186,
    "train_samples_per_second": 3.144,
    "train_steps_per_second": 0.098,
    "total_flos": 2.7898729612959744e+17,
    "train_loss": 0.01657172168294589,
    "epoch": 1.0
  },
  "log": [
    {
      "loss": 0.022410409152507783,
      "grad_norm": 0.31484320759773254,
      "learning_rate": 4.736842105263158e-05,
      "epoch": 0.02666666666666667,
      "step": 10
    },
    {
      "loss": 0.025908684730529784,
      "grad_norm": 0.12033871561288834,
      "learning_rate": 0.0001,
      "epoch": 0.05333333333333334,
      "step": 20
    },
    {
      "loss": 0.033026579022407535,
      "grad_norm": 0.2651446461677551,
      "learning_rate": 9.980543805476446e-05,
      "epoch": 0.08,
      "step": 30
    },
    {
      "loss": 0.018129727244377135,
      "grad_norm": 0.10938066244125366,
      "learning_rate": 9.922326639307917e-05,
      "epoch": 0.10666666666666667,
      "step": 40
    },
    {
      "loss": 0.021136780083179475,
      "grad_norm": 0.023365505039691925,
      "learning_rate": 9.825801575298248e-05,
      "epoch": 0.13333333333333333,
      "step": 50
    },
    {
      "loss": 0.01657077521085739,
      "grad_norm": 0.0763176754117012,
      "learning_rate": 9.691719817616147e-05,
      "epoch": 0.16,
      "step": 60
    },
    {
      "loss": 0.01797577887773514,
      "grad_norm": 0.2973942756652832,
      "learning_rate": 9.521124854565425e-05,
      "epoch": 0.18666666666666668,
      "step": 70
    },
    {
      "loss": 0.027637118101119997,
      "grad_norm": 0.08629462122917175,
      "learning_rate": 9.315344337660421e-05,
      "epoch": 0.21333333333333335,
      "step": 80
    },
    {
      "loss": 0.019443552196025848,
      "grad_norm": 0.06489592045545578,
      "learning_rate": 9.075979749207561e-05,
      "epoch": 0.24,
      "step": 90
    },
    {
      "loss": 0.021283210813999177,
      "grad_norm": 0.217499241232872,
      "learning_rate": 8.80489393880484e-05,
      "epoch": 0.26666666666666666,
      "step": 100
    },
    {
      "eval_loss": 0.032591983675956726,
      "eval_runtime": 17.6588,
      "eval_samples_per_second": 11.326,
      "eval_steps_per_second": 1.416,
      "epoch": 0.26666666666666666,
      "step": 100
    },
    {
      "loss": 0.014644630253314972,
      "grad_norm": 0.172653466463089,
      "learning_rate": 8.504196625756166e-05,
      "epoch": 0.29333333333333333,
      "step": 110
    },
    {
      "loss": 0.014805102348327636,
      "grad_norm": 0.10182410478591919,
      "learning_rate": 8.176227980227694e-05,
      "epoch": 0.32,
      "step": 120
    },
    {
      "loss": 0.014555910229682922,
      "grad_norm": 0.17371761798858643,
      "learning_rate": 7.823540410925435e-05,
      "epoch": 0.3466666666666667,
      "step": 130
    },
    {
      "loss": 0.016314414143562318,
      "grad_norm": 0.022570302709937096,
      "learning_rate": 7.448878701031142e-05,
      "epoch": 0.37333333333333335,
      "step": 140
    },
    {
      "loss": 0.01566801071166992,
      "grad_norm": 0.2855406701564789,
      "learning_rate": 7.055158646988109e-05,
      "epoch": 0.4,
      "step": 150
    },
    {
      "loss": 0.018659861385822298,
      "grad_norm": 0.11542737483978271,
      "learning_rate": 6.64544436638005e-05,
      "epoch": 0.4266666666666667,
      "step": 160
    },
    {
      "loss": 0.013596612215042114,
      "grad_norm": 0.16074834764003754,
      "learning_rate": 6.222924451504001e-05,
      "epoch": 0.4533333333333333,
      "step": 170
    },
    {
      "loss": 0.020363916456699372,
      "grad_norm": 0.04103618487715721,
      "learning_rate": 5.79088715422152e-05,
      "epoch": 0.48,
      "step": 180
    },
    {
      "loss": 0.014784714579582215,
      "grad_norm": 0.13508240878582,
      "learning_rate": 5.352694795211555e-05,
      "epoch": 0.5066666666666667,
      "step": 190
    },
    {
      "loss": 0.008975905925035476,
      "grad_norm": 0.028504809364676476,
      "learning_rate": 4.911757596784357e-05,
      "epoch": 0.5333333333333333,
      "step": 200
    },
    {
      "eval_loss": 0.026406440883874893,
      "eval_runtime": 17.6523,
      "eval_samples_per_second": 11.33,
      "eval_steps_per_second": 1.416,
      "epoch": 0.5333333333333333,
      "step": 200
    },
    {
      "loss": 0.017752039432525634,
      "grad_norm": 0.07383177429437637,
      "learning_rate": 4.471507142902036e-05,
      "epoch": 0.56,
      "step": 210
    },
    {
      "loss": 0.011054360121488572,
      "grad_norm": 0.16143670678138733,
      "learning_rate": 4.035369672952516e-05,
      "epoch": 0.5866666666666667,
      "step": 220
    },
    {
      "loss": 0.010980958491563797,
      "grad_norm": 0.12218093127012253,
      "learning_rate": 3.6067394171175394e-05,
      "epoch": 0.6133333333333333,
      "step": 230
    },
    {
      "loss": 0.012676803767681122,
      "grad_norm": 0.24066784977912903,
      "learning_rate": 3.188952180851589e-05,
      "epoch": 0.64,
      "step": 240
    },
    {
      "loss": 0.011154239624738693,
      "grad_norm": 0.15749654173851013,
      "learning_rate": 2.785259384049959e-05,
      "epoch": 0.6666666666666666,
      "step": 250
    },
    {
      "loss": 0.01723717600107193,
      "grad_norm": 0.12799593806266785,
      "learning_rate": 2.3988027569455895e-05,
      "epoch": 0.6933333333333334,
      "step": 260
    },
    {
      "loss": 0.01156546175479889,
      "grad_norm": 0.1376238912343979,
      "learning_rate": 2.0325898896632177e-05,
      "epoch": 0.72,
      "step": 270
    },
    {
      "loss": 0.013197347521781921,
      "grad_norm": 0.2866266965866089,
      "learning_rate": 1.689470825715998e-05,
      "epoch": 0.7466666666666667,
      "step": 280
    },
    {
      "loss": 0.014573681354522704,
      "grad_norm": 0.02412877231836319,
      "learning_rate": 1.3721158816050873e-05,
      "epoch": 0.7733333333333333,
      "step": 290
    },
    {
      "loss": 0.025699126720428466,
      "grad_norm": 0.1795499622821808,
      "learning_rate": 1.0829948651407374e-05,
      "epoch": 0.8,
      "step": 300
    },
    {
      "eval_loss": 0.026917733252048492,
      "eval_runtime": 17.6625,
      "eval_samples_per_second": 11.323,
      "eval_steps_per_second": 1.415,
      "epoch": 0.8,
      "step": 300
    },
    {
      "loss": 0.020623017847537995,
      "grad_norm": 0.1386154145002365,
      "learning_rate": 8.243578542178226e-06,
      "epoch": 0.8266666666666667,
      "step": 310
    },
    {
      "loss": 0.012144556641578675,
      "grad_norm": 0.20349618792533875,
      "learning_rate": 5.982176856345445e-06,
      "epoch": 0.8533333333333334,
      "step": 320
    },
    {
      "loss": 0.02048766314983368,
      "grad_norm": 0.24547676742076874,
      "learning_rate": 4.0633429023472e-06,
      "epoch": 0.88,
      "step": 330
    },
    {
      "loss": 0.012237246334552764,
      "grad_norm": 0.07478773593902588,
      "learning_rate": 2.50200996285046e-06,
      "epoch": 0.9066666666666666,
      "step": 340
    },
    {
      "loss": 0.009454116225242615,
      "grad_norm": 0.1210315078496933,
      "learning_rate": 1.3103290768099797e-06,
      "epoch": 0.9333333333333333,
      "step": 350
    },
    {
      "loss": 0.009869556128978729,
      "grad_norm": 0.0482204295694828,
      "learning_rate": 4.975744742772848e-07,
      "epoch": 0.96,
      "step": 360
    },
    {
      "loss": 0.010671104490756988,
      "grad_norm": 0.08062546700239182,
      "learning_rate": 7.007139991108135e-08,
      "epoch": 0.9866666666666667,
      "step": 370
    },
    {
      "eval_loss": 0.024143896996974945,
      "eval_runtime": 17.6568,
      "eval_samples_per_second": 11.327,
      "eval_steps_per_second": 1.416,
      "epoch": 1.0,
      "step": 375
    },
    {
      "train_runtime": 3817.0186,
      "train_samples_per_second": 3.144,
      "train_steps_per_second": 0.098,
      "total_flos": 2.7898729612959744e+17,
      "train_loss": 0.01657172168294589,
      "epoch": 1.0,
      "step": 375
    }
  ],
  "peak_gpu_gb": 17.891663074493408
}