{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 3188,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006273525721455458,
      "grad_norm": 4.55400276184082,
      "learning_rate": 0.0,
      "loss": 1.5808,
      "memory/device_mem_reserved(gib)": 63.4,
      "memory/max_mem_active(gib)": 59.02,
      "memory/max_mem_allocated(gib)": 57.84,
      "step": 1
    },
    {
      "epoch": 0.0012547051442910915,
      "grad_norm": 7.68040132522583,
      "learning_rate": 1.5723270440251573e-08,
      "loss": 1.474,
      "memory/device_mem_reserved(gib)": 66.65,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2
    },
    {
      "epoch": 0.0018820577164366374,
      "grad_norm": 10.7412691116333,
      "learning_rate": 3.1446540880503146e-08,
      "loss": 1.5559,
      "memory/device_mem_reserved(gib)": 66.65,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3
    },
    {
      "epoch": 0.002509410288582183,
      "grad_norm": 11.039576530456543,
      "learning_rate": 4.716981132075472e-08,
      "loss": 1.4446,
      "memory/device_mem_reserved(gib)": 66.65,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 4
    },
    {
      "epoch": 0.003136762860727729,
      "grad_norm": 6.22249174118042,
      "learning_rate": 6.289308176100629e-08,
      "loss": 1.515,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 5
    },
    {
      "epoch": 0.0037641154328732747,
      "grad_norm": 9.491832733154297,
      "learning_rate": 7.861635220125787e-08,
      "loss": 1.6619,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 6
    },
    {
      "epoch": 0.00439146800501882,
      "grad_norm": 5.316587924957275,
      "learning_rate": 9.433962264150944e-08,
      "loss": 1.4984,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 7
    },
    {
      "epoch": 0.005018820577164366,
      "grad_norm": 10.27668285369873,
      "learning_rate": 1.1006289308176101e-07,
      "loss": 1.459,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 8
    },
    {
      "epoch": 0.0056461731493099125,
      "grad_norm": 10.63105297088623,
      "learning_rate": 1.2578616352201258e-07,
      "loss": 1.5212,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 9
    },
    {
      "epoch": 0.006273525721455458,
      "grad_norm": 10.851506233215332,
      "learning_rate": 1.4150943396226417e-07,
      "loss": 1.5022,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 10
    },
    {
      "epoch": 0.006900878293601004,
      "grad_norm": 11.121560096740723,
      "learning_rate": 1.5723270440251575e-07,
      "loss": 1.4822,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 11
    },
    {
      "epoch": 0.0075282308657465494,
      "grad_norm": 11.112455368041992,
      "learning_rate": 1.729559748427673e-07,
      "loss": 1.4818,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 12
    },
    {
      "epoch": 0.008155583437892095,
      "grad_norm": 10.980988502502441,
      "learning_rate": 1.886792452830189e-07,
      "loss": 1.5752,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 13
    },
    {
      "epoch": 0.00878293601003764,
      "grad_norm": 11.245274543762207,
      "learning_rate": 2.0440251572327044e-07,
      "loss": 1.6465,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 14
    },
    {
      "epoch": 0.009410288582183186,
      "grad_norm": 8.604133605957031,
      "learning_rate": 2.2012578616352203e-07,
      "loss": 1.6738,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 15
    },
    {
      "epoch": 0.010037641154328732,
      "grad_norm": 10.810846328735352,
      "learning_rate": 2.358490566037736e-07,
      "loss": 1.6457,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 16
    },
    {
      "epoch": 0.01066499372647428,
      "grad_norm": 10.287492752075195,
      "learning_rate": 2.5157232704402517e-07,
      "loss": 1.5768,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 17
    },
    {
      "epoch": 0.011292346298619825,
      "grad_norm": 6.959763050079346,
      "learning_rate": 2.672955974842768e-07,
      "loss": 1.5683,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 18
    },
    {
      "epoch": 0.01191969887076537,
      "grad_norm": 12.073090553283691,
      "learning_rate": 2.8301886792452833e-07,
      "loss": 1.5883,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 19
    },
    {
      "epoch": 0.012547051442910916,
      "grad_norm": 12.099584579467773,
      "learning_rate": 2.987421383647799e-07,
      "loss": 1.5578,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 20
    },
    {
      "epoch": 0.013174404015056462,
      "grad_norm": 9.198031425476074,
      "learning_rate": 3.144654088050315e-07,
      "loss": 1.5087,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 21
    },
    {
      "epoch": 0.013801756587202008,
      "grad_norm": 10.079482078552246,
      "learning_rate": 3.3018867924528305e-07,
      "loss": 1.5111,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 22
    },
    {
      "epoch": 0.014429109159347553,
      "grad_norm": 12.30569076538086,
      "learning_rate": 3.459119496855346e-07,
      "loss": 1.6295,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 23
    },
    {
      "epoch": 0.015056461731493099,
      "grad_norm": 12.143986701965332,
      "learning_rate": 3.6163522012578617e-07,
      "loss": 1.6378,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 24
    },
    {
      "epoch": 0.015683814303638646,
      "grad_norm": 10.82630729675293,
      "learning_rate": 3.773584905660378e-07,
      "loss": 1.6042,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 25
    },
    {
      "epoch": 0.01631116687578419,
      "grad_norm": 10.716633796691895,
      "learning_rate": 3.9308176100628933e-07,
      "loss": 1.6034,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 26
    },
    {
      "epoch": 0.016938519447929738,
      "grad_norm": 11.65971565246582,
      "learning_rate": 4.088050314465409e-07,
      "loss": 1.5007,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 27
    },
    {
      "epoch": 0.01756587202007528,
      "grad_norm": 8.470008850097656,
      "learning_rate": 4.2452830188679244e-07,
      "loss": 1.671,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 28
    },
    {
      "epoch": 0.01819322459222083,
      "grad_norm": 8.622583389282227,
      "learning_rate": 4.4025157232704405e-07,
      "loss": 1.4135,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 29
    },
    {
      "epoch": 0.018820577164366373,
      "grad_norm": 8.237770080566406,
      "learning_rate": 4.559748427672956e-07,
      "loss": 1.5056,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 30
    },
    {
      "epoch": 0.01944792973651192,
      "grad_norm": 10.933939933776855,
      "learning_rate": 4.716981132075472e-07,
      "loss": 1.5356,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 31
    },
    {
      "epoch": 0.020075282308657464,
      "grad_norm": 11.000118255615234,
      "learning_rate": 4.874213836477988e-07,
      "loss": 1.5723,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 32
    },
    {
      "epoch": 0.02070263488080301,
      "grad_norm": 9.162322044372559,
      "learning_rate": 5.031446540880503e-07,
      "loss": 1.5293,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 33
    },
    {
      "epoch": 0.02132998745294856,
      "grad_norm": 9.679073333740234,
      "learning_rate": 5.18867924528302e-07,
      "loss": 1.6125,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 34
    },
    {
      "epoch": 0.021957340025094103,
      "grad_norm": 8.7017183303833,
      "learning_rate": 5.345911949685535e-07,
      "loss": 1.6079,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 35
    },
    {
      "epoch": 0.02258469259723965,
      "grad_norm": 7.953933238983154,
      "learning_rate": 5.503144654088051e-07,
      "loss": 1.4798,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 36
    },
    {
      "epoch": 0.023212045169385194,
      "grad_norm": 8.445432662963867,
      "learning_rate": 5.660377358490567e-07,
      "loss": 1.4155,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 37
    },
    {
      "epoch": 0.02383939774153074,
      "grad_norm": 9.527459144592285,
      "learning_rate": 5.817610062893082e-07,
      "loss": 1.556,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 38
    },
    {
      "epoch": 0.024466750313676285,
      "grad_norm": 9.838929176330566,
      "learning_rate": 5.974842767295598e-07,
      "loss": 1.6388,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 39
    },
    {
      "epoch": 0.025094102885821833,
      "grad_norm": 7.818563461303711,
      "learning_rate": 6.132075471698113e-07,
      "loss": 1.6555,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 40
    },
    {
      "epoch": 0.025721455457967377,
      "grad_norm": 6.137913227081299,
      "learning_rate": 6.28930817610063e-07,
      "loss": 1.4206,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 41
    },
    {
      "epoch": 0.026348808030112924,
      "grad_norm": 9.544657707214355,
      "learning_rate": 6.446540880503145e-07,
      "loss": 1.5458,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 42
    },
    {
      "epoch": 0.026976160602258468,
      "grad_norm": 9.101583480834961,
      "learning_rate": 6.603773584905661e-07,
      "loss": 1.6298,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 43
    },
    {
      "epoch": 0.027603513174404015,
      "grad_norm": 6.088347434997559,
      "learning_rate": 6.761006289308177e-07,
      "loss": 1.5088,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 44
    },
    {
      "epoch": 0.028230865746549563,
      "grad_norm": 8.540587425231934,
      "learning_rate": 6.918238993710692e-07,
      "loss": 1.4617,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 45
    },
    {
      "epoch": 0.028858218318695106,
      "grad_norm": 8.382627487182617,
      "learning_rate": 7.075471698113208e-07,
      "loss": 1.5994,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 46
    },
    {
      "epoch": 0.029485570890840654,
      "grad_norm": 8.526695251464844,
      "learning_rate": 7.232704402515723e-07,
      "loss": 1.4959,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 47
    },
    {
      "epoch": 0.030112923462986198,
      "grad_norm": 4.549665451049805,
      "learning_rate": 7.389937106918239e-07,
      "loss": 1.3477,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 48
    },
    {
      "epoch": 0.030740276035131745,
      "grad_norm": 6.675533771514893,
      "learning_rate": 7.547169811320755e-07,
      "loss": 1.3748,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 49
    },
    {
      "epoch": 0.03136762860727729,
      "grad_norm": 6.710190773010254,
      "learning_rate": 7.704402515723271e-07,
      "loss": 1.4118,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 50
    },
    {
      "epoch": 0.031994981179422836,
      "grad_norm": 5.698192119598389,
      "learning_rate": 7.861635220125787e-07,
      "loss": 1.5622,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 51
    },
    {
      "epoch": 0.03262233375156838,
      "grad_norm": 5.854632377624512,
      "learning_rate": 8.018867924528302e-07,
      "loss": 1.4148,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 52
    },
    {
      "epoch": 0.033249686323713924,
      "grad_norm": 5.814241409301758,
      "learning_rate": 8.176100628930818e-07,
      "loss": 1.4957,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 53
    },
    {
      "epoch": 0.033877038895859475,
      "grad_norm": 5.345610618591309,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.5477,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 54
    },
    {
      "epoch": 0.03450439146800502,
      "grad_norm": 4.159165859222412,
      "learning_rate": 8.490566037735849e-07,
      "loss": 1.6153,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 55
    },
    {
      "epoch": 0.03513174404015056,
      "grad_norm": 4.605583667755127,
      "learning_rate": 8.647798742138364e-07,
      "loss": 1.5115,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 56
    },
    {
      "epoch": 0.035759096612296114,
      "grad_norm": 4.393484115600586,
      "learning_rate": 8.805031446540881e-07,
      "loss": 1.4937,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 57
    },
    {
      "epoch": 0.03638644918444166,
      "grad_norm": 4.844731330871582,
      "learning_rate": 8.962264150943397e-07,
      "loss": 1.4964,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 58
    },
    {
      "epoch": 0.0370138017565872,
      "grad_norm": 5.815843105316162,
      "learning_rate": 9.119496855345912e-07,
      "loss": 1.3776,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 59
    },
    {
      "epoch": 0.037641154328732745,
      "grad_norm": 4.103832721710205,
      "learning_rate": 9.276729559748428e-07,
      "loss": 1.5472,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 60
    },
    {
      "epoch": 0.038268506900878296,
      "grad_norm": 4.0624003410339355,
      "learning_rate": 9.433962264150944e-07,
      "loss": 1.4647,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 61
    },
    {
      "epoch": 0.03889585947302384,
      "grad_norm": 3.943577527999878,
      "learning_rate": 9.59119496855346e-07,
      "loss": 1.4935,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 62
    },
    {
      "epoch": 0.039523212045169384,
      "grad_norm": 3.511507034301758,
      "learning_rate": 9.748427672955975e-07,
      "loss": 1.481,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 63
    },
    {
      "epoch": 0.04015056461731493,
      "grad_norm": 3.368314027786255,
      "learning_rate": 9.90566037735849e-07,
      "loss": 1.436,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 64
    },
    {
      "epoch": 0.04077791718946048,
      "grad_norm": 6.15948486328125,
      "learning_rate": 1.0062893081761007e-06,
      "loss": 1.4033,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 65
    },
    {
      "epoch": 0.04140526976160602,
      "grad_norm": 3.0553529262542725,
      "learning_rate": 1.0220125786163524e-06,
      "loss": 1.4416,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 66
    },
    {
      "epoch": 0.04203262233375157,
      "grad_norm": 3.3963119983673096,
      "learning_rate": 1.037735849056604e-06,
      "loss": 1.3565,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 67
    },
    {
      "epoch": 0.04265997490589712,
      "grad_norm": 3.019916534423828,
      "learning_rate": 1.0534591194968555e-06,
      "loss": 1.5527,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 68
    },
    {
      "epoch": 0.04328732747804266,
      "grad_norm": 2.330040693283081,
      "learning_rate": 1.069182389937107e-06,
      "loss": 1.3976,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 69
    },
    {
      "epoch": 0.043914680050188205,
      "grad_norm": 3.104090929031372,
      "learning_rate": 1.0849056603773587e-06,
      "loss": 1.3887,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 70
    },
    {
      "epoch": 0.04454203262233375,
      "grad_norm": 2.3439829349517822,
      "learning_rate": 1.1006289308176102e-06,
      "loss": 1.3351,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 71
    },
    {
      "epoch": 0.0451693851944793,
      "grad_norm": 2.272207021713257,
      "learning_rate": 1.1163522012578618e-06,
      "loss": 1.4638,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 72
    },
    {
      "epoch": 0.045796737766624844,
      "grad_norm": 3.1712934970855713,
      "learning_rate": 1.1320754716981133e-06,
      "loss": 1.4946,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 73
    },
    {
      "epoch": 0.04642409033877039,
      "grad_norm": 2.8735315799713135,
      "learning_rate": 1.1477987421383649e-06,
      "loss": 1.3761,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 74
    },
    {
      "epoch": 0.04705144291091593,
      "grad_norm": 2.4971110820770264,
      "learning_rate": 1.1635220125786164e-06,
      "loss": 1.3517,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 75
    },
    {
      "epoch": 0.04767879548306148,
      "grad_norm": 3.0293681621551514,
      "learning_rate": 1.179245283018868e-06,
      "loss": 1.3868,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 76
    },
    {
      "epoch": 0.04830614805520703,
      "grad_norm": 2.4946517944335938,
      "learning_rate": 1.1949685534591195e-06,
      "loss": 1.432,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 77
    },
    {
      "epoch": 0.04893350062735257,
      "grad_norm": 2.731900691986084,
      "learning_rate": 1.210691823899371e-06,
      "loss": 1.4344,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 78
    },
    {
      "epoch": 0.04956085319949812,
      "grad_norm": 2.6176462173461914,
      "learning_rate": 1.2264150943396227e-06,
      "loss": 1.5104,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 79
    },
    {
      "epoch": 0.050188205771643665,
      "grad_norm": 2.2624993324279785,
      "learning_rate": 1.2421383647798742e-06,
      "loss": 1.3613,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 80
    },
    {
      "epoch": 0.05081555834378921,
      "grad_norm": 2.183051347732544,
      "learning_rate": 1.257861635220126e-06,
      "loss": 1.466,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 81
    },
    {
      "epoch": 0.05144291091593475,
      "grad_norm": 2.4126477241516113,
      "learning_rate": 1.2735849056603775e-06,
      "loss": 1.4675,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 82
    },
    {
      "epoch": 0.052070263488080304,
      "grad_norm": 2.0408334732055664,
      "learning_rate": 1.289308176100629e-06,
      "loss": 1.5206,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 83
    },
    {
      "epoch": 0.05269761606022585,
      "grad_norm": 2.421983242034912,
      "learning_rate": 1.3050314465408807e-06,
      "loss": 1.3339,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 84
    },
    {
      "epoch": 0.05332496863237139,
      "grad_norm": 2.067976236343384,
      "learning_rate": 1.3207547169811322e-06,
      "loss": 1.2658,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 85
    },
    {
      "epoch": 0.053952321204516936,
      "grad_norm": 2.595181465148926,
      "learning_rate": 1.3364779874213838e-06,
      "loss": 1.4028,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 86
    },
    {
      "epoch": 0.054579673776662486,
      "grad_norm": 2.261770486831665,
      "learning_rate": 1.3522012578616353e-06,
      "loss": 1.4594,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 87
    },
    {
      "epoch": 0.05520702634880803,
      "grad_norm": 1.8053139448165894,
      "learning_rate": 1.3679245283018869e-06,
      "loss": 1.4086,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 88
    },
    {
      "epoch": 0.055834378920953574,
      "grad_norm": 2.1403069496154785,
      "learning_rate": 1.3836477987421384e-06,
      "loss": 1.4465,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 89
    },
    {
      "epoch": 0.056461731493099125,
      "grad_norm": 1.8229817152023315,
      "learning_rate": 1.39937106918239e-06,
      "loss": 1.2884,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 90
    },
    {
      "epoch": 0.05708908406524467,
      "grad_norm": 2.258371591567993,
      "learning_rate": 1.4150943396226415e-06,
      "loss": 1.5038,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 91
    },
    {
      "epoch": 0.05771643663739021,
      "grad_norm": 2.1894209384918213,
      "learning_rate": 1.430817610062893e-06,
      "loss": 1.4794,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 92
    },
    {
      "epoch": 0.05834378920953576,
      "grad_norm": 2.3533987998962402,
      "learning_rate": 1.4465408805031447e-06,
      "loss": 1.44,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 93
    },
    {
      "epoch": 0.05897114178168131,
      "grad_norm": 2.299332618713379,
      "learning_rate": 1.4622641509433962e-06,
      "loss": 1.3913,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 94
    },
    {
      "epoch": 0.05959849435382685,
      "grad_norm": 4.128480911254883,
      "learning_rate": 1.4779874213836478e-06,
      "loss": 1.3982,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 95
    },
    {
      "epoch": 0.060225846925972396,
      "grad_norm": 2.248939037322998,
      "learning_rate": 1.4937106918238995e-06,
      "loss": 1.483,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 96
    },
    {
      "epoch": 0.06085319949811794,
      "grad_norm": 1.1961462497711182,
      "learning_rate": 1.509433962264151e-06,
      "loss": 1.2478,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 97
    },
    {
      "epoch": 0.06148055207026349,
      "grad_norm": 2.1051137447357178,
      "learning_rate": 1.5251572327044027e-06,
      "loss": 1.3615,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 98
    },
    {
      "epoch": 0.062107904642409034,
      "grad_norm": 2.1795237064361572,
      "learning_rate": 1.5408805031446542e-06,
      "loss": 1.4017,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 99
    },
    {
      "epoch": 0.06273525721455459,
      "grad_norm": 1.4975708723068237,
      "learning_rate": 1.5566037735849058e-06,
      "loss": 1.2858,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 100
    },
    {
      "epoch": 0.06336260978670012,
      "grad_norm": 2.0895345211029053,
      "learning_rate": 1.5723270440251573e-06,
      "loss": 1.3706,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 101
    },
    {
      "epoch": 0.06398996235884567,
      "grad_norm": 2.0322093963623047,
      "learning_rate": 1.5880503144654089e-06,
      "loss": 1.3236,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 102
    },
    {
      "epoch": 0.06461731493099122,
      "grad_norm": 2.1316373348236084,
      "learning_rate": 1.6037735849056604e-06,
      "loss": 1.4114,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 103
    },
    {
      "epoch": 0.06524466750313676,
      "grad_norm": 1.7316523790359497,
      "learning_rate": 1.619496855345912e-06,
      "loss": 1.2191,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 104
    },
    {
      "epoch": 0.06587202007528231,
      "grad_norm": 2.152735710144043,
      "learning_rate": 1.6352201257861635e-06,
      "loss": 1.3812,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 105
    },
    {
      "epoch": 0.06649937264742785,
      "grad_norm": 1.5850223302841187,
      "learning_rate": 1.650943396226415e-06,
      "loss": 1.1812,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 106
    },
    {
      "epoch": 0.0671267252195734,
      "grad_norm": 1.7916295528411865,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.4622,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 107
    },
    {
      "epoch": 0.06775407779171895,
      "grad_norm": 2.106351375579834,
      "learning_rate": 1.6823899371069182e-06,
      "loss": 1.4152,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 108
    },
    {
      "epoch": 0.06838143036386449,
      "grad_norm": 2.268888235092163,
      "learning_rate": 1.6981132075471698e-06,
      "loss": 1.2981,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 109
    },
    {
      "epoch": 0.06900878293601004,
      "grad_norm": 1.7028484344482422,
      "learning_rate": 1.7138364779874213e-06,
      "loss": 1.5788,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 110
    },
    {
      "epoch": 0.06963613550815559,
      "grad_norm": 1.8516031503677368,
      "learning_rate": 1.7295597484276729e-06,
      "loss": 1.4352,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 111
    },
    {
      "epoch": 0.07026348808030113,
      "grad_norm": 2.121943235397339,
      "learning_rate": 1.7452830188679247e-06,
      "loss": 1.3917,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 112
    },
    {
      "epoch": 0.07089084065244668,
      "grad_norm": 2.1828010082244873,
      "learning_rate": 1.7610062893081762e-06,
      "loss": 1.322,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 113
    },
    {
      "epoch": 0.07151819322459223,
      "grad_norm": 2.0953001976013184,
      "learning_rate": 1.7767295597484278e-06,
      "loss": 1.3502,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 114
    },
    {
      "epoch": 0.07214554579673776,
      "grad_norm": 1.852333426475525,
      "learning_rate": 1.7924528301886793e-06,
      "loss": 1.2831,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 115
    },
    {
      "epoch": 0.07277289836888332,
      "grad_norm": 5.071216106414795,
      "learning_rate": 1.8081761006289309e-06,
      "loss": 1.4053,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 116
    },
    {
      "epoch": 0.07340025094102885,
      "grad_norm": 1.746447205543518,
      "learning_rate": 1.8238993710691824e-06,
      "loss": 1.3858,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 117
    },
    {
      "epoch": 0.0740276035131744,
      "grad_norm": 1.3724796772003174,
      "learning_rate": 1.839622641509434e-06,
      "loss": 1.4171,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 118
    },
    {
      "epoch": 0.07465495608531995,
      "grad_norm": 1.8758927583694458,
      "learning_rate": 1.8553459119496855e-06,
      "loss": 1.4434,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 119
    },
    {
      "epoch": 0.07528230865746549,
      "grad_norm": 1.7063547372817993,
      "learning_rate": 1.871069182389937e-06,
      "loss": 1.3846,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 120
    },
    {
      "epoch": 0.07590966122961104,
      "grad_norm": 1.6393623352050781,
      "learning_rate": 1.8867924528301889e-06,
      "loss": 1.4012,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 121
    },
    {
      "epoch": 0.07653701380175659,
      "grad_norm": 1.9628385305404663,
      "learning_rate": 1.9025157232704406e-06,
      "loss": 1.4207,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 122
    },
    {
      "epoch": 0.07716436637390213,
      "grad_norm": 1.8672785758972168,
      "learning_rate": 1.918238993710692e-06,
      "loss": 1.3259,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 123
    },
    {
      "epoch": 0.07779171894604768,
      "grad_norm": 1.5363185405731201,
      "learning_rate": 1.9339622641509438e-06,
      "loss": 1.3228,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 124
    },
    {
      "epoch": 0.07841907151819323,
      "grad_norm": 2.0102760791778564,
      "learning_rate": 1.949685534591195e-06,
      "loss": 1.4722,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 125
    },
    {
      "epoch": 0.07904642409033877,
      "grad_norm": 2.0831358432769775,
      "learning_rate": 1.965408805031447e-06,
      "loss": 1.3728,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 126
    },
    {
      "epoch": 0.07967377666248432,
      "grad_norm": 1.8129521608352661,
      "learning_rate": 1.981132075471698e-06,
      "loss": 1.4643,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 127
    },
    {
      "epoch": 0.08030112923462986,
      "grad_norm": 4.2833380699157715,
      "learning_rate": 1.99685534591195e-06,
      "loss": 1.266,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 128
    },
    {
      "epoch": 0.0809284818067754,
      "grad_norm": 2.023083448410034,
      "learning_rate": 2.0125786163522013e-06,
      "loss": 1.3666,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 129
    },
    {
      "epoch": 0.08155583437892096,
      "grad_norm": 2.003934144973755,
      "learning_rate": 2.028301886792453e-06,
      "loss": 1.3563,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 130
    },
    {
      "epoch": 0.0821831869510665,
      "grad_norm": 1.747253179550171,
      "learning_rate": 2.044025157232705e-06,
      "loss": 1.3771,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 131
    },
    {
      "epoch": 0.08281053952321205,
      "grad_norm": 1.769203782081604,
      "learning_rate": 2.059748427672956e-06,
      "loss": 1.3521,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 132
    },
    {
      "epoch": 0.0834378920953576,
      "grad_norm": 2.124824285507202,
      "learning_rate": 2.075471698113208e-06,
      "loss": 1.5221,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 133
    },
    {
      "epoch": 0.08406524466750313,
      "grad_norm": 2.0400173664093018,
      "learning_rate": 2.0911949685534593e-06,
      "loss": 1.3352,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 134
    },
    {
      "epoch": 0.08469259723964868,
      "grad_norm": 1.806525468826294,
      "learning_rate": 2.106918238993711e-06,
      "loss": 1.2829,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 135
    },
    {
      "epoch": 0.08531994981179424,
      "grad_norm": 1.6417590379714966,
      "learning_rate": 2.1226415094339624e-06,
      "loss": 1.4122,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 136
    },
    {
      "epoch": 0.08594730238393977,
      "grad_norm": 1.7979159355163574,
      "learning_rate": 2.138364779874214e-06,
      "loss": 1.3852,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 137
    },
    {
      "epoch": 0.08657465495608532,
      "grad_norm": 1.5453232526779175,
      "learning_rate": 2.1540880503144655e-06,
      "loss": 1.2895,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 138
    },
    {
      "epoch": 0.08720200752823086,
      "grad_norm": 2.0955095291137695,
      "learning_rate": 2.1698113207547173e-06,
      "loss": 1.2654,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 139
    },
    {
      "epoch": 0.08782936010037641,
      "grad_norm": 1.9225586652755737,
      "learning_rate": 2.1855345911949687e-06,
      "loss": 1.3794,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 140
    },
    {
      "epoch": 0.08845671267252196,
      "grad_norm": 1.9370880126953125,
      "learning_rate": 2.2012578616352204e-06,
      "loss": 1.2875,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 141
    },
    {
      "epoch": 0.0890840652446675,
      "grad_norm": 1.9599952697753906,
      "learning_rate": 2.2169811320754718e-06,
      "loss": 1.3845,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 142
    },
    {
      "epoch": 0.08971141781681305,
      "grad_norm": 1.8032351732254028,
      "learning_rate": 2.2327044025157235e-06,
      "loss": 1.2829,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 143
    },
    {
      "epoch": 0.0903387703889586,
      "grad_norm": 1.683026671409607,
      "learning_rate": 2.248427672955975e-06,
      "loss": 1.3368,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 144
    },
    {
      "epoch": 0.09096612296110414,
      "grad_norm": 1.9588282108306885,
      "learning_rate": 2.2641509433962266e-06,
      "loss": 1.5689,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 145
    },
    {
      "epoch": 0.09159347553324969,
      "grad_norm": 1.8645468950271606,
      "learning_rate": 2.2798742138364784e-06,
      "loss": 1.4278,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 146
    },
    {
      "epoch": 0.09222082810539524,
      "grad_norm": 1.889079213142395,
      "learning_rate": 2.2955974842767298e-06,
      "loss": 1.2831,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 147
    },
    {
      "epoch": 0.09284818067754078,
      "grad_norm": 1.615443468093872,
      "learning_rate": 2.3113207547169815e-06,
      "loss": 1.2778,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 148
    },
    {
      "epoch": 0.09347553324968633,
      "grad_norm": 2.068561315536499,
      "learning_rate": 2.327044025157233e-06,
      "loss": 1.3043,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 149
    },
    {
      "epoch": 0.09410288582183186,
      "grad_norm": 2.1261227130889893,
      "learning_rate": 2.3427672955974846e-06,
      "loss": 1.3721,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 150
    },
    {
      "epoch": 0.09473023839397741,
      "grad_norm": 1.621612310409546,
      "learning_rate": 2.358490566037736e-06,
      "loss": 1.4434,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 151
    },
    {
      "epoch": 0.09535759096612297,
      "grad_norm": 1.688284993171692,
      "learning_rate": 2.3742138364779878e-06,
      "loss": 1.2438,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 152
    },
    {
      "epoch": 0.0959849435382685,
      "grad_norm": 1.7449889183044434,
      "learning_rate": 2.389937106918239e-06,
      "loss": 1.2775,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 153
    },
    {
      "epoch": 0.09661229611041405,
      "grad_norm": 1.9485979080200195,
      "learning_rate": 2.405660377358491e-06,
      "loss": 1.3454,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 154
    },
    {
      "epoch": 0.0972396486825596,
      "grad_norm": 1.9239331483840942,
      "learning_rate": 2.421383647798742e-06,
      "loss": 1.3666,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 155
    },
    {
      "epoch": 0.09786700125470514,
      "grad_norm": 1.554479956626892,
      "learning_rate": 2.437106918238994e-06,
      "loss": 1.4377,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 156
    },
    {
      "epoch": 0.09849435382685069,
      "grad_norm": 1.6431803703308105,
      "learning_rate": 2.4528301886792453e-06,
      "loss": 1.3351,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 157
    },
    {
      "epoch": 0.09912170639899624,
      "grad_norm": 2.31290340423584,
      "learning_rate": 2.468553459119497e-06,
      "loss": 1.4224,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 158
    },
    {
      "epoch": 0.09974905897114178,
      "grad_norm": 1.8860207796096802,
      "learning_rate": 2.4842767295597484e-06,
      "loss": 1.3442,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 159
    },
    {
      "epoch": 0.10037641154328733,
      "grad_norm": 7.125235080718994,
      "learning_rate": 2.5e-06,
      "loss": 1.471,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 160
    },
    {
      "epoch": 0.10100376411543287,
      "grad_norm": 1.5071229934692383,
      "learning_rate": 2.515723270440252e-06,
      "loss": 1.2873,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 161
    },
    {
      "epoch": 0.10163111668757842,
      "grad_norm": 1.965579867362976,
      "learning_rate": 2.5314465408805033e-06,
      "loss": 1.5114,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 162
    },
    {
      "epoch": 0.10225846925972397,
      "grad_norm": 1.856428861618042,
      "learning_rate": 2.547169811320755e-06,
      "loss": 1.3347,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 163
    },
    {
      "epoch": 0.1028858218318695,
      "grad_norm": 2.000871419906616,
      "learning_rate": 2.5628930817610064e-06,
      "loss": 1.3671,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 164
    },
    {
      "epoch": 0.10351317440401506,
      "grad_norm": 3.1658947467803955,
      "learning_rate": 2.578616352201258e-06,
      "loss": 1.1915,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 165
    },
    {
      "epoch": 0.10414052697616061,
      "grad_norm": 1.8465994596481323,
      "learning_rate": 2.5943396226415095e-06,
      "loss": 1.5443,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 166
    },
    {
      "epoch": 0.10476787954830614,
      "grad_norm": 7.598590850830078,
      "learning_rate": 2.6100628930817613e-06,
      "loss": 1.3744,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 167
    },
    {
      "epoch": 0.1053952321204517,
      "grad_norm": 1.9127484560012817,
      "learning_rate": 2.6257861635220127e-06,
      "loss": 1.323,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 168
    },
    {
      "epoch": 0.10602258469259725,
      "grad_norm": 2.021780014038086,
      "learning_rate": 2.6415094339622644e-06,
      "loss": 1.3888,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 169
    },
    {
      "epoch": 0.10664993726474278,
      "grad_norm": 1.603609323501587,
      "learning_rate": 2.6572327044025158e-06,
      "loss": 1.3605,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 170
    },
    {
      "epoch": 0.10727728983688833,
      "grad_norm": 1.784544587135315,
      "learning_rate": 2.6729559748427675e-06,
      "loss": 1.3892,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 171
    },
    {
      "epoch": 0.10790464240903387,
      "grad_norm": 1.7793283462524414,
      "learning_rate": 2.688679245283019e-06,
      "loss": 1.4657,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 172
    },
    {
      "epoch": 0.10853199498117942,
      "grad_norm": 1.8067797422409058,
      "learning_rate": 2.7044025157232706e-06,
      "loss": 1.3574,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 173
    },
    {
      "epoch": 0.10915934755332497,
      "grad_norm": 1.8207420110702515,
      "learning_rate": 2.720125786163522e-06,
      "loss": 1.5282,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 174
    },
    {
      "epoch": 0.10978670012547051,
      "grad_norm": 1.8373732566833496,
      "learning_rate": 2.7358490566037738e-06,
      "loss": 1.3553,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 175
    },
    {
      "epoch": 0.11041405269761606,
      "grad_norm": 1.8807525634765625,
      "learning_rate": 2.7515723270440255e-06,
      "loss": 1.4968,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 176
    },
    {
      "epoch": 0.11104140526976161,
      "grad_norm": 1.7542294263839722,
      "learning_rate": 2.767295597484277e-06,
      "loss": 1.4336,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 177
    },
    {
      "epoch": 0.11166875784190715,
      "grad_norm": 1.7253546714782715,
      "learning_rate": 2.7830188679245286e-06,
      "loss": 1.3561,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 178
    },
    {
      "epoch": 0.1122961104140527,
      "grad_norm": 1.9645566940307617,
      "learning_rate": 2.79874213836478e-06,
      "loss": 1.3704,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 179
    },
    {
      "epoch": 0.11292346298619825,
      "grad_norm": 1.717736840248108,
      "learning_rate": 2.8144654088050318e-06,
      "loss": 1.2457,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 180
    },
    {
      "epoch": 0.11355081555834379,
      "grad_norm": 1.9592612981796265,
      "learning_rate": 2.830188679245283e-06,
      "loss": 1.6151,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 181
    },
    {
      "epoch": 0.11417816813048934,
      "grad_norm": 1.8280335664749146,
      "learning_rate": 2.845911949685535e-06,
      "loss": 1.2718,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 182
    },
    {
      "epoch": 0.11480552070263488,
      "grad_norm": 1.8632508516311646,
      "learning_rate": 2.861635220125786e-06,
      "loss": 1.5193,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 183
    },
    {
      "epoch": 0.11543287327478043,
      "grad_norm": 1.5125095844268799,
      "learning_rate": 2.877358490566038e-06,
      "loss": 1.3563,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 184
    },
    {
      "epoch": 0.11606022584692598,
      "grad_norm": 3.884819269180298,
      "learning_rate": 2.8930817610062893e-06,
      "loss": 1.4289,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 185
    },
    {
      "epoch": 0.11668757841907151,
      "grad_norm": 2.071526527404785,
      "learning_rate": 2.908805031446541e-06,
      "loss": 1.4305,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 186
    },
    {
      "epoch": 0.11731493099121706,
      "grad_norm": 1.6391427516937256,
      "learning_rate": 2.9245283018867924e-06,
      "loss": 1.434,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 187
    },
    {
      "epoch": 0.11794228356336262,
      "grad_norm": 1.7116525173187256,
      "learning_rate": 2.940251572327044e-06,
      "loss": 1.3891,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 188
    },
    {
      "epoch": 0.11856963613550815,
      "grad_norm": 2.0668036937713623,
      "learning_rate": 2.9559748427672955e-06,
      "loss": 1.4199,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 189
    },
    {
      "epoch": 0.1191969887076537,
      "grad_norm": 1.774177074432373,
      "learning_rate": 2.9716981132075473e-06,
      "loss": 1.3575,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 190
    },
    {
      "epoch": 0.11982434127979925,
      "grad_norm": 1.622582197189331,
      "learning_rate": 2.987421383647799e-06,
      "loss": 1.3625,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 191
    },
    {
      "epoch": 0.12045169385194479,
      "grad_norm": 1.9640077352523804,
      "learning_rate": 3.0031446540880504e-06,
      "loss": 1.3435,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 192
    },
    {
      "epoch": 0.12107904642409034,
      "grad_norm": 1.9893110990524292,
      "learning_rate": 3.018867924528302e-06,
      "loss": 1.3283,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 193
    },
    {
      "epoch": 0.12170639899623588,
      "grad_norm": 1.7824875116348267,
      "learning_rate": 3.0345911949685535e-06,
      "loss": 1.3816,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 194
    },
    {
      "epoch": 0.12233375156838143,
      "grad_norm": 1.4995529651641846,
      "learning_rate": 3.0503144654088053e-06,
      "loss": 1.2341,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 195
    },
    {
      "epoch": 0.12296110414052698,
      "grad_norm": 1.909799337387085,
      "learning_rate": 3.0660377358490567e-06,
      "loss": 1.396,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 196
    },
    {
      "epoch": 0.12358845671267252,
      "grad_norm": 1.9422545433044434,
      "learning_rate": 3.0817610062893084e-06,
      "loss": 1.4268,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 197
    },
    {
      "epoch": 0.12421580928481807,
      "grad_norm": 1.5427347421646118,
      "learning_rate": 3.0974842767295598e-06,
      "loss": 1.2787,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 198
    },
    {
      "epoch": 0.12484316185696362,
      "grad_norm": 2.090911865234375,
      "learning_rate": 3.1132075471698115e-06,
      "loss": 1.3448,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 199
    },
    {
      "epoch": 0.12547051442910917,
      "grad_norm": 1.820373296737671,
      "learning_rate": 3.128930817610063e-06,
      "loss": 1.4195,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 200
    },
    {
      "epoch": 0.1260978670012547,
      "grad_norm": 2.1247458457946777,
      "learning_rate": 3.1446540880503146e-06,
      "loss": 1.3468,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 201
    },
    {
      "epoch": 0.12672521957340024,
      "grad_norm": 2.0211782455444336,
      "learning_rate": 3.160377358490566e-06,
      "loss": 1.2897,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 202
    },
    {
      "epoch": 0.1273525721455458,
      "grad_norm": 2.2064476013183594,
      "learning_rate": 3.1761006289308178e-06,
      "loss": 1.3444,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 203
    },
    {
      "epoch": 0.12797992471769135,
      "grad_norm": 1.8434877395629883,
      "learning_rate": 3.191823899371069e-06,
      "loss": 1.3776,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 204
    },
    {
      "epoch": 0.1286072772898369,
      "grad_norm": 1.9339540004730225,
      "learning_rate": 3.207547169811321e-06,
      "loss": 1.2157,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 205
    },
    {
      "epoch": 0.12923462986198245,
      "grad_norm": 1.8434295654296875,
      "learning_rate": 3.2232704402515726e-06,
      "loss": 1.3971,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 206
    },
    {
      "epoch": 0.12986198243412797,
      "grad_norm": 1.5623953342437744,
      "learning_rate": 3.238993710691824e-06,
      "loss": 1.3002,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 207
    },
    {
      "epoch": 0.13048933500627352,
      "grad_norm": 1.600899577140808,
      "learning_rate": 3.2547169811320758e-06,
      "loss": 1.4586,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 208
    },
    {
      "epoch": 0.13111668757841907,
      "grad_norm": 2.4249284267425537,
      "learning_rate": 3.270440251572327e-06,
      "loss": 1.3632,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 209
    },
    {
      "epoch": 0.13174404015056462,
      "grad_norm": 1.8537931442260742,
      "learning_rate": 3.286163522012579e-06,
      "loss": 1.3831,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 210
    },
    {
      "epoch": 0.13237139272271017,
      "grad_norm": 1.504931092262268,
      "learning_rate": 3.30188679245283e-06,
      "loss": 1.2481,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 211
    },
    {
      "epoch": 0.1329987452948557,
      "grad_norm": 1.8068492412567139,
      "learning_rate": 3.317610062893082e-06,
      "loss": 1.3971,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 212
    },
    {
      "epoch": 0.13362609786700125,
      "grad_norm": 1.5803545713424683,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.3475,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 213
    },
    {
      "epoch": 0.1342534504391468,
      "grad_norm": 1.7603237628936768,
      "learning_rate": 3.349056603773585e-06,
      "loss": 1.3331,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 214
    },
    {
      "epoch": 0.13488080301129235,
      "grad_norm": 1.8488463163375854,
      "learning_rate": 3.3647798742138364e-06,
      "loss": 1.3273,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 215
    },
    {
      "epoch": 0.1355081555834379,
      "grad_norm": 1.9468002319335938,
      "learning_rate": 3.380503144654088e-06,
      "loss": 1.3886,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 216
    },
    {
      "epoch": 0.13613550815558345,
      "grad_norm": 2.5950703620910645,
      "learning_rate": 3.3962264150943395e-06,
      "loss": 1.316,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 217
    },
    {
      "epoch": 0.13676286072772897,
      "grad_norm": 1.8121410608291626,
      "learning_rate": 3.4119496855345913e-06,
      "loss": 1.3347,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 218
    },
    {
      "epoch": 0.13739021329987453,
      "grad_norm": 1.9044947624206543,
      "learning_rate": 3.4276729559748427e-06,
      "loss": 1.4193,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 219
    },
    {
      "epoch": 0.13801756587202008,
      "grad_norm": 1.817305564880371,
      "learning_rate": 3.4433962264150944e-06,
      "loss": 1.2194,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 220
    },
    {
      "epoch": 0.13864491844416563,
      "grad_norm": 1.3980709314346313,
      "learning_rate": 3.4591194968553458e-06,
      "loss": 1.256,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 221
    },
    {
      "epoch": 0.13927227101631118,
      "grad_norm": 1.7033888101577759,
      "learning_rate": 3.4748427672955975e-06,
      "loss": 1.3148,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 222
    },
    {
      "epoch": 0.1398996235884567,
      "grad_norm": 1.9811346530914307,
      "learning_rate": 3.4905660377358493e-06,
      "loss": 1.2855,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 223
    },
    {
      "epoch": 0.14052697616060225,
      "grad_norm": 1.5731050968170166,
      "learning_rate": 3.5062893081761007e-06,
      "loss": 1.3025,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 224
    },
    {
      "epoch": 0.1411543287327478,
      "grad_norm": 1.8436543941497803,
      "learning_rate": 3.5220125786163524e-06,
      "loss": 1.3267,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 225
    },
    {
      "epoch": 0.14178168130489335,
      "grad_norm": 1.9756402969360352,
      "learning_rate": 3.5377358490566038e-06,
      "loss": 1.403,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 226
    },
    {
      "epoch": 0.1424090338770389,
      "grad_norm": 1.7208794355392456,
      "learning_rate": 3.5534591194968555e-06,
      "loss": 1.2933,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 227
    },
    {
      "epoch": 0.14303638644918445,
      "grad_norm": 1.8221158981323242,
      "learning_rate": 3.569182389937107e-06,
      "loss": 1.3209,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 228
    },
    {
      "epoch": 0.14366373902132998,
      "grad_norm": 1.9259345531463623,
      "learning_rate": 3.5849056603773586e-06,
      "loss": 1.4681,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 229
    },
    {
      "epoch": 0.14429109159347553,
      "grad_norm": 1.8323369026184082,
      "learning_rate": 3.60062893081761e-06,
      "loss": 1.3018,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 230
    },
    {
      "epoch": 0.14491844416562108,
      "grad_norm": 1.7952187061309814,
      "learning_rate": 3.6163522012578618e-06,
      "loss": 1.411,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 231
    },
    {
      "epoch": 0.14554579673776663,
      "grad_norm": 1.8569618463516235,
      "learning_rate": 3.632075471698113e-06,
      "loss": 1.3432,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 232
    },
    {
      "epoch": 0.14617314930991218,
      "grad_norm": 1.9454624652862549,
      "learning_rate": 3.647798742138365e-06,
      "loss": 1.3542,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 233
    },
    {
      "epoch": 0.1468005018820577,
      "grad_norm": 1.7847152948379517,
      "learning_rate": 3.6635220125786162e-06,
      "loss": 1.4037,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 234
    },
    {
      "epoch": 0.14742785445420326,
      "grad_norm": 1.9342718124389648,
      "learning_rate": 3.679245283018868e-06,
      "loss": 1.4753,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 235
    },
    {
      "epoch": 0.1480552070263488,
      "grad_norm": 1.8230245113372803,
      "learning_rate": 3.6949685534591193e-06,
      "loss": 1.3496,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 236
    },
    {
      "epoch": 0.14868255959849436,
      "grad_norm": 2.1563334465026855,
      "learning_rate": 3.710691823899371e-06,
      "loss": 1.3971,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 237
    },
    {
      "epoch": 0.1493099121706399,
      "grad_norm": 1.9064549207687378,
      "learning_rate": 3.726415094339623e-06,
      "loss": 1.3381,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 238
    },
    {
      "epoch": 0.14993726474278546,
      "grad_norm": 1.7572532892227173,
      "learning_rate": 3.742138364779874e-06,
      "loss": 1.3343,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 239
    },
    {
      "epoch": 0.15056461731493098,
      "grad_norm": 1.8741847276687622,
      "learning_rate": 3.7578616352201264e-06,
      "loss": 1.2935,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 240
    },
    {
      "epoch": 0.15119196988707653,
      "grad_norm": 21.64093589782715,
      "learning_rate": 3.7735849056603777e-06,
      "loss": 1.3308,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 241
    },
    {
      "epoch": 0.15181932245922208,
      "grad_norm": 2.0540499687194824,
      "learning_rate": 3.7893081761006295e-06,
      "loss": 1.3383,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 242
    },
    {
      "epoch": 0.15244667503136763,
      "grad_norm": 2.0343210697174072,
      "learning_rate": 3.8050314465408813e-06,
      "loss": 1.3605,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 243
    },
    {
      "epoch": 0.15307402760351319,
      "grad_norm": 1.7628991603851318,
      "learning_rate": 3.820754716981133e-06,
      "loss": 1.2637,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 244
    },
    {
      "epoch": 0.1537013801756587,
      "grad_norm": 1.801661491394043,
      "learning_rate": 3.836477987421384e-06,
      "loss": 1.4673,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 245
    },
    {
      "epoch": 0.15432873274780426,
      "grad_norm": 1.7599327564239502,
      "learning_rate": 3.852201257861636e-06,
      "loss": 1.3397,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 246
    },
    {
      "epoch": 0.1549560853199498,
      "grad_norm": 1.9988664388656616,
      "learning_rate": 3.8679245283018875e-06,
      "loss": 1.2941,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 247
    },
    {
      "epoch": 0.15558343789209536,
      "grad_norm": 1.2556072473526,
      "learning_rate": 3.883647798742139e-06,
      "loss": 1.2417,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 248
    },
    {
      "epoch": 0.1562107904642409,
      "grad_norm": 2.0255300998687744,
      "learning_rate": 3.89937106918239e-06,
      "loss": 1.455,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 249
    },
    {
      "epoch": 0.15683814303638646,
      "grad_norm": 1.8777093887329102,
      "learning_rate": 3.915094339622642e-06,
      "loss": 1.244,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 250
    },
    {
      "epoch": 0.15746549560853199,
      "grad_norm": 2.1528234481811523,
      "learning_rate": 3.930817610062894e-06,
      "loss": 1.3446,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 251
    },
    {
      "epoch": 0.15809284818067754,
      "grad_norm": 1.7366938591003418,
      "learning_rate": 3.9465408805031455e-06,
      "loss": 1.2959,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 252
    },
    {
      "epoch": 0.1587202007528231,
      "grad_norm": 1.7935433387756348,
      "learning_rate": 3.962264150943396e-06,
      "loss": 1.3411,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 253
    },
    {
      "epoch": 0.15934755332496864,
      "grad_norm": 1.8638379573822021,
      "learning_rate": 3.977987421383648e-06,
      "loss": 1.3481,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 254
    },
    {
      "epoch": 0.1599749058971142,
      "grad_norm": 1.9182302951812744,
      "learning_rate": 3.9937106918239e-06,
      "loss": 1.3058,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 255
    },
    {
      "epoch": 0.1606022584692597,
      "grad_norm": 1.832619309425354,
      "learning_rate": 4.009433962264152e-06,
      "loss": 1.352,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 256
    },
    {
      "epoch": 0.16122961104140526,
      "grad_norm": 1.579271674156189,
      "learning_rate": 4.025157232704403e-06,
      "loss": 1.2741,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 257
    },
    {
      "epoch": 0.1618569636135508,
      "grad_norm": 1.567428708076477,
      "learning_rate": 4.040880503144654e-06,
      "loss": 1.2891,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 258
    },
    {
      "epoch": 0.16248431618569636,
      "grad_norm": 1.7496509552001953,
      "learning_rate": 4.056603773584906e-06,
      "loss": 1.2356,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 259
    },
    {
      "epoch": 0.16311166875784192,
      "grad_norm": 1.7312959432601929,
      "learning_rate": 4.072327044025158e-06,
      "loss": 1.3896,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 260
    },
    {
      "epoch": 0.16373902132998747,
      "grad_norm": 2.2512001991271973,
      "learning_rate": 4.08805031446541e-06,
      "loss": 1.353,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 261
    },
    {
      "epoch": 0.164366373902133,
      "grad_norm": 1.853335976600647,
      "learning_rate": 4.103773584905661e-06,
      "loss": 1.5254,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 262
    },
    {
      "epoch": 0.16499372647427854,
      "grad_norm": 1.9108259677886963,
      "learning_rate": 4.119496855345912e-06,
      "loss": 1.4425,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 263
    },
    {
      "epoch": 0.1656210790464241,
      "grad_norm": 1.7109558582305908,
      "learning_rate": 4.135220125786164e-06,
      "loss": 1.1786,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 264
    },
    {
      "epoch": 0.16624843161856964,
      "grad_norm": 1.9971568584442139,
      "learning_rate": 4.150943396226416e-06,
      "loss": 1.4076,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 265
    },
    {
      "epoch": 0.1668757841907152,
      "grad_norm": 1.3846633434295654,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.4011,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 266
    },
    {
      "epoch": 0.16750313676286072,
      "grad_norm": 1.7542074918746948,
      "learning_rate": 4.182389937106919e-06,
      "loss": 1.2915,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 267
    },
    {
      "epoch": 0.16813048933500627,
      "grad_norm": 1.7762579917907715,
      "learning_rate": 4.19811320754717e-06,
      "loss": 1.3584,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 268
    },
    {
      "epoch": 0.16875784190715182,
      "grad_norm": 1.8659406900405884,
      "learning_rate": 4.213836477987422e-06,
      "loss": 1.3343,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 269
    },
    {
      "epoch": 0.16938519447929737,
      "grad_norm": 1.7126916646957397,
      "learning_rate": 4.229559748427673e-06,
      "loss": 1.4461,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 270
    },
    {
      "epoch": 0.17001254705144292,
      "grad_norm": 2.4897971153259277,
      "learning_rate": 4.245283018867925e-06,
      "loss": 1.3599,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 271
    },
    {
      "epoch": 0.17063989962358847,
      "grad_norm": 1.7607709169387817,
      "learning_rate": 4.261006289308177e-06,
      "loss": 1.2669,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 272
    },
    {
      "epoch": 0.171267252195734,
      "grad_norm": 1.936936378479004,
      "learning_rate": 4.276729559748428e-06,
      "loss": 1.4449,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 273
    },
    {
      "epoch": 0.17189460476787954,
      "grad_norm": 1.7051970958709717,
      "learning_rate": 4.29245283018868e-06,
      "loss": 1.2865,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 274
    },
    {
      "epoch": 0.1725219573400251,
      "grad_norm": 1.7027671337127686,
      "learning_rate": 4.308176100628931e-06,
      "loss": 1.2152,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 275
    },
    {
      "epoch": 0.17314930991217065,
      "grad_norm": 1.8415780067443848,
      "learning_rate": 4.323899371069183e-06,
      "loss": 1.3654,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 276
    },
    {
      "epoch": 0.1737766624843162,
      "grad_norm": 2.0088579654693604,
      "learning_rate": 4.339622641509435e-06,
      "loss": 1.416,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 277
    },
    {
      "epoch": 0.17440401505646172,
      "grad_norm": 1.8947542905807495,
      "learning_rate": 4.355345911949686e-06,
      "loss": 1.2923,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 278
    },
    {
      "epoch": 0.17503136762860727,
      "grad_norm": 1.9341015815734863,
      "learning_rate": 4.371069182389937e-06,
      "loss": 1.3498,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 279
    },
    {
      "epoch": 0.17565872020075282,
      "grad_norm": 1.9172481298446655,
      "learning_rate": 4.386792452830189e-06,
      "loss": 1.4128,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 280
    },
    {
      "epoch": 0.17628607277289837,
      "grad_norm": 1.8716827630996704,
      "learning_rate": 4.402515723270441e-06,
      "loss": 1.3574,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 281
    },
    {
      "epoch": 0.17691342534504392,
      "grad_norm": 1.814761996269226,
      "learning_rate": 4.418238993710693e-06,
      "loss": 1.3448,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 282
    },
    {
      "epoch": 0.17754077791718947,
      "grad_norm": 2.0237176418304443,
      "learning_rate": 4.4339622641509435e-06,
      "loss": 1.2693,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 283
    },
    {
      "epoch": 0.178168130489335,
      "grad_norm": 1.7121891975402832,
      "learning_rate": 4.449685534591195e-06,
      "loss": 1.2457,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 284
    },
    {
      "epoch": 0.17879548306148055,
      "grad_norm": 1.781483769416809,
      "learning_rate": 4.465408805031447e-06,
      "loss": 1.2307,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 285
    },
    {
      "epoch": 0.1794228356336261,
      "grad_norm": 1.9639188051223755,
      "learning_rate": 4.481132075471699e-06,
      "loss": 1.2809,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 286
    },
    {
      "epoch": 0.18005018820577165,
      "grad_norm": 2.1548235416412354,
      "learning_rate": 4.49685534591195e-06,
      "loss": 1.2848,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 287
    },
    {
      "epoch": 0.1806775407779172,
      "grad_norm": 1.7896254062652588,
      "learning_rate": 4.5125786163522015e-06,
      "loss": 1.3772,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 288
    },
    {
      "epoch": 0.18130489335006272,
      "grad_norm": 1.9206440448760986,
      "learning_rate": 4.528301886792453e-06,
      "loss": 1.3501,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 289
    },
    {
      "epoch": 0.18193224592220827,
      "grad_norm": 1.570934534072876,
      "learning_rate": 4.544025157232705e-06,
      "loss": 1.2086,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 290
    },
    {
      "epoch": 0.18255959849435383,
      "grad_norm": 1.782419204711914,
      "learning_rate": 4.559748427672957e-06,
      "loss": 1.3928,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 291
    },
    {
      "epoch": 0.18318695106649938,
      "grad_norm": 2.629718780517578,
      "learning_rate": 4.575471698113208e-06,
      "loss": 1.3189,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 292
    },
    {
      "epoch": 0.18381430363864493,
      "grad_norm": 1.9206874370574951,
      "learning_rate": 4.5911949685534595e-06,
      "loss": 1.3112,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 293
    },
    {
      "epoch": 0.18444165621079048,
      "grad_norm": 1.8122892379760742,
      "learning_rate": 4.606918238993711e-06,
      "loss": 1.2539,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 294
    },
    {
      "epoch": 0.185069008782936,
      "grad_norm": 1.8378740549087524,
      "learning_rate": 4.622641509433963e-06,
      "loss": 1.307,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 295
    },
    {
      "epoch": 0.18569636135508155,
      "grad_norm": 1.7635501623153687,
      "learning_rate": 4.638364779874214e-06,
      "loss": 1.3125,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 296
    },
    {
      "epoch": 0.1863237139272271,
      "grad_norm": 1.7572214603424072,
      "learning_rate": 4.654088050314466e-06,
      "loss": 1.4464,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 297
    },
    {
      "epoch": 0.18695106649937265,
      "grad_norm": 1.8707910776138306,
      "learning_rate": 4.6698113207547175e-06,
      "loss": 1.4683,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 298
    },
    {
      "epoch": 0.1875784190715182,
      "grad_norm": 1.9102506637573242,
      "learning_rate": 4.685534591194969e-06,
      "loss": 1.3676,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 299
    },
    {
      "epoch": 0.18820577164366373,
      "grad_norm": 1.95762038230896,
      "learning_rate": 4.70125786163522e-06,
      "loss": 1.4814,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 300
    },
    {
      "epoch": 0.18883312421580928,
      "grad_norm": 1.8276519775390625,
      "learning_rate": 4.716981132075472e-06,
      "loss": 1.3073,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 301
    },
    {
      "epoch": 0.18946047678795483,
      "grad_norm": 1.9433486461639404,
      "learning_rate": 4.732704402515724e-06,
      "loss": 1.2964,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 302
    },
    {
      "epoch": 0.19008782936010038,
      "grad_norm": 1.7516218423843384,
      "learning_rate": 4.7484276729559755e-06,
      "loss": 1.3246,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 303
    },
    {
      "epoch": 0.19071518193224593,
      "grad_norm": 1.6977726221084595,
      "learning_rate": 4.764150943396227e-06,
      "loss": 1.2985,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 304
    },
    {
      "epoch": 0.19134253450439148,
      "grad_norm": 1.5477553606033325,
      "learning_rate": 4.779874213836478e-06,
      "loss": 1.2887,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 305
    },
    {
      "epoch": 0.191969887076537,
      "grad_norm": 2.1542649269104004,
      "learning_rate": 4.79559748427673e-06,
      "loss": 1.3762,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 306
    },
    {
      "epoch": 0.19259723964868256,
      "grad_norm": 1.8527969121932983,
      "learning_rate": 4.811320754716982e-06,
      "loss": 1.2428,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 307
    },
    {
      "epoch": 0.1932245922208281,
      "grad_norm": 2.14410400390625,
      "learning_rate": 4.8270440251572335e-06,
      "loss": 1.2875,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 308
    },
    {
      "epoch": 0.19385194479297366,
      "grad_norm": 1.9714504480361938,
      "learning_rate": 4.842767295597484e-06,
      "loss": 1.6028,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 309
    },
    {
      "epoch": 0.1944792973651192,
      "grad_norm": 2.064124345779419,
      "learning_rate": 4.858490566037736e-06,
      "loss": 1.3189,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 310
    },
    {
      "epoch": 0.19510664993726473,
      "grad_norm": 1.954937219619751,
      "learning_rate": 4.874213836477988e-06,
      "loss": 1.3709,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 311
    },
    {
      "epoch": 0.19573400250941028,
      "grad_norm": 1.5165770053863525,
      "learning_rate": 4.88993710691824e-06,
      "loss": 1.4661,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 312
    },
    {
      "epoch": 0.19636135508155583,
      "grad_norm": 1.927662968635559,
      "learning_rate": 4.905660377358491e-06,
      "loss": 1.2719,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 313
    },
    {
      "epoch": 0.19698870765370138,
      "grad_norm": 1.7934112548828125,
      "learning_rate": 4.921383647798742e-06,
      "loss": 1.5374,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 314
    },
    {
      "epoch": 0.19761606022584693,
      "grad_norm": 2.1834499835968018,
      "learning_rate": 4.937106918238994e-06,
      "loss": 1.3703,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 315
    },
    {
      "epoch": 0.19824341279799249,
      "grad_norm": 1.7504669427871704,
      "learning_rate": 4.952830188679246e-06,
      "loss": 1.3915,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 316
    },
    {
      "epoch": 0.198870765370138,
      "grad_norm": 1.904638409614563,
      "learning_rate": 4.968553459119497e-06,
      "loss": 1.2474,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 317
    },
    {
      "epoch": 0.19949811794228356,
      "grad_norm": 1.925824761390686,
      "learning_rate": 4.984276729559749e-06,
      "loss": 1.3469,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 318
    },
    {
      "epoch": 0.2001254705144291,
      "grad_norm": 2.1691761016845703,
      "learning_rate": 5e-06,
      "loss": 1.2512,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 319
    },
    {
      "epoch": 0.20075282308657466,
      "grad_norm": 1.7544150352478027,
      "learning_rate": 4.999998502227263e-06,
      "loss": 1.3736,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 320
    },
    {
      "epoch": 0.2013801756587202,
      "grad_norm": 1.6073588132858276,
      "learning_rate": 4.9999940089108415e-06,
      "loss": 1.4896,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 321
    },
    {
      "epoch": 0.20200752823086573,
      "grad_norm": 1.6996064186096191,
      "learning_rate": 4.999986520056122e-06,
      "loss": 1.2907,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 322
    },
    {
      "epoch": 0.20263488080301129,
      "grad_norm": 1.7334671020507812,
      "learning_rate": 4.999976035672078e-06,
      "loss": 1.3704,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 323
    },
    {
      "epoch": 0.20326223337515684,
      "grad_norm": 1.9705431461334229,
      "learning_rate": 4.999962555771272e-06,
      "loss": 1.2347,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 324
    },
    {
      "epoch": 0.2038895859473024,
      "grad_norm": 1.3459328413009644,
      "learning_rate": 4.999946080369854e-06,
      "loss": 1.3108,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 325
    },
    {
      "epoch": 0.20451693851944794,
      "grad_norm": 1.8672802448272705,
      "learning_rate": 4.999926609487568e-06,
      "loss": 1.3796,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 326
    },
    {
      "epoch": 0.2051442910915935,
      "grad_norm": 2.1950390338897705,
      "learning_rate": 4.9999041431477425e-06,
      "loss": 1.2373,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 327
    },
    {
      "epoch": 0.205771643663739,
      "grad_norm": 1.8165322542190552,
      "learning_rate": 4.999878681377297e-06,
      "loss": 1.2377,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 328
    },
    {
      "epoch": 0.20639899623588456,
      "grad_norm": 1.840531587600708,
      "learning_rate": 4.999850224206741e-06,
      "loss": 1.3705,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 329
    },
    {
      "epoch": 0.20702634880803011,
      "grad_norm": 1.6710518598556519,
      "learning_rate": 4.999818771670172e-06,
      "loss": 1.3876,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 330
    },
    {
      "epoch": 0.20765370138017566,
      "grad_norm": 2.126668930053711,
      "learning_rate": 4.9997843238052776e-06,
      "loss": 1.3916,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 331
    },
    {
      "epoch": 0.20828105395232122,
      "grad_norm": 1.5642105340957642,
      "learning_rate": 4.999746880653333e-06,
      "loss": 1.2617,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 332
    },
    {
      "epoch": 0.20890840652446674,
      "grad_norm": 1.7182477712631226,
      "learning_rate": 4.999706442259205e-06,
      "loss": 1.3659,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 333
    },
    {
      "epoch": 0.2095357590966123,
      "grad_norm": 1.906502604484558,
      "learning_rate": 4.999663008671344e-06,
      "loss": 1.3145,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 334
    },
    {
      "epoch": 0.21016311166875784,
      "grad_norm": 2.0408036708831787,
      "learning_rate": 4.999616579941797e-06,
      "loss": 1.3193,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 335
    },
    {
      "epoch": 0.2107904642409034,
      "grad_norm": 1.7874103784561157,
      "learning_rate": 4.999567156126193e-06,
      "loss": 1.308,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 336
    },
    {
      "epoch": 0.21141781681304894,
      "grad_norm": 1.9260340929031372,
      "learning_rate": 4.999514737283754e-06,
      "loss": 1.3081,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 337
    },
    {
      "epoch": 0.2120451693851945,
      "grad_norm": 1.7831474542617798,
      "learning_rate": 4.999459323477288e-06,
      "loss": 1.2228,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 338
    },
    {
      "epoch": 0.21267252195734002,
      "grad_norm": 1.9563218355178833,
      "learning_rate": 4.999400914773193e-06,
      "loss": 1.3612,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 339
    },
    {
      "epoch": 0.21329987452948557,
      "grad_norm": 3.149996280670166,
      "learning_rate": 4.999339511241458e-06,
      "loss": 1.2988,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 340
    },
    {
      "epoch": 0.21392722710163112,
      "grad_norm": 1.7452430725097656,
      "learning_rate": 4.999275112955654e-06,
      "loss": 1.3759,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 341
    },
    {
      "epoch": 0.21455457967377667,
      "grad_norm": 1.843584656715393,
      "learning_rate": 4.999207719992947e-06,
      "loss": 1.397,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 342
    },
    {
      "epoch": 0.21518193224592222,
      "grad_norm": 1.9594056606292725,
      "learning_rate": 4.999137332434086e-06,
      "loss": 1.2943,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 343
    },
    {
      "epoch": 0.21580928481806774,
      "grad_norm": 1.8866289854049683,
      "learning_rate": 4.999063950363413e-06,
      "loss": 1.4503,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 344
    },
    {
      "epoch": 0.2164366373902133,
      "grad_norm": 1.8794231414794922,
      "learning_rate": 4.9989875738688555e-06,
      "loss": 1.2822,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 345
    },
    {
      "epoch": 0.21706398996235884,
      "grad_norm": 2.0103776454925537,
      "learning_rate": 4.998908203041926e-06,
      "loss": 1.3155,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 346
    },
    {
      "epoch": 0.2176913425345044,
      "grad_norm": 1.9219310283660889,
      "learning_rate": 4.9988258379777334e-06,
      "loss": 1.3347,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 347
    },
    {
      "epoch": 0.21831869510664995,
      "grad_norm": 1.6140512228012085,
      "learning_rate": 4.998740478774965e-06,
      "loss": 1.2699,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 348
    },
    {
      "epoch": 0.2189460476787955,
      "grad_norm": 1.647599220275879,
      "learning_rate": 4.998652125535901e-06,
      "loss": 1.3667,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 349
    },
    {
      "epoch": 0.21957340025094102,
      "grad_norm": 1.810556173324585,
      "learning_rate": 4.998560778366407e-06,
      "loss": 1.4081,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 350
    },
    {
      "epoch": 0.22020075282308657,
      "grad_norm": 1.6720383167266846,
      "learning_rate": 4.998466437375938e-06,
      "loss": 1.309,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 351
    },
    {
      "epoch": 0.22082810539523212,
      "grad_norm": 1.6353473663330078,
      "learning_rate": 4.998369102677535e-06,
      "loss": 1.2393,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 352
    },
    {
      "epoch": 0.22145545796737767,
      "grad_norm": 1.7175465822219849,
      "learning_rate": 4.9982687743878256e-06,
      "loss": 1.318,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 353
    },
    {
      "epoch": 0.22208281053952322,
      "grad_norm": 1.627525806427002,
      "learning_rate": 4.998165452627025e-06,
      "loss": 1.2159,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 354
    },
    {
      "epoch": 0.22271016311166875,
      "grad_norm": 1.8175550699234009,
      "learning_rate": 4.998059137518936e-06,
      "loss": 1.2853,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 355
    },
    {
      "epoch": 0.2233375156838143,
      "grad_norm": 2.5346550941467285,
      "learning_rate": 4.997949829190947e-06,
      "loss": 1.3752,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 356
    },
    {
      "epoch": 0.22396486825595985,
      "grad_norm": 1.8762712478637695,
      "learning_rate": 4.997837527774033e-06,
      "loss": 1.3113,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 357
    },
    {
      "epoch": 0.2245922208281054,
      "grad_norm": 1.645166277885437,
      "learning_rate": 4.997722233402755e-06,
      "loss": 1.3383,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 358
    },
    {
      "epoch": 0.22521957340025095,
      "grad_norm": 1.9214521646499634,
      "learning_rate": 4.997603946215262e-06,
      "loss": 1.1769,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 359
    },
    {
      "epoch": 0.2258469259723965,
      "grad_norm": 1.7410590648651123,
      "learning_rate": 4.997482666353287e-06,
      "loss": 1.2254,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 360
    },
    {
      "epoch": 0.22647427854454202,
      "grad_norm": 2.080554723739624,
      "learning_rate": 4.99735839396215e-06,
      "loss": 1.2844,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 361
    },
    {
      "epoch": 0.22710163111668757,
      "grad_norm": 1.9182201623916626,
      "learning_rate": 4.997231129190757e-06,
      "loss": 1.3778,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 362
    },
    {
      "epoch": 0.22772898368883313,
      "grad_norm": 2.0027408599853516,
      "learning_rate": 4.997100872191598e-06,
      "loss": 1.3227,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 363
    },
    {
      "epoch": 0.22835633626097868,
      "grad_norm": 1.7964129447937012,
      "learning_rate": 4.99696762312075e-06,
      "loss": 1.3391,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 364
    },
    {
      "epoch": 0.22898368883312423,
      "grad_norm": 1.8464184999465942,
      "learning_rate": 4.996831382137873e-06,
      "loss": 1.4158,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 365
    },
    {
      "epoch": 0.22961104140526975,
      "grad_norm": 1.7289762496948242,
      "learning_rate": 4.996692149406216e-06,
      "loss": 1.3439,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 366
    },
    {
      "epoch": 0.2302383939774153,
      "grad_norm": 1.4288086891174316,
      "learning_rate": 4.996549925092609e-06,
      "loss": 1.2174,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 367
    },
    {
      "epoch": 0.23086574654956085,
      "grad_norm": 1.8290668725967407,
      "learning_rate": 4.996404709367466e-06,
      "loss": 1.3256,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 368
    },
    {
      "epoch": 0.2314930991217064,
      "grad_norm": 1.6571376323699951,
      "learning_rate": 4.99625650240479e-06,
      "loss": 1.2822,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 369
    },
    {
      "epoch": 0.23212045169385195,
      "grad_norm": 1.8437656164169312,
      "learning_rate": 4.9961053043821636e-06,
      "loss": 1.3772,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 370
    },
    {
      "epoch": 0.2327478042659975,
      "grad_norm": 1.6965371370315552,
      "learning_rate": 4.9959511154807555e-06,
      "loss": 1.2846,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 371
    },
    {
      "epoch": 0.23337515683814303,
      "grad_norm": 1.8487476110458374,
      "learning_rate": 4.995793935885317e-06,
      "loss": 1.3123,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 372
    },
    {
      "epoch": 0.23400250941028858,
      "grad_norm": 1.8760329484939575,
      "learning_rate": 4.995633765784185e-06,
      "loss": 1.2402,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 373
    },
    {
      "epoch": 0.23462986198243413,
      "grad_norm": 1.8349744081497192,
      "learning_rate": 4.9954706053692766e-06,
      "loss": 1.2657,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 374
    },
    {
      "epoch": 0.23525721455457968,
      "grad_norm": 1.6800428628921509,
      "learning_rate": 4.995304454836095e-06,
      "loss": 1.2008,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 375
    },
    {
      "epoch": 0.23588456712672523,
      "grad_norm": 1.647479772567749,
      "learning_rate": 4.9951353143837235e-06,
      "loss": 1.2184,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 376
    },
    {
      "epoch": 0.23651191969887075,
      "grad_norm": 1.886382818222046,
      "learning_rate": 4.99496318421483e-06,
      "loss": 1.3817,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 377
    },
    {
      "epoch": 0.2371392722710163,
      "grad_norm": 1.8178256750106812,
      "learning_rate": 4.9947880645356636e-06,
      "loss": 1.4071,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 378
    },
    {
      "epoch": 0.23776662484316186,
      "grad_norm": 1.8750951290130615,
      "learning_rate": 4.994609955556057e-06,
      "loss": 1.2923,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 379
    },
    {
      "epoch": 0.2383939774153074,
      "grad_norm": 1.7878093719482422,
      "learning_rate": 4.994428857489422e-06,
      "loss": 1.3081,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 380
    },
    {
      "epoch": 0.23902132998745296,
      "grad_norm": 2.013211488723755,
      "learning_rate": 4.994244770552755e-06,
      "loss": 1.4391,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 381
    },
    {
      "epoch": 0.2396486825595985,
      "grad_norm": 1.7840601205825806,
      "learning_rate": 4.994057694966632e-06,
      "loss": 1.283,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 382
    },
    {
      "epoch": 0.24027603513174403,
      "grad_norm": 1.9290210008621216,
      "learning_rate": 4.993867630955209e-06,
      "loss": 1.2982,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 383
    },
    {
      "epoch": 0.24090338770388958,
      "grad_norm": 1.8992472887039185,
      "learning_rate": 4.993674578746225e-06,
      "loss": 1.2167,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 384
    },
    {
      "epoch": 0.24153074027603513,
      "grad_norm": 2.0222549438476562,
      "learning_rate": 4.993478538571e-06,
      "loss": 1.3329,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 385
    },
    {
      "epoch": 0.24215809284818068,
      "grad_norm": 1.8096097707748413,
      "learning_rate": 4.99327951066443e-06,
      "loss": 1.4154,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 386
    },
    {
      "epoch": 0.24278544542032623,
      "grad_norm": 1.4684665203094482,
      "learning_rate": 4.9930774952649975e-06,
      "loss": 1.3255,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 387
    },
    {
      "epoch": 0.24341279799247176,
      "grad_norm": 1.7699239253997803,
      "learning_rate": 4.9928724926147586e-06,
      "loss": 1.3138,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 388
    },
    {
      "epoch": 0.2440401505646173,
      "grad_norm": 1.8916990756988525,
      "learning_rate": 4.992664502959351e-06,
      "loss": 1.5344,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 389
    },
    {
      "epoch": 0.24466750313676286,
      "grad_norm": 2.013256311416626,
      "learning_rate": 4.992453526547993e-06,
      "loss": 1.4169,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 390
    },
    {
      "epoch": 0.2452948557089084,
      "grad_norm": 1.911051630973816,
      "learning_rate": 4.9922395636334806e-06,
      "loss": 1.3485,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 391
    },
    {
      "epoch": 0.24592220828105396,
      "grad_norm": 2.74343204498291,
      "learning_rate": 4.992022614472187e-06,
      "loss": 1.25,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 392
    },
    {
      "epoch": 0.2465495608531995,
      "grad_norm": 1.8582487106323242,
      "learning_rate": 4.991802679324064e-06,
      "loss": 1.3874,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 393
    },
    {
      "epoch": 0.24717691342534504,
      "grad_norm": 1.8683292865753174,
      "learning_rate": 4.991579758452644e-06,
      "loss": 1.3666,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 394
    },
    {
      "epoch": 0.24780426599749059,
      "grad_norm": 1.898242712020874,
      "learning_rate": 4.991353852125032e-06,
      "loss": 1.2622,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 395
    },
    {
      "epoch": 0.24843161856963614,
      "grad_norm": 1.793607234954834,
      "learning_rate": 4.991124960611916e-06,
      "loss": 1.3744,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 396
    },
    {
      "epoch": 0.2490589711417817,
      "grad_norm": 1.8810088634490967,
      "learning_rate": 4.990893084187557e-06,
      "loss": 1.3638,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 397
    },
    {
      "epoch": 0.24968632371392724,
      "grad_norm": 1.5801982879638672,
      "learning_rate": 4.990658223129793e-06,
      "loss": 1.3114,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 398
    },
    {
      "epoch": 0.25031367628607276,
      "grad_norm": 1.5382050275802612,
      "learning_rate": 4.9904203777200375e-06,
      "loss": 1.2568,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 399
    },
    {
      "epoch": 0.25094102885821834,
      "grad_norm": 1.5019408464431763,
      "learning_rate": 4.990179548243285e-06,
      "loss": 1.3077,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 400
    },
    {
      "epoch": 0.25156838143036386,
      "grad_norm": 1.5238393545150757,
      "learning_rate": 4.989935734988098e-06,
      "loss": 1.255,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 401
    },
    {
      "epoch": 0.2521957340025094,
      "grad_norm": 1.8959760665893555,
      "learning_rate": 4.98968893824662e-06,
      "loss": 1.2656,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 402
    },
    {
      "epoch": 0.25282308657465496,
      "grad_norm": 1.4434200525283813,
      "learning_rate": 4.989439158314566e-06,
      "loss": 1.1504,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 403
    },
    {
      "epoch": 0.2534504391468005,
      "grad_norm": 1.9673125743865967,
      "learning_rate": 4.989186395491229e-06,
      "loss": 1.4144,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 404
    },
    {
      "epoch": 0.25407779171894607,
      "grad_norm": 1.916951060295105,
      "learning_rate": 4.988930650079472e-06,
      "loss": 1.257,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 405
    },
    {
      "epoch": 0.2547051442910916,
      "grad_norm": 1.7569973468780518,
      "learning_rate": 4.988671922385735e-06,
      "loss": 1.4437,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 406
    },
    {
      "epoch": 0.2553324968632371,
      "grad_norm": 1.8653000593185425,
      "learning_rate": 4.988410212720029e-06,
      "loss": 1.2732,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 407
    },
    {
      "epoch": 0.2559598494353827,
      "grad_norm": 2.0402140617370605,
      "learning_rate": 4.9881455213959405e-06,
      "loss": 1.3318,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 408
    },
    {
      "epoch": 0.2565872020075282,
      "grad_norm": 1.4472616910934448,
      "learning_rate": 4.987877848730627e-06,
      "loss": 1.2535,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 409
    },
    {
      "epoch": 0.2572145545796738,
      "grad_norm": 1.8016327619552612,
      "learning_rate": 4.9876071950448185e-06,
      "loss": 1.4344,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 410
    },
    {
      "epoch": 0.2578419071518193,
      "grad_norm": 1.69692063331604,
      "learning_rate": 4.987333560662817e-06,
      "loss": 1.2204,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 411
    },
    {
      "epoch": 0.2584692597239649,
      "grad_norm": 1.9340769052505493,
      "learning_rate": 4.987056945912498e-06,
      "loss": 1.3763,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 412
    },
    {
      "epoch": 0.2590966122961104,
      "grad_norm": 1.7981717586517334,
      "learning_rate": 4.986777351125304e-06,
      "loss": 1.2048,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 413
    },
    {
      "epoch": 0.25972396486825594,
      "grad_norm": 8.244365692138672,
      "learning_rate": 4.986494776636251e-06,
      "loss": 1.4393,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 414
    },
    {
      "epoch": 0.2603513174404015,
      "grad_norm": 1.6280755996704102,
      "learning_rate": 4.986209222783925e-06,
      "loss": 1.3547,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 415
    },
    {
      "epoch": 0.26097867001254704,
      "grad_norm": 1.932045817375183,
      "learning_rate": 4.985920689910484e-06,
      "loss": 1.3862,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 416
    },
    {
      "epoch": 0.2616060225846926,
      "grad_norm": 2.241225004196167,
      "learning_rate": 4.98562917836165e-06,
      "loss": 1.2758,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 417
    },
    {
      "epoch": 0.26223337515683814,
      "grad_norm": 1.9528926610946655,
      "learning_rate": 4.9853346884867195e-06,
      "loss": 1.5007,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 418
    },
    {
      "epoch": 0.26286072772898367,
      "grad_norm": 2.021772861480713,
      "learning_rate": 4.985037220638556e-06,
      "loss": 1.3199,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 419
    },
    {
      "epoch": 0.26348808030112925,
      "grad_norm": 2.033529758453369,
      "learning_rate": 4.984736775173588e-06,
      "loss": 1.388,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 420
    },
    {
      "epoch": 0.26411543287327477,
      "grad_norm": 2.2411134243011475,
      "learning_rate": 4.984433352451818e-06,
      "loss": 1.2186,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 421
    },
    {
      "epoch": 0.26474278544542035,
      "grad_norm": 1.808010458946228,
      "learning_rate": 4.984126952836811e-06,
      "loss": 1.3176,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 422
    },
    {
      "epoch": 0.26537013801756587,
      "grad_norm": 1.856432318687439,
      "learning_rate": 4.983817576695701e-06,
      "loss": 1.2231,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 423
    },
    {
      "epoch": 0.2659974905897114,
      "grad_norm": 1.4289281368255615,
      "learning_rate": 4.983505224399188e-06,
      "loss": 1.3955,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 424
    },
    {
      "epoch": 0.266624843161857,
      "grad_norm": 1.582891821861267,
      "learning_rate": 4.9831898963215386e-06,
      "loss": 1.2563,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 425
    },
    {
      "epoch": 0.2672521957340025,
      "grad_norm": 1.8547396659851074,
      "learning_rate": 4.982871592840583e-06,
      "loss": 1.3904,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 426
    },
    {
      "epoch": 0.2678795483061481,
      "grad_norm": 1.6647510528564453,
      "learning_rate": 4.982550314337721e-06,
      "loss": 1.2344,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 427
    },
    {
      "epoch": 0.2685069008782936,
      "grad_norm": 1.7917332649230957,
      "learning_rate": 4.9822260611979125e-06,
      "loss": 1.4628,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 428
    },
    {
      "epoch": 0.2691342534504391,
      "grad_norm": 1.7590855360031128,
      "learning_rate": 4.981898833809684e-06,
      "loss": 1.3563,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 429
    },
    {
      "epoch": 0.2697616060225847,
      "grad_norm": 1.706697702407837,
      "learning_rate": 4.981568632565125e-06,
      "loss": 1.2132,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 430
    },
    {
      "epoch": 0.2703889585947302,
      "grad_norm": 1.829269528388977,
      "learning_rate": 4.9812354578598876e-06,
      "loss": 1.4478,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 431
    },
    {
      "epoch": 0.2710163111668758,
      "grad_norm": 1.4316096305847168,
      "learning_rate": 4.98089931009319e-06,
      "loss": 1.2116,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 432
    },
    {
      "epoch": 0.2716436637390213,
      "grad_norm": 1.522438645362854,
      "learning_rate": 4.98056018966781e-06,
      "loss": 1.2716,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 433
    },
    {
      "epoch": 0.2722710163111669,
      "grad_norm": 1.749516248703003,
      "learning_rate": 4.980218096990087e-06,
      "loss": 1.3609,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 434
    },
    {
      "epoch": 0.2728983688833124,
      "grad_norm": 2.074244499206543,
      "learning_rate": 4.979873032469923e-06,
      "loss": 1.1601,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 435
    },
    {
      "epoch": 0.27352572145545795,
      "grad_norm": 2.087233066558838,
      "learning_rate": 4.979524996520781e-06,
      "loss": 1.3867,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 436
    },
    {
      "epoch": 0.2741530740276035,
      "grad_norm": 2.0224671363830566,
      "learning_rate": 4.979173989559684e-06,
      "loss": 1.4594,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 437
    },
    {
      "epoch": 0.27478042659974905,
      "grad_norm": 1.8196134567260742,
      "learning_rate": 4.978820012007213e-06,
      "loss": 1.3352,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 438
    },
    {
      "epoch": 0.27540777917189463,
      "grad_norm": 2.074153423309326,
      "learning_rate": 4.978463064287513e-06,
      "loss": 1.4756,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 439
    },
    {
      "epoch": 0.27603513174404015,
      "grad_norm": 1.6791142225265503,
      "learning_rate": 4.978103146828285e-06,
      "loss": 1.326,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 440
    },
    {
      "epoch": 0.2766624843161857,
      "grad_norm": 2.263253688812256,
      "learning_rate": 4.977740260060787e-06,
      "loss": 1.5105,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 441
    },
    {
      "epoch": 0.27728983688833125,
      "grad_norm": 1.7390024662017822,
      "learning_rate": 4.977374404419838e-06,
      "loss": 1.1394,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 442
    },
    {
      "epoch": 0.2779171894604768,
      "grad_norm": 1.5287703275680542,
      "learning_rate": 4.977005580343811e-06,
      "loss": 1.4403,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 443
    },
    {
      "epoch": 0.27854454203262236,
      "grad_norm": 1.9817328453063965,
      "learning_rate": 4.9766337882746395e-06,
      "loss": 1.3576,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 444
    },
    {
      "epoch": 0.2791718946047679,
      "grad_norm": 3.146590232849121,
      "learning_rate": 4.976259028657812e-06,
      "loss": 1.394,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 445
    },
    {
      "epoch": 0.2797992471769134,
      "grad_norm": 1.9648598432540894,
      "learning_rate": 4.97588130194237e-06,
      "loss": 1.4508,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 446
    },
    {
      "epoch": 0.280426599749059,
      "grad_norm": 1.9081323146820068,
      "learning_rate": 4.975500608580914e-06,
      "loss": 1.3358,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 447
    },
    {
      "epoch": 0.2810539523212045,
      "grad_norm": 1.8519835472106934,
      "learning_rate": 4.975116949029598e-06,
      "loss": 1.3292,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 448
    },
    {
      "epoch": 0.2816813048933501,
      "grad_norm": 1.6948096752166748,
      "learning_rate": 4.974730323748129e-06,
      "loss": 1.1885,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 449
    },
    {
      "epoch": 0.2823086574654956,
      "grad_norm": 1.8216960430145264,
      "learning_rate": 4.9743407331997696e-06,
      "loss": 1.3733,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 450
    },
    {
      "epoch": 0.28293601003764113,
      "grad_norm": 1.7822445631027222,
      "learning_rate": 4.973948177851333e-06,
      "loss": 1.4895,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 451
    },
    {
      "epoch": 0.2835633626097867,
      "grad_norm": 1.525521159172058,
      "learning_rate": 4.973552658173186e-06,
      "loss": 1.1041,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 452
    },
    {
      "epoch": 0.28419071518193223,
      "grad_norm": 1.5609532594680786,
      "learning_rate": 4.973154174639249e-06,
      "loss": 1.2549,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 453
    },
    {
      "epoch": 0.2848180677540778,
      "grad_norm": 1.5253825187683105,
      "learning_rate": 4.972752727726992e-06,
      "loss": 1.2703,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 454
    },
    {
      "epoch": 0.28544542032622333,
      "grad_norm": 1.5149940252304077,
      "learning_rate": 4.972348317917434e-06,
      "loss": 1.2457,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 455
    },
    {
      "epoch": 0.2860727728983689,
      "grad_norm": 1.466367483139038,
      "learning_rate": 4.971940945695148e-06,
      "loss": 1.2483,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 456
    },
    {
      "epoch": 0.28670012547051443,
      "grad_norm": 2.159257650375366,
      "learning_rate": 4.971530611548254e-06,
      "loss": 1.2765,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 457
    },
    {
      "epoch": 0.28732747804265996,
      "grad_norm": 1.6494646072387695,
      "learning_rate": 4.971117315968423e-06,
      "loss": 1.263,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 458
    },
    {
      "epoch": 0.28795483061480553,
      "grad_norm": 1.8701457977294922,
      "learning_rate": 4.970701059450872e-06,
      "loss": 1.392,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 459
    },
    {
      "epoch": 0.28858218318695106,
      "grad_norm": 1.613142490386963,
      "learning_rate": 4.970281842494368e-06,
      "loss": 1.1785,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 460
    },
    {
      "epoch": 0.28920953575909664,
      "grad_norm": 1.840915560722351,
      "learning_rate": 4.969859665601223e-06,
      "loss": 1.3366,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 461
    },
    {
      "epoch": 0.28983688833124216,
      "grad_norm": 1.8302453756332397,
      "learning_rate": 4.969434529277299e-06,
      "loss": 1.4625,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 462
    },
    {
      "epoch": 0.2904642409033877,
      "grad_norm": 1.698702096939087,
      "learning_rate": 4.969006434032e-06,
      "loss": 1.3302,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 463
    },
    {
      "epoch": 0.29109159347553326,
      "grad_norm": 2.026204824447632,
      "learning_rate": 4.96857538037828e-06,
      "loss": 1.2724,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 464
    },
    {
      "epoch": 0.2917189460476788,
      "grad_norm": 1.7603907585144043,
      "learning_rate": 4.968141368832633e-06,
      "loss": 1.2931,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 465
    },
    {
      "epoch": 0.29234629861982436,
      "grad_norm": 2.7061665058135986,
      "learning_rate": 4.9677043999151e-06,
      "loss": 1.3099,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 466
    },
    {
      "epoch": 0.2929736511919699,
      "grad_norm": 1.795343041419983,
      "learning_rate": 4.967264474149267e-06,
      "loss": 1.2425,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 467
    },
    {
      "epoch": 0.2936010037641154,
      "grad_norm": 1.4162352085113525,
      "learning_rate": 4.966821592062259e-06,
      "loss": 1.3789,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 468
    },
    {
      "epoch": 0.294228356336261,
      "grad_norm": 1.8810973167419434,
      "learning_rate": 4.966375754184746e-06,
      "loss": 1.2759,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 469
    },
    {
      "epoch": 0.2948557089084065,
      "grad_norm": 1.919083833694458,
      "learning_rate": 4.965926961050939e-06,
      "loss": 1.2544,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 470
    },
    {
      "epoch": 0.2954830614805521,
      "grad_norm": 1.7765024900436401,
      "learning_rate": 4.965475213198589e-06,
      "loss": 1.2828,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 471
    },
    {
      "epoch": 0.2961104140526976,
      "grad_norm": 1.5920195579528809,
      "learning_rate": 4.965020511168991e-06,
      "loss": 1.3994,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 472
    },
    {
      "epoch": 0.29673776662484314,
      "grad_norm": 1.7678951025009155,
      "learning_rate": 4.964562855506976e-06,
      "loss": 1.3297,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 473
    },
    {
      "epoch": 0.2973651191969887,
      "grad_norm": 1.9965953826904297,
      "learning_rate": 4.964102246760915e-06,
      "loss": 1.271,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 474
    },
    {
      "epoch": 0.29799247176913424,
      "grad_norm": 2.709869623184204,
      "learning_rate": 4.963638685482717e-06,
      "loss": 1.2818,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 475
    },
    {
      "epoch": 0.2986198243412798,
      "grad_norm": 1.530746579170227,
      "learning_rate": 4.963172172227831e-06,
      "loss": 1.2644,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 476
    },
    {
      "epoch": 0.29924717691342534,
      "grad_norm": 1.8586333990097046,
      "learning_rate": 4.9627027075552425e-06,
      "loss": 1.3388,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 477
    },
    {
      "epoch": 0.2998745294855709,
      "grad_norm": 1.7000927925109863,
      "learning_rate": 4.96223029202747e-06,
      "loss": 1.3127,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 478
    },
    {
      "epoch": 0.30050188205771644,
      "grad_norm": 1.7719290256500244,
      "learning_rate": 4.961754926210572e-06,
      "loss": 1.2862,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 479
    },
    {
      "epoch": 0.30112923462986196,
      "grad_norm": 2.3949620723724365,
      "learning_rate": 4.961276610674141e-06,
      "loss": 1.2543,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 480
    },
    {
      "epoch": 0.30175658720200754,
      "grad_norm": 1.7187825441360474,
      "learning_rate": 4.960795345991302e-06,
      "loss": 1.1969,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 481
    },
    {
      "epoch": 0.30238393977415307,
      "grad_norm": 1.8787022829055786,
      "learning_rate": 4.960311132738715e-06,
      "loss": 1.2647,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 482
    },
    {
      "epoch": 0.30301129234629864,
      "grad_norm": 1.5814067125320435,
      "learning_rate": 4.959823971496575e-06,
      "loss": 1.2246,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 483
    },
    {
      "epoch": 0.30363864491844417,
      "grad_norm": 1.7932416200637817,
      "learning_rate": 4.959333862848605e-06,
      "loss": 1.3038,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 484
    },
    {
      "epoch": 0.3042659974905897,
      "grad_norm": 2.22336483001709,
      "learning_rate": 4.958840807382064e-06,
      "loss": 1.3414,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 485
    },
    {
      "epoch": 0.30489335006273527,
      "grad_norm": 1.9425820112228394,
      "learning_rate": 4.958344805687739e-06,
      "loss": 1.2395,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 486
    },
    {
      "epoch": 0.3055207026348808,
      "grad_norm": 1.7737921476364136,
      "learning_rate": 4.9578458583599495e-06,
      "loss": 1.2689,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 487
    },
    {
      "epoch": 0.30614805520702637,
      "grad_norm": 1.9404109716415405,
      "learning_rate": 4.957343965996542e-06,
      "loss": 1.302,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 488
    },
    {
      "epoch": 0.3067754077791719,
      "grad_norm": 1.6181503534317017,
      "learning_rate": 4.956839129198892e-06,
      "loss": 1.3497,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 489
    },
    {
      "epoch": 0.3074027603513174,
      "grad_norm": 2.0537965297698975,
      "learning_rate": 4.956331348571907e-06,
      "loss": 1.3725,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 490
    },
    {
      "epoch": 0.308030112923463,
      "grad_norm": 1.6368733644485474,
      "learning_rate": 4.955820624724018e-06,
      "loss": 1.2585,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 491
    },
    {
      "epoch": 0.3086574654956085,
      "grad_norm": 1.745303750038147,
      "learning_rate": 4.955306958267183e-06,
      "loss": 1.2816,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 492
    },
    {
      "epoch": 0.3092848180677541,
      "grad_norm": 1.9312869310379028,
      "learning_rate": 4.954790349816886e-06,
      "loss": 1.2817,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 493
    },
    {
      "epoch": 0.3099121706398996,
      "grad_norm": 1.7578679323196411,
      "learning_rate": 4.954270799992138e-06,
      "loss": 1.3172,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 494
    },
    {
      "epoch": 0.31053952321204514,
      "grad_norm": 1.7783771753311157,
      "learning_rate": 4.953748309415472e-06,
      "loss": 1.3384,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 495
    },
    {
      "epoch": 0.3111668757841907,
      "grad_norm": 1.5775705575942993,
      "learning_rate": 4.9532228787129456e-06,
      "loss": 1.1949,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 496
    },
    {
      "epoch": 0.31179422835633624,
      "grad_norm": 1.4879812002182007,
      "learning_rate": 4.9526945085141405e-06,
      "loss": 1.2281,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 497
    },
    {
      "epoch": 0.3124215809284818,
      "grad_norm": 1.920745611190796,
      "learning_rate": 4.952163199452159e-06,
      "loss": 1.3227,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 498
    },
    {
      "epoch": 0.31304893350062735,
      "grad_norm": 1.9175952672958374,
      "learning_rate": 4.951628952163625e-06,
      "loss": 1.359,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 499
    },
    {
      "epoch": 0.3136762860727729,
      "grad_norm": 1.4619158506393433,
      "learning_rate": 4.951091767288683e-06,
      "loss": 1.2466,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 500
    },
    {
      "epoch": 0.31430363864491845,
      "grad_norm": 1.8956916332244873,
      "learning_rate": 4.950551645470998e-06,
      "loss": 1.3183,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 501
    },
    {
      "epoch": 0.31493099121706397,
      "grad_norm": 1.4829291105270386,
      "learning_rate": 4.950008587357755e-06,
      "loss": 1.2069,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 502
    },
    {
      "epoch": 0.31555834378920955,
      "grad_norm": 2.078758955001831,
      "learning_rate": 4.949462593599654e-06,
      "loss": 1.2592,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 503
    },
    {
      "epoch": 0.3161856963613551,
      "grad_norm": 1.880047082901001,
      "learning_rate": 4.948913664850917e-06,
      "loss": 1.2492,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 504
    },
    {
      "epoch": 0.31681304893350065,
      "grad_norm": 1.900512933731079,
      "learning_rate": 4.9483618017692784e-06,
      "loss": 1.2715,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 505
    },
    {
      "epoch": 0.3174404015056462,
      "grad_norm": 1.6285508871078491,
      "learning_rate": 4.947807005015991e-06,
      "loss": 1.2036,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 506
    },
    {
      "epoch": 0.3180677540777917,
      "grad_norm": 1.93606436252594,
      "learning_rate": 4.947249275255823e-06,
      "loss": 1.372,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 507
    },
    {
      "epoch": 0.3186951066499373,
      "grad_norm": 1.9985902309417725,
      "learning_rate": 4.9466886131570565e-06,
      "loss": 1.3982,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 508
    },
    {
      "epoch": 0.3193224592220828,
      "grad_norm": 1.8069493770599365,
      "learning_rate": 4.946125019391486e-06,
      "loss": 1.3255,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 509
    },
    {
      "epoch": 0.3199498117942284,
      "grad_norm": 1.8266329765319824,
      "learning_rate": 4.94555849463442e-06,
      "loss": 1.2351,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 510
    },
    {
      "epoch": 0.3205771643663739,
      "grad_norm": 2.004852056503296,
      "learning_rate": 4.94498903956468e-06,
      "loss": 1.2493,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 511
    },
    {
      "epoch": 0.3212045169385194,
      "grad_norm": 1.9460428953170776,
      "learning_rate": 4.944416654864596e-06,
      "loss": 1.4126,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 512
    },
    {
      "epoch": 0.321831869510665,
      "grad_norm": 1.6291266679763794,
      "learning_rate": 4.943841341220011e-06,
      "loss": 1.4045,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 513
    },
    {
      "epoch": 0.3224592220828105,
      "grad_norm": 1.5005308389663696,
      "learning_rate": 4.943263099320275e-06,
      "loss": 1.4401,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 514
    },
    {
      "epoch": 0.3230865746549561,
      "grad_norm": 1.3146636486053467,
      "learning_rate": 4.942681929858249e-06,
      "loss": 1.3803,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 515
    },
    {
      "epoch": 0.3237139272271016,
      "grad_norm": 1.5461806058883667,
      "learning_rate": 4.942097833530299e-06,
      "loss": 1.2672,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 516
    },
    {
      "epoch": 0.32434127979924715,
      "grad_norm": 1.8553651571273804,
      "learning_rate": 4.9415108110363025e-06,
      "loss": 1.3967,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 517
    },
    {
      "epoch": 0.32496863237139273,
      "grad_norm": 1.5972548723220825,
      "learning_rate": 4.940920863079639e-06,
      "loss": 1.2384,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 518
    },
    {
      "epoch": 0.32559598494353825,
      "grad_norm": 1.6900385618209839,
      "learning_rate": 4.940327990367196e-06,
      "loss": 1.2429,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 519
    },
    {
      "epoch": 0.32622333751568383,
      "grad_norm": 1.8452892303466797,
      "learning_rate": 4.939732193609361e-06,
      "loss": 1.3469,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 520
    },
    {
      "epoch": 0.32685069008782935,
      "grad_norm": 1.7989273071289062,
      "learning_rate": 4.939133473520033e-06,
      "loss": 1.3273,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 521
    },
    {
      "epoch": 0.32747804265997493,
      "grad_norm": 1.6270055770874023,
      "learning_rate": 4.9385318308166065e-06,
      "loss": 1.2904,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 522
    },
    {
      "epoch": 0.32810539523212046,
      "grad_norm": 1.8746994733810425,
      "learning_rate": 4.937927266219982e-06,
      "loss": 1.4421,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 523
    },
    {
      "epoch": 0.328732747804266,
      "grad_norm": 1.5117107629776,
      "learning_rate": 4.937319780454559e-06,
      "loss": 1.2808,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 524
    },
    {
      "epoch": 0.32936010037641156,
      "grad_norm": 2.1387412548065186,
      "learning_rate": 4.936709374248238e-06,
      "loss": 1.1458,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 525
    },
    {
      "epoch": 0.3299874529485571,
      "grad_norm": 1.901404857635498,
      "learning_rate": 4.936096048332421e-06,
      "loss": 1.418,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 526
    },
    {
      "epoch": 0.33061480552070266,
      "grad_norm": 11.465993881225586,
      "learning_rate": 4.935479803442002e-06,
      "loss": 1.325,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 527
    },
    {
      "epoch": 0.3312421580928482,
      "grad_norm": 1.0712916851043701,
      "learning_rate": 4.934860640315381e-06,
      "loss": 1.1727,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 528
    },
    {
      "epoch": 0.3318695106649937,
      "grad_norm": 1.8514409065246582,
      "learning_rate": 4.934238559694448e-06,
      "loss": 1.3749,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 529
    },
    {
      "epoch": 0.3324968632371393,
      "grad_norm": 1.7288488149642944,
      "learning_rate": 4.933613562324592e-06,
      "loss": 1.2642,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 530
    },
    {
      "epoch": 0.3331242158092848,
      "grad_norm": 1.8730841875076294,
      "learning_rate": 4.932985648954696e-06,
      "loss": 1.3008,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 531
    },
    {
      "epoch": 0.3337515683814304,
      "grad_norm": 1.5771578550338745,
      "learning_rate": 4.932354820337138e-06,
      "loss": 1.2411,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 532
    },
    {
      "epoch": 0.3343789209535759,
      "grad_norm": 1.9279652833938599,
      "learning_rate": 4.931721077227788e-06,
      "loss": 1.209,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 533
    },
    {
      "epoch": 0.33500627352572143,
      "grad_norm": 2.032771110534668,
      "learning_rate": 4.931084420386009e-06,
      "loss": 1.2686,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 534
    },
    {
      "epoch": 0.335633626097867,
      "grad_norm": 1.446815013885498,
      "learning_rate": 4.9304448505746535e-06,
      "loss": 1.2038,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 535
    },
    {
      "epoch": 0.33626097867001253,
      "grad_norm": 2.149000644683838,
      "learning_rate": 4.929802368560066e-06,
      "loss": 1.2825,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 536
    },
    {
      "epoch": 0.3368883312421581,
      "grad_norm": 1.8099359273910522,
      "learning_rate": 4.929156975112081e-06,
      "loss": 1.2651,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 537
    },
    {
      "epoch": 0.33751568381430364,
      "grad_norm": 2.0547146797180176,
      "learning_rate": 4.92850867100402e-06,
      "loss": 1.3422,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 538
    },
    {
      "epoch": 0.33814303638644916,
      "grad_norm": 1.6774853467941284,
      "learning_rate": 4.927857457012693e-06,
      "loss": 1.4274,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 539
    },
    {
      "epoch": 0.33877038895859474,
      "grad_norm": 1.6937648057937622,
      "learning_rate": 4.927203333918396e-06,
      "loss": 1.2117,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 540
    },
    {
      "epoch": 0.33939774153074026,
      "grad_norm": 1.93661630153656,
      "learning_rate": 4.926546302504912e-06,
      "loss": 1.2139,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 541
    },
    {
      "epoch": 0.34002509410288584,
      "grad_norm": 2.4098830223083496,
      "learning_rate": 4.925886363559507e-06,
      "loss": 1.2046,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 542
    },
    {
      "epoch": 0.34065244667503136,
      "grad_norm": 1.7324174642562866,
      "learning_rate": 4.925223517872934e-06,
      "loss": 1.3581,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 543
    },
    {
      "epoch": 0.34127979924717694,
      "grad_norm": 2.022440195083618,
      "learning_rate": 4.924557766239424e-06,
      "loss": 1.246,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 544
    },
    {
      "epoch": 0.34190715181932246,
      "grad_norm": 1.8990328311920166,
      "learning_rate": 4.923889109456693e-06,
      "loss": 1.3062,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 545
    },
    {
      "epoch": 0.342534504391468,
      "grad_norm": 2.065661668777466,
      "learning_rate": 4.923217548325939e-06,
      "loss": 1.2956,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 546
    },
    {
      "epoch": 0.34316185696361357,
      "grad_norm": 1.897823691368103,
      "learning_rate": 4.92254308365184e-06,
      "loss": 1.2868,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 547
    },
    {
      "epoch": 0.3437892095357591,
      "grad_norm": 1.8894264698028564,
      "learning_rate": 4.9218657162425496e-06,
      "loss": 1.2409,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 548
    },
    {
      "epoch": 0.34441656210790467,
      "grad_norm": 1.9177500009536743,
      "learning_rate": 4.921185446909702e-06,
      "loss": 1.3804,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 549
    },
    {
      "epoch": 0.3450439146800502,
      "grad_norm": 1.6005748510360718,
      "learning_rate": 4.920502276468408e-06,
      "loss": 1.1235,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 550
    },
    {
      "epoch": 0.3456712672521957,
      "grad_norm": 1.9727932214736938,
      "learning_rate": 4.919816205737257e-06,
      "loss": 1.2998,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 551
    },
    {
      "epoch": 0.3462986198243413,
      "grad_norm": 1.958331823348999,
      "learning_rate": 4.919127235538309e-06,
      "loss": 1.1596,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 552
    },
    {
      "epoch": 0.3469259723964868,
      "grad_norm": 1.9714314937591553,
      "learning_rate": 4.918435366697102e-06,
      "loss": 1.2857,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 553
    },
    {
      "epoch": 0.3475533249686324,
      "grad_norm": 1.689508318901062,
      "learning_rate": 4.917740600042645e-06,
      "loss": 1.4039,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 554
    },
    {
      "epoch": 0.3481806775407779,
      "grad_norm": 2.0137124061584473,
      "learning_rate": 4.91704293640742e-06,
      "loss": 1.2008,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 555
    },
    {
      "epoch": 0.34880803011292344,
      "grad_norm": 1.9370338916778564,
      "learning_rate": 4.9163423766273815e-06,
      "loss": 1.3269,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 556
    },
    {
      "epoch": 0.349435382685069,
      "grad_norm": 2.1233015060424805,
      "learning_rate": 4.915638921541952e-06,
      "loss": 1.4172,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 557
    },
    {
      "epoch": 0.35006273525721454,
      "grad_norm": 1.8814067840576172,
      "learning_rate": 4.914932571994024e-06,
      "loss": 1.2576,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 558
    },
    {
      "epoch": 0.3506900878293601,
      "grad_norm": 1.962165117263794,
      "learning_rate": 4.9142233288299595e-06,
      "loss": 1.2706,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 559
    },
    {
      "epoch": 0.35131744040150564,
      "grad_norm": 1.8706358671188354,
      "learning_rate": 4.913511192899586e-06,
      "loss": 1.3171,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 560
    },
    {
      "epoch": 0.35194479297365117,
      "grad_norm": 1.8167012929916382,
      "learning_rate": 4.9127961650561966e-06,
      "loss": 1.4102,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 561
    },
    {
      "epoch": 0.35257214554579674,
      "grad_norm": 2.4058690071105957,
      "learning_rate": 4.912078246156552e-06,
      "loss": 1.1793,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 562
    },
    {
      "epoch": 0.35319949811794227,
      "grad_norm": 1.899332880973816,
      "learning_rate": 4.911357437060875e-06,
      "loss": 1.2112,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 563
    },
    {
      "epoch": 0.35382685069008785,
      "grad_norm": 2.099212169647217,
      "learning_rate": 4.9106337386328524e-06,
      "loss": 1.3686,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 564
    },
    {
      "epoch": 0.35445420326223337,
      "grad_norm": 1.7897380590438843,
      "learning_rate": 4.909907151739634e-06,
      "loss": 1.3741,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 565
    },
    {
      "epoch": 0.35508155583437895,
      "grad_norm": 1.8345693349838257,
      "learning_rate": 4.909177677251828e-06,
      "loss": 1.3606,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 566
    },
    {
      "epoch": 0.35570890840652447,
      "grad_norm": 1.7460086345672607,
      "learning_rate": 4.908445316043503e-06,
      "loss": 1.3124,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 567
    },
    {
      "epoch": 0.35633626097867,
      "grad_norm": 1.9590387344360352,
      "learning_rate": 4.90771006899219e-06,
      "loss": 1.3128,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 568
    },
    {
      "epoch": 0.3569636135508156,
      "grad_norm": 1.4945135116577148,
      "learning_rate": 4.906971936978874e-06,
      "loss": 1.4596,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 569
    },
    {
      "epoch": 0.3575909661229611,
      "grad_norm": 1.649003267288208,
      "learning_rate": 4.906230920887998e-06,
      "loss": 1.2446,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 570
    },
    {
      "epoch": 0.3582183186951067,
      "grad_norm": 1.7148494720458984,
      "learning_rate": 4.905487021607462e-06,
      "loss": 1.2175,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 571
    },
    {
      "epoch": 0.3588456712672522,
      "grad_norm": 1.9987525939941406,
      "learning_rate": 4.904740240028618e-06,
      "loss": 1.3247,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 572
    },
    {
      "epoch": 0.3594730238393977,
      "grad_norm": 1.8224921226501465,
      "learning_rate": 4.903990577046275e-06,
      "loss": 1.455,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 573
    },
    {
      "epoch": 0.3601003764115433,
      "grad_norm": 1.7764469385147095,
      "learning_rate": 4.903238033558692e-06,
      "loss": 1.4298,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 574
    },
    {
      "epoch": 0.3607277289836888,
      "grad_norm": 1.8746356964111328,
      "learning_rate": 4.90248261046758e-06,
      "loss": 1.1855,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 575
    },
    {
      "epoch": 0.3613550815558344,
      "grad_norm": 1.7384833097457886,
      "learning_rate": 4.901724308678102e-06,
      "loss": 1.2095,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 576
    },
    {
      "epoch": 0.3619824341279799,
      "grad_norm": 1.754146695137024,
      "learning_rate": 4.900963129098868e-06,
      "loss": 1.2735,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 577
    },
    {
      "epoch": 0.36260978670012545,
      "grad_norm": 1.7974838018417358,
      "learning_rate": 4.900199072641937e-06,
      "loss": 1.2829,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 578
    },
    {
      "epoch": 0.363237139272271,
      "grad_norm": 1.8893485069274902,
      "learning_rate": 4.899432140222816e-06,
      "loss": 1.2911,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 579
    },
    {
      "epoch": 0.36386449184441655,
      "grad_norm": 1.7964454889297485,
      "learning_rate": 4.898662332760456e-06,
      "loss": 1.2089,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 580
    },
    {
      "epoch": 0.3644918444165621,
      "grad_norm": 1.9865061044692993,
      "learning_rate": 4.897889651177257e-06,
      "loss": 1.2704,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 581
    },
    {
      "epoch": 0.36511919698870765,
      "grad_norm": 1.8877671957015991,
      "learning_rate": 4.897114096399058e-06,
      "loss": 1.2918,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 582
    },
    {
      "epoch": 0.3657465495608532,
      "grad_norm": 1.8955698013305664,
      "learning_rate": 4.896335669355143e-06,
      "loss": 1.3619,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 583
    },
    {
      "epoch": 0.36637390213299875,
      "grad_norm": 1.6800644397735596,
      "learning_rate": 4.895554370978238e-06,
      "loss": 1.174,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 584
    },
    {
      "epoch": 0.3670012547051443,
      "grad_norm": 1.6480708122253418,
      "learning_rate": 4.894770202204509e-06,
      "loss": 1.1286,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 585
    },
    {
      "epoch": 0.36762860727728985,
      "grad_norm": 1.616126537322998,
      "learning_rate": 4.8939831639735594e-06,
      "loss": 1.2139,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 586
    },
    {
      "epoch": 0.3682559598494354,
      "grad_norm": 2.3213493824005127,
      "learning_rate": 4.893193257228436e-06,
      "loss": 1.4502,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 587
    },
    {
      "epoch": 0.36888331242158096,
      "grad_norm": 1.7523118257522583,
      "learning_rate": 4.892400482915617e-06,
      "loss": 1.3381,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 588
    },
    {
      "epoch": 0.3695106649937265,
      "grad_norm": 2.03842830657959,
      "learning_rate": 4.89160484198502e-06,
      "loss": 1.2769,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 589
    },
    {
      "epoch": 0.370138017565872,
      "grad_norm": 1.9613364934921265,
      "learning_rate": 4.890806335389996e-06,
      "loss": 1.1924,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 590
    },
    {
      "epoch": 0.3707653701380176,
      "grad_norm": 1.6123595237731934,
      "learning_rate": 4.89000496408733e-06,
      "loss": 1.2848,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 591
    },
    {
      "epoch": 0.3713927227101631,
      "grad_norm": 1.4787251949310303,
      "learning_rate": 4.889200729037241e-06,
      "loss": 1.3938,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 592
    },
    {
      "epoch": 0.3720200752823087,
      "grad_norm": 1.888031005859375,
      "learning_rate": 4.888393631203377e-06,
      "loss": 1.3367,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 593
    },
    {
      "epoch": 0.3726474278544542,
      "grad_norm": 1.948343276977539,
      "learning_rate": 4.887583671552815e-06,
      "loss": 1.3012,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 594
    },
    {
      "epoch": 0.37327478042659973,
      "grad_norm": 1.8316664695739746,
      "learning_rate": 4.886770851056068e-06,
      "loss": 1.2778,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 595
    },
    {
      "epoch": 0.3739021329987453,
      "grad_norm": 1.860657811164856,
      "learning_rate": 4.88595517068707e-06,
      "loss": 1.1374,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 596
    },
    {
      "epoch": 0.37452948557089083,
      "grad_norm": 1.8296133279800415,
      "learning_rate": 4.885136631423184e-06,
      "loss": 1.3081,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 597
    },
    {
      "epoch": 0.3751568381430364,
      "grad_norm": 1.8158042430877686,
      "learning_rate": 4.884315234245198e-06,
      "loss": 1.3434,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 598
    },
    {
      "epoch": 0.37578419071518193,
      "grad_norm": 1.9907175302505493,
      "learning_rate": 4.883490980137327e-06,
      "loss": 1.2084,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 599
    },
    {
      "epoch": 0.37641154328732745,
      "grad_norm": 2.1032094955444336,
      "learning_rate": 4.882663870087206e-06,
      "loss": 1.2768,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 600
    },
    {
      "epoch": 0.37703889585947303,
      "grad_norm": 1.874523401260376,
      "learning_rate": 4.881833905085892e-06,
      "loss": 1.2316,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 601
    },
    {
      "epoch": 0.37766624843161856,
      "grad_norm": 1.8574604988098145,
      "learning_rate": 4.881001086127866e-06,
      "loss": 1.3143,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 602
    },
    {
      "epoch": 0.37829360100376414,
      "grad_norm": 3.8513078689575195,
      "learning_rate": 4.880165414211027e-06,
      "loss": 1.2601,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 603
    },
    {
      "epoch": 0.37892095357590966,
      "grad_norm": 1.8717108964920044,
      "learning_rate": 4.8793268903366905e-06,
      "loss": 1.1758,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 604
    },
    {
      "epoch": 0.3795483061480552,
      "grad_norm": 1.7282662391662598,
      "learning_rate": 4.878485515509594e-06,
      "loss": 1.2174,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 605
    },
    {
      "epoch": 0.38017565872020076,
      "grad_norm": 2.066521167755127,
      "learning_rate": 4.8776412907378845e-06,
      "loss": 1.2972,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 606
    },
    {
      "epoch": 0.3808030112923463,
      "grad_norm": 1.7320374250411987,
      "learning_rate": 4.876794217033129e-06,
      "loss": 1.211,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 607
    },
    {
      "epoch": 0.38143036386449186,
      "grad_norm": 1.6623154878616333,
      "learning_rate": 4.875944295410308e-06,
      "loss": 1.2735,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 608
    },
    {
      "epoch": 0.3820577164366374,
      "grad_norm": 1.917044997215271,
      "learning_rate": 4.875091526887813e-06,
      "loss": 1.3399,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 609
    },
    {
      "epoch": 0.38268506900878296,
      "grad_norm": 2.031456470489502,
      "learning_rate": 4.874235912487444e-06,
      "loss": 1.1696,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 610
    },
    {
      "epoch": 0.3833124215809285,
      "grad_norm": 1.8612998723983765,
      "learning_rate": 4.8733774532344164e-06,
      "loss": 1.4226,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 611
    },
    {
      "epoch": 0.383939774153074,
      "grad_norm": 1.7743910551071167,
      "learning_rate": 4.872516150157351e-06,
      "loss": 1.2365,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 612
    },
    {
      "epoch": 0.3845671267252196,
      "grad_norm": 1.7323930263519287,
      "learning_rate": 4.871652004288275e-06,
      "loss": 1.1926,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 613
    },
    {
      "epoch": 0.3851944792973651,
      "grad_norm": 1.93533194065094,
      "learning_rate": 4.870785016662627e-06,
      "loss": 1.1668,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 614
    },
    {
      "epoch": 0.3858218318695107,
      "grad_norm": 3.5375280380249023,
      "learning_rate": 4.869915188319244e-06,
      "loss": 1.3051,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 615
    },
    {
      "epoch": 0.3864491844416562,
      "grad_norm": 2.405059814453125,
      "learning_rate": 4.8690425203003735e-06,
      "loss": 1.4259,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 616
    },
    {
      "epoch": 0.38707653701380174,
      "grad_norm": 1.943349003791809,
      "learning_rate": 4.86816701365166e-06,
      "loss": 1.202,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 617
    },
    {
      "epoch": 0.3877038895859473,
      "grad_norm": 1.9087837934494019,
      "learning_rate": 4.867288669422151e-06,
      "loss": 1.1793,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 618
    },
    {
      "epoch": 0.38833124215809284,
      "grad_norm": 1.8294497728347778,
      "learning_rate": 4.866407488664296e-06,
      "loss": 1.1885,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 619
    },
    {
      "epoch": 0.3889585947302384,
      "grad_norm": 1.836368441581726,
      "learning_rate": 4.865523472433942e-06,
      "loss": 1.2184,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 620
    },
    {
      "epoch": 0.38958594730238394,
      "grad_norm": 1.4587712287902832,
      "learning_rate": 4.8646366217903314e-06,
      "loss": 1.1466,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 621
    },
    {
      "epoch": 0.39021329987452946,
      "grad_norm": 2.0479862689971924,
      "learning_rate": 4.863746937796107e-06,
      "loss": 1.3139,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 622
    },
    {
      "epoch": 0.39084065244667504,
      "grad_norm": 1.8100470304489136,
      "learning_rate": 4.862854421517303e-06,
      "loss": 1.3105,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 623
    },
    {
      "epoch": 0.39146800501882056,
      "grad_norm": 1.9524677991867065,
      "learning_rate": 4.861959074023348e-06,
      "loss": 1.2959,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 624
    },
    {
      "epoch": 0.39209535759096614,
      "grad_norm": 1.7042853832244873,
      "learning_rate": 4.8610608963870665e-06,
      "loss": 1.3388,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 625
    },
    {
      "epoch": 0.39272271016311167,
      "grad_norm": 1.6724315881729126,
      "learning_rate": 4.860159889684668e-06,
      "loss": 1.22,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 626
    },
    {
      "epoch": 0.3933500627352572,
      "grad_norm": 1.8373382091522217,
      "learning_rate": 4.859256054995758e-06,
      "loss": 1.2115,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 627
    },
    {
      "epoch": 0.39397741530740277,
      "grad_norm": 1.9262843132019043,
      "learning_rate": 4.8583493934033245e-06,
      "loss": 1.3984,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 628
    },
    {
      "epoch": 0.3946047678795483,
      "grad_norm": 2.0423967838287354,
      "learning_rate": 4.857439905993748e-06,
      "loss": 1.3022,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 629
    },
    {
      "epoch": 0.39523212045169387,
      "grad_norm": 2.023704767227173,
      "learning_rate": 4.856527593856794e-06,
      "loss": 1.2688,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 630
    },
    {
      "epoch": 0.3958594730238394,
      "grad_norm": 2.0022521018981934,
      "learning_rate": 4.855612458085608e-06,
      "loss": 1.3507,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 631
    },
    {
      "epoch": 0.39648682559598497,
      "grad_norm": 2.0341274738311768,
      "learning_rate": 4.854694499776726e-06,
      "loss": 1.2695,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 632
    },
    {
      "epoch": 0.3971141781681305,
      "grad_norm": 1.9634671211242676,
      "learning_rate": 4.853773720030059e-06,
      "loss": 1.1983,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 633
    },
    {
      "epoch": 0.397741530740276,
      "grad_norm": 1.9861356019973755,
      "learning_rate": 4.8528501199489045e-06,
      "loss": 1.5236,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 634
    },
    {
      "epoch": 0.3983688833124216,
      "grad_norm": 1.6151918172836304,
      "learning_rate": 4.851923700639935e-06,
      "loss": 1.1524,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 635
    },
    {
      "epoch": 0.3989962358845671,
      "grad_norm": 1.7349764108657837,
      "learning_rate": 4.850994463213206e-06,
      "loss": 1.2273,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 636
    },
    {
      "epoch": 0.3996235884567127,
      "grad_norm": 2.104606866836548,
      "learning_rate": 4.850062408782144e-06,
      "loss": 1.2654,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 637
    },
    {
      "epoch": 0.4002509410288582,
      "grad_norm": 1.8367067575454712,
      "learning_rate": 4.849127538463554e-06,
      "loss": 1.344,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 638
    },
    {
      "epoch": 0.40087829360100374,
      "grad_norm": 1.7443346977233887,
      "learning_rate": 4.848189853377615e-06,
      "loss": 1.2501,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 639
    },
    {
      "epoch": 0.4015056461731493,
      "grad_norm": 1.8433117866516113,
      "learning_rate": 4.8472493546478794e-06,
      "loss": 1.3459,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 640
    },
    {
      "epoch": 0.40213299874529485,
      "grad_norm": 1.8732486963272095,
      "learning_rate": 4.846306043401268e-06,
      "loss": 1.2625,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 641
    },
    {
      "epoch": 0.4027603513174404,
      "grad_norm": 1.844283103942871,
      "learning_rate": 4.845359920768074e-06,
      "loss": 1.2765,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 642
    },
    {
      "epoch": 0.40338770388958595,
      "grad_norm": 1.86635422706604,
      "learning_rate": 4.8444109878819585e-06,
      "loss": 1.4083,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 643
    },
    {
      "epoch": 0.40401505646173147,
      "grad_norm": 1.824976921081543,
      "learning_rate": 4.843459245879952e-06,
      "loss": 1.3423,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 644
    },
    {
      "epoch": 0.40464240903387705,
      "grad_norm": 1.6277782917022705,
      "learning_rate": 4.842504695902447e-06,
      "loss": 1.131,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 645
    },
    {
      "epoch": 0.40526976160602257,
      "grad_norm": 1.6990870237350464,
      "learning_rate": 4.841547339093203e-06,
      "loss": 1.2938,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 646
    },
    {
      "epoch": 0.40589711417816815,
      "grad_norm": 1.830122470855713,
      "learning_rate": 4.8405871765993435e-06,
      "loss": 1.239,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 647
    },
    {
      "epoch": 0.4065244667503137,
      "grad_norm": 1.927756905555725,
      "learning_rate": 4.839624209571352e-06,
      "loss": 1.3042,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 648
    },
    {
      "epoch": 0.4071518193224592,
      "grad_norm": 1.8679585456848145,
      "learning_rate": 4.838658439163072e-06,
      "loss": 1.3488,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 649
    },
    {
      "epoch": 0.4077791718946048,
      "grad_norm": 1.6993918418884277,
      "learning_rate": 4.837689866531709e-06,
      "loss": 1.4785,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 650
    },
    {
      "epoch": 0.4084065244667503,
      "grad_norm": 1.6975336074829102,
      "learning_rate": 4.836718492837823e-06,
      "loss": 1.2338,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 651
    },
    {
      "epoch": 0.4090338770388959,
      "grad_norm": 1.8254927396774292,
      "learning_rate": 4.835744319245333e-06,
      "loss": 1.3636,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 652
    },
    {
      "epoch": 0.4096612296110414,
      "grad_norm": 1.6572511196136475,
      "learning_rate": 4.83476734692151e-06,
      "loss": 1.1809,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 653
    },
    {
      "epoch": 0.410288582183187,
      "grad_norm": 1.943337082862854,
      "learning_rate": 4.833787577036981e-06,
      "loss": 1.2904,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 654
    },
    {
      "epoch": 0.4109159347553325,
      "grad_norm": 1.8604482412338257,
      "learning_rate": 4.832805010765724e-06,
      "loss": 1.3767,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 655
    },
    {
      "epoch": 0.411543287327478,
      "grad_norm": 1.9489023685455322,
      "learning_rate": 4.831819649285067e-06,
      "loss": 1.2878,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 656
    },
    {
      "epoch": 0.4121706398996236,
      "grad_norm": 1.7244187593460083,
      "learning_rate": 4.830831493775689e-06,
      "loss": 1.12,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 657
    },
    {
      "epoch": 0.4127979924717691,
      "grad_norm": 1.3886579275131226,
      "learning_rate": 4.829840545421615e-06,
      "loss": 1.2602,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 658
    },
    {
      "epoch": 0.4134253450439147,
      "grad_norm": 2.1388373374938965,
      "learning_rate": 4.828846805410219e-06,
      "loss": 1.4233,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 659
    },
    {
      "epoch": 0.41405269761606023,
      "grad_norm": 1.8400318622589111,
      "learning_rate": 4.827850274932216e-06,
      "loss": 1.2721,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 660
    },
    {
      "epoch": 0.41468005018820575,
      "grad_norm": 2.0351569652557373,
      "learning_rate": 4.826850955181669e-06,
      "loss": 1.2682,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 661
    },
    {
      "epoch": 0.41530740276035133,
      "grad_norm": 1.9729537963867188,
      "learning_rate": 4.8258488473559794e-06,
      "loss": 1.243,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 662
    },
    {
      "epoch": 0.41593475533249685,
      "grad_norm": 1.8061813116073608,
      "learning_rate": 4.8248439526558925e-06,
      "loss": 1.3179,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 663
    },
    {
      "epoch": 0.41656210790464243,
      "grad_norm": 1.7892892360687256,
      "learning_rate": 4.8238362722854905e-06,
      "loss": 1.3885,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 664
    },
    {
      "epoch": 0.41718946047678795,
      "grad_norm": 1.5382590293884277,
      "learning_rate": 4.822825807452195e-06,
      "loss": 1.2888,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 665
    },
    {
      "epoch": 0.4178168130489335,
      "grad_norm": 1.8710049390792847,
      "learning_rate": 4.821812559366763e-06,
      "loss": 1.3353,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 666
    },
    {
      "epoch": 0.41844416562107906,
      "grad_norm": 1.8414349555969238,
      "learning_rate": 4.820796529243287e-06,
      "loss": 1.3406,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 667
    },
    {
      "epoch": 0.4190715181932246,
      "grad_norm": 1.5640252828598022,
      "learning_rate": 4.819777718299192e-06,
      "loss": 1.2562,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 668
    },
    {
      "epoch": 0.41969887076537016,
      "grad_norm": 1.8525867462158203,
      "learning_rate": 4.8187561277552376e-06,
      "loss": 1.4028,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 669
    },
    {
      "epoch": 0.4203262233375157,
      "grad_norm": 1.936888337135315,
      "learning_rate": 4.81773175883551e-06,
      "loss": 1.141,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 670
    },
    {
      "epoch": 0.4209535759096612,
      "grad_norm": 1.4375437498092651,
      "learning_rate": 4.816704612767427e-06,
      "loss": 1.1689,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 671
    },
    {
      "epoch": 0.4215809284818068,
      "grad_norm": 1.9113013744354248,
      "learning_rate": 4.815674690781735e-06,
      "loss": 1.3321,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 672
    },
    {
      "epoch": 0.4222082810539523,
      "grad_norm": 1.830393671989441,
      "learning_rate": 4.814641994112505e-06,
      "loss": 1.4788,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 673
    },
    {
      "epoch": 0.4228356336260979,
      "grad_norm": 1.8348015546798706,
      "learning_rate": 4.813606523997132e-06,
      "loss": 1.3408,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 674
    },
    {
      "epoch": 0.4234629861982434,
      "grad_norm": 1.7932466268539429,
      "learning_rate": 4.812568281676335e-06,
      "loss": 1.2546,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 675
    },
    {
      "epoch": 0.424090338770389,
      "grad_norm": 1.4640141725540161,
      "learning_rate": 4.811527268394157e-06,
      "loss": 1.234,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 676
    },
    {
      "epoch": 0.4247176913425345,
      "grad_norm": 1.623489499092102,
      "learning_rate": 4.8104834853979565e-06,
      "loss": 1.3964,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 677
    },
    {
      "epoch": 0.42534504391468003,
      "grad_norm": 1.85269033908844,
      "learning_rate": 4.809436933938415e-06,
      "loss": 1.1884,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 678
    },
    {
      "epoch": 0.4259723964868256,
      "grad_norm": 1.5965299606323242,
      "learning_rate": 4.8083876152695285e-06,
      "loss": 1.2753,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 679
    },
    {
      "epoch": 0.42659974905897113,
      "grad_norm": 1.5630314350128174,
      "learning_rate": 4.80733553064861e-06,
      "loss": 1.3914,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 680
    },
    {
      "epoch": 0.4272271016311167,
      "grad_norm": 1.7506418228149414,
      "learning_rate": 4.806280681336286e-06,
      "loss": 1.4983,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 681
    },
    {
      "epoch": 0.42785445420326224,
      "grad_norm": 1.6316046714782715,
      "learning_rate": 4.805223068596497e-06,
      "loss": 1.3471,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 682
    },
    {
      "epoch": 0.42848180677540776,
      "grad_norm": 1.502759337425232,
      "learning_rate": 4.804162693696494e-06,
      "loss": 1.3321,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 683
    },
    {
      "epoch": 0.42910915934755334,
      "grad_norm": 1.963027000427246,
      "learning_rate": 4.803099557906836e-06,
      "loss": 1.2162,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 684
    },
    {
      "epoch": 0.42973651191969886,
      "grad_norm": 1.7412725687026978,
      "learning_rate": 4.802033662501392e-06,
      "loss": 1.2524,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 685
    },
    {
      "epoch": 0.43036386449184444,
      "grad_norm": 2.261122226715088,
      "learning_rate": 4.800965008757339e-06,
      "loss": 1.3432,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 686
    },
    {
      "epoch": 0.43099121706398996,
      "grad_norm": 1.758253812789917,
      "learning_rate": 4.799893597955157e-06,
      "loss": 1.2964,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 687
    },
    {
      "epoch": 0.4316185696361355,
      "grad_norm": 1.7574810981750488,
      "learning_rate": 4.7988194313786275e-06,
      "loss": 1.3334,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 688
    },
    {
      "epoch": 0.43224592220828106,
      "grad_norm": 1.8565386533737183,
      "learning_rate": 4.797742510314838e-06,
      "loss": 1.3402,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 689
    },
    {
      "epoch": 0.4328732747804266,
      "grad_norm": 1.7075424194335938,
      "learning_rate": 4.796662836054176e-06,
      "loss": 1.2549,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 690
    },
    {
      "epoch": 0.43350062735257217,
      "grad_norm": 1.8174091577529907,
      "learning_rate": 4.795580409890325e-06,
      "loss": 1.21,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 691
    },
    {
      "epoch": 0.4341279799247177,
      "grad_norm": 2.091965675354004,
      "learning_rate": 4.794495233120268e-06,
      "loss": 1.3602,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 692
    },
    {
      "epoch": 0.4347553324968632,
      "grad_norm": 1.618514060974121,
      "learning_rate": 4.793407307044285e-06,
      "loss": 1.3823,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 693
    },
    {
      "epoch": 0.4353826850690088,
      "grad_norm": 1.4373396635055542,
      "learning_rate": 4.792316632965947e-06,
      "loss": 1.1906,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 694
    },
    {
      "epoch": 0.4360100376411543,
      "grad_norm": 1.89703369140625,
      "learning_rate": 4.791223212192121e-06,
      "loss": 1.2905,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 695
    },
    {
      "epoch": 0.4366373902132999,
      "grad_norm": 1.8535451889038086,
      "learning_rate": 4.790127046032963e-06,
      "loss": 1.2045,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 696
    },
    {
      "epoch": 0.4372647427854454,
      "grad_norm": 1.8034179210662842,
      "learning_rate": 4.789028135801919e-06,
      "loss": 1.2953,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 697
    },
    {
      "epoch": 0.437892095357591,
      "grad_norm": 1.4916350841522217,
      "learning_rate": 4.787926482815724e-06,
      "loss": 1.2158,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 698
    },
    {
      "epoch": 0.4385194479297365,
      "grad_norm": 1.2732455730438232,
      "learning_rate": 4.786822088394397e-06,
      "loss": 1.405,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 699
    },
    {
      "epoch": 0.43914680050188204,
      "grad_norm": 1.8891061544418335,
      "learning_rate": 4.7857149538612454e-06,
      "loss": 1.2653,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 700
    },
    {
      "epoch": 0.4397741530740276,
      "grad_norm": 2.2417843341827393,
      "learning_rate": 4.784605080542857e-06,
      "loss": 1.3876,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 701
    },
    {
      "epoch": 0.44040150564617314,
      "grad_norm": 1.9334285259246826,
      "learning_rate": 4.783492469769103e-06,
      "loss": 1.3257,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 702
    },
    {
      "epoch": 0.4410288582183187,
      "grad_norm": 1.7055387496948242,
      "learning_rate": 4.7823771228731325e-06,
      "loss": 1.2261,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 703
    },
    {
      "epoch": 0.44165621079046424,
      "grad_norm": 1.9559547901153564,
      "learning_rate": 4.7812590411913755e-06,
      "loss": 1.2834,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 704
    },
    {
      "epoch": 0.44228356336260977,
      "grad_norm": 1.906968116760254,
      "learning_rate": 4.780138226063537e-06,
      "loss": 1.3762,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 705
    },
    {
      "epoch": 0.44291091593475534,
      "grad_norm": 2.4029104709625244,
      "learning_rate": 4.779014678832599e-06,
      "loss": 1.3537,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 706
    },
    {
      "epoch": 0.44353826850690087,
      "grad_norm": 1.7167798280715942,
      "learning_rate": 4.777888400844815e-06,
      "loss": 1.219,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 707
    },
    {
      "epoch": 0.44416562107904645,
      "grad_norm": 1.8012571334838867,
      "learning_rate": 4.7767593934497145e-06,
      "loss": 1.3654,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 708
    },
    {
      "epoch": 0.44479297365119197,
      "grad_norm": 2.2008590698242188,
      "learning_rate": 4.775627658000091e-06,
      "loss": 1.2487,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 709
    },
    {
      "epoch": 0.4454203262233375,
      "grad_norm": 1.9626473188400269,
      "learning_rate": 4.774493195852012e-06,
      "loss": 1.2511,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 710
    },
    {
      "epoch": 0.44604767879548307,
      "grad_norm": 2.1587445735931396,
      "learning_rate": 4.773356008364812e-06,
      "loss": 1.2596,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 711
    },
    {
      "epoch": 0.4466750313676286,
      "grad_norm": 1.9012348651885986,
      "learning_rate": 4.772216096901089e-06,
      "loss": 1.315,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 712
    },
    {
      "epoch": 0.4473023839397742,
      "grad_norm": 1.6769298315048218,
      "learning_rate": 4.771073462826704e-06,
      "loss": 1.1517,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 713
    },
    {
      "epoch": 0.4479297365119197,
      "grad_norm": 1.7850003242492676,
      "learning_rate": 4.769928107510784e-06,
      "loss": 1.319,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 714
    },
    {
      "epoch": 0.4485570890840652,
      "grad_norm": 1.8586574792861938,
      "learning_rate": 4.768780032325714e-06,
      "loss": 1.3514,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 715
    },
    {
      "epoch": 0.4491844416562108,
      "grad_norm": 1.946988821029663,
      "learning_rate": 4.767629238647138e-06,
      "loss": 1.2864,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 716
    },
    {
      "epoch": 0.4498117942283563,
      "grad_norm": 1.9582651853561401,
      "learning_rate": 4.766475727853959e-06,
      "loss": 1.2603,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 717
    },
    {
      "epoch": 0.4504391468005019,
      "grad_norm": 1.9077582359313965,
      "learning_rate": 4.765319501328332e-06,
      "loss": 1.3582,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 718
    },
    {
      "epoch": 0.4510664993726474,
      "grad_norm": 1.8661820888519287,
      "learning_rate": 4.7641605604556725e-06,
      "loss": 1.2798,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 719
    },
    {
      "epoch": 0.451693851944793,
      "grad_norm": 1.7664010524749756,
      "learning_rate": 4.762998906624643e-06,
      "loss": 1.2414,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 720
    },
    {
      "epoch": 0.4523212045169385,
      "grad_norm": 1.740552544593811,
      "learning_rate": 4.761834541227157e-06,
      "loss": 1.3123,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 721
    },
    {
      "epoch": 0.45294855708908405,
      "grad_norm": 2.0045013427734375,
      "learning_rate": 4.760667465658378e-06,
      "loss": 1.3067,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 722
    },
    {
      "epoch": 0.4535759096612296,
      "grad_norm": 1.4887140989303589,
      "learning_rate": 4.75949768131672e-06,
      "loss": 1.2057,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 723
    },
    {
      "epoch": 0.45420326223337515,
      "grad_norm": 4.028663635253906,
      "learning_rate": 4.758325189603838e-06,
      "loss": 1.2178,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 724
    },
    {
      "epoch": 0.4548306148055207,
      "grad_norm": 1.8994674682617188,
      "learning_rate": 4.757149991924633e-06,
      "loss": 1.2722,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 725
    },
    {
      "epoch": 0.45545796737766625,
      "grad_norm": 1.8742094039916992,
      "learning_rate": 4.755972089687248e-06,
      "loss": 1.2098,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 726
    },
    {
      "epoch": 0.4560853199498118,
      "grad_norm": 1.8946446180343628,
      "learning_rate": 4.754791484303068e-06,
      "loss": 1.2502,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 727
    },
    {
      "epoch": 0.45671267252195735,
      "grad_norm": 1.503798246383667,
      "learning_rate": 4.753608177186714e-06,
      "loss": 1.2383,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 728
    },
    {
      "epoch": 0.4573400250941029,
      "grad_norm": 1.4462285041809082,
      "learning_rate": 4.752422169756048e-06,
      "loss": 1.2647,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 729
    },
    {
      "epoch": 0.45796737766624845,
      "grad_norm": 1.9991223812103271,
      "learning_rate": 4.751233463432165e-06,
      "loss": 1.2988,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 730
    },
    {
      "epoch": 0.458594730238394,
      "grad_norm": 1.5803600549697876,
      "learning_rate": 4.7500420596393935e-06,
      "loss": 1.2609,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 731
    },
    {
      "epoch": 0.4592220828105395,
      "grad_norm": 1.5404584407806396,
      "learning_rate": 4.748847959805297e-06,
      "loss": 1.2166,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 732
    },
    {
      "epoch": 0.4598494353826851,
      "grad_norm": 1.8132827281951904,
      "learning_rate": 4.7476511653606676e-06,
      "loss": 1.3939,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 733
    },
    {
      "epoch": 0.4604767879548306,
      "grad_norm": 1.800467610359192,
      "learning_rate": 4.7464516777395234e-06,
      "loss": 1.2325,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 734
    },
    {
      "epoch": 0.4611041405269762,
      "grad_norm": 1.9034929275512695,
      "learning_rate": 4.7452494983791155e-06,
      "loss": 1.2562,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 735
    },
    {
      "epoch": 0.4617314930991217,
      "grad_norm": 1.7843937873840332,
      "learning_rate": 4.744044628719916e-06,
      "loss": 1.3663,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 736
    },
    {
      "epoch": 0.4623588456712672,
      "grad_norm": 2.562595844268799,
      "learning_rate": 4.74283707020562e-06,
      "loss": 1.2311,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 737
    },
    {
      "epoch": 0.4629861982434128,
      "grad_norm": 1.9093017578125,
      "learning_rate": 4.741626824283149e-06,
      "loss": 1.3225,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 738
    },
    {
      "epoch": 0.46361355081555833,
      "grad_norm": 1.5254753828048706,
      "learning_rate": 4.740413892402639e-06,
      "loss": 1.161,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 739
    },
    {
      "epoch": 0.4642409033877039,
      "grad_norm": 1.795628547668457,
      "learning_rate": 4.739198276017449e-06,
      "loss": 1.2754,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 740
    },
    {
      "epoch": 0.46486825595984943,
      "grad_norm": 1.8925135135650635,
      "learning_rate": 4.7379799765841525e-06,
      "loss": 1.208,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 741
    },
    {
      "epoch": 0.465495608531995,
      "grad_norm": 1.8105072975158691,
      "learning_rate": 4.736758995562536e-06,
      "loss": 1.3267,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 742
    },
    {
      "epoch": 0.46612296110414053,
      "grad_norm": 1.910277009010315,
      "learning_rate": 4.735535334415604e-06,
      "loss": 1.1361,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 743
    },
    {
      "epoch": 0.46675031367628605,
      "grad_norm": 1.7848541736602783,
      "learning_rate": 4.734308994609568e-06,
      "loss": 1.405,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 744
    },
    {
      "epoch": 0.46737766624843163,
      "grad_norm": 1.724311351776123,
      "learning_rate": 4.73307997761385e-06,
      "loss": 1.2611,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 745
    },
    {
      "epoch": 0.46800501882057716,
      "grad_norm": 1.8376044034957886,
      "learning_rate": 4.731848284901082e-06,
      "loss": 1.4162,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 746
    },
    {
      "epoch": 0.46863237139272274,
      "grad_norm": 1.8465354442596436,
      "learning_rate": 4.7306139179471e-06,
      "loss": 1.3957,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 747
    },
    {
      "epoch": 0.46925972396486826,
      "grad_norm": 1.9396543502807617,
      "learning_rate": 4.7293768782309446e-06,
      "loss": 1.2662,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 748
    },
    {
      "epoch": 0.4698870765370138,
      "grad_norm": 1.952086091041565,
      "learning_rate": 4.72813716723486e-06,
      "loss": 1.2586,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 749
    },
    {
      "epoch": 0.47051442910915936,
      "grad_norm": 1.8261383771896362,
      "learning_rate": 4.726894786444289e-06,
      "loss": 1.2385,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 750
    },
    {
      "epoch": 0.4711417816813049,
      "grad_norm": 1.561482548713684,
      "learning_rate": 4.725649737347877e-06,
      "loss": 1.1562,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 751
    },
    {
      "epoch": 0.47176913425345046,
      "grad_norm": 1.9331833124160767,
      "learning_rate": 4.724402021437463e-06,
      "loss": 1.1425,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 752
    },
    {
      "epoch": 0.472396486825596,
      "grad_norm": 1.876682162284851,
      "learning_rate": 4.723151640208084e-06,
      "loss": 1.327,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 753
    },
    {
      "epoch": 0.4730238393977415,
      "grad_norm": 1.957084059715271,
      "learning_rate": 4.721898595157969e-06,
      "loss": 1.3078,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 754
    },
    {
      "epoch": 0.4736511919698871,
      "grad_norm": 1.9063243865966797,
      "learning_rate": 4.720642887788538e-06,
      "loss": 1.4485,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 755
    },
    {
      "epoch": 0.4742785445420326,
      "grad_norm": 1.6841647624969482,
      "learning_rate": 4.719384519604404e-06,
      "loss": 1.2789,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 756
    },
    {
      "epoch": 0.4749058971141782,
      "grad_norm": 1.864952802658081,
      "learning_rate": 4.718123492113367e-06,
      "loss": 1.1969,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 757
    },
    {
      "epoch": 0.4755332496863237,
      "grad_norm": 1.3406996726989746,
      "learning_rate": 4.716859806826412e-06,
      "loss": 1.2043,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 758
    },
    {
      "epoch": 0.47616060225846923,
      "grad_norm": 2.2283079624176025,
      "learning_rate": 4.7155934652577095e-06,
      "loss": 1.3071,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 759
    },
    {
      "epoch": 0.4767879548306148,
      "grad_norm": 1.748877763748169,
      "learning_rate": 4.714324468924614e-06,
      "loss": 1.3739,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 760
    },
    {
      "epoch": 0.47741530740276034,
      "grad_norm": 1.9405159950256348,
      "learning_rate": 4.71305281934766e-06,
      "loss": 1.1123,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 761
    },
    {
      "epoch": 0.4780426599749059,
      "grad_norm": 1.994694709777832,
      "learning_rate": 4.71177851805056e-06,
      "loss": 1.2621,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 762
    },
    {
      "epoch": 0.47867001254705144,
      "grad_norm": 1.9419386386871338,
      "learning_rate": 4.7105015665602064e-06,
      "loss": 1.2656,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 763
    },
    {
      "epoch": 0.479297365119197,
      "grad_norm": 1.8195141553878784,
      "learning_rate": 4.709221966406664e-06,
      "loss": 1.1962,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 764
    },
    {
      "epoch": 0.47992471769134254,
      "grad_norm": 1.7010425329208374,
      "learning_rate": 4.707939719123175e-06,
      "loss": 1.3749,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 765
    },
    {
      "epoch": 0.48055207026348806,
      "grad_norm": 1.808499813079834,
      "learning_rate": 4.70665482624615e-06,
      "loss": 1.1261,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 766
    },
    {
      "epoch": 0.48117942283563364,
      "grad_norm": 1.8877878189086914,
      "learning_rate": 4.705367289315172e-06,
      "loss": 1.358,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 767
    },
    {
      "epoch": 0.48180677540777916,
      "grad_norm": 2.219843626022339,
      "learning_rate": 4.704077109872989e-06,
      "loss": 1.3687,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 768
    },
    {
      "epoch": 0.48243412797992474,
      "grad_norm": 1.7579948902130127,
      "learning_rate": 4.702784289465521e-06,
      "loss": 1.2922,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 769
    },
    {
      "epoch": 0.48306148055207027,
      "grad_norm": 1.767750859260559,
      "learning_rate": 4.701488829641845e-06,
      "loss": 1.2044,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 770
    },
    {
      "epoch": 0.4836888331242158,
      "grad_norm": 1.6337237358093262,
      "learning_rate": 4.700190731954209e-06,
      "loss": 1.4174,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 771
    },
    {
      "epoch": 0.48431618569636137,
      "grad_norm": 1.8537286520004272,
      "learning_rate": 4.698889997958013e-06,
      "loss": 1.4132,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 772
    },
    {
      "epoch": 0.4849435382685069,
      "grad_norm": 1.842587947845459,
      "learning_rate": 4.697586629211822e-06,
      "loss": 1.2961,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 773
    },
    {
      "epoch": 0.48557089084065247,
      "grad_norm": 1.8020635843276978,
      "learning_rate": 4.696280627277356e-06,
      "loss": 1.2704,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 774
    },
    {
      "epoch": 0.486198243412798,
      "grad_norm": 2.015141248703003,
      "learning_rate": 4.694971993719491e-06,
      "loss": 1.276,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 775
    },
    {
      "epoch": 0.4868255959849435,
      "grad_norm": 1.6180552244186401,
      "learning_rate": 4.6936607301062545e-06,
      "loss": 1.2509,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 776
    },
    {
      "epoch": 0.4874529485570891,
      "grad_norm": 2.057633876800537,
      "learning_rate": 4.692346838008828e-06,
      "loss": 1.2712,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 777
    },
    {
      "epoch": 0.4880803011292346,
      "grad_norm": 1.8549116849899292,
      "learning_rate": 4.691030319001538e-06,
      "loss": 1.2617,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 778
    },
    {
      "epoch": 0.4887076537013802,
      "grad_norm": 1.8069357872009277,
      "learning_rate": 4.689711174661864e-06,
      "loss": 1.1867,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 779
    },
    {
      "epoch": 0.4893350062735257,
      "grad_norm": 2.0871284008026123,
      "learning_rate": 4.688389406570427e-06,
      "loss": 1.2402,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 780
    },
    {
      "epoch": 0.48996235884567124,
      "grad_norm": 1.4803322553634644,
      "learning_rate": 4.687065016310996e-06,
      "loss": 1.2458,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 781
    },
    {
      "epoch": 0.4905897114178168,
      "grad_norm": 1.9475903511047363,
      "learning_rate": 4.685738005470478e-06,
      "loss": 1.3481,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 782
    },
    {
      "epoch": 0.49121706398996234,
      "grad_norm": 1.8729878664016724,
      "learning_rate": 4.684408375638921e-06,
      "loss": 1.3966,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 783
    },
    {
      "epoch": 0.4918444165621079,
      "grad_norm": 1.7364298105239868,
      "learning_rate": 4.683076128409512e-06,
      "loss": 1.2473,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 784
    },
    {
      "epoch": 0.49247176913425345,
      "grad_norm": 1.6857995986938477,
      "learning_rate": 4.681741265378574e-06,
      "loss": 1.3229,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 785
    },
    {
      "epoch": 0.493099121706399,
      "grad_norm": 2.2026307582855225,
      "learning_rate": 4.680403788145566e-06,
      "loss": 1.298,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 786
    },
    {
      "epoch": 0.49372647427854455,
      "grad_norm": 2.3991730213165283,
      "learning_rate": 4.679063698313074e-06,
      "loss": 1.2762,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 787
    },
    {
      "epoch": 0.49435382685069007,
      "grad_norm": 1.9600284099578857,
      "learning_rate": 4.6777209974868194e-06,
      "loss": 1.3338,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 788
    },
    {
      "epoch": 0.49498117942283565,
      "grad_norm": 1.7276560068130493,
      "learning_rate": 4.676375687275653e-06,
      "loss": 1.2017,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 789
    },
    {
      "epoch": 0.49560853199498117,
      "grad_norm": 2.1582796573638916,
      "learning_rate": 4.675027769291547e-06,
      "loss": 1.2509,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 790
    },
    {
      "epoch": 0.49623588456712675,
      "grad_norm": 1.7889536619186401,
      "learning_rate": 4.673677245149602e-06,
      "loss": 1.3825,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 791
    },
    {
      "epoch": 0.4968632371392723,
      "grad_norm": 1.567894697189331,
      "learning_rate": 4.672324116468042e-06,
      "loss": 1.2841,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 792
    },
    {
      "epoch": 0.4974905897114178,
      "grad_norm": 1.4516322612762451,
      "learning_rate": 4.67096838486821e-06,
      "loss": 1.2557,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 793
    },
    {
      "epoch": 0.4981179422835634,
      "grad_norm": 1.992506980895996,
      "learning_rate": 4.669610051974566e-06,
      "loss": 1.2688,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 794
    },
    {
      "epoch": 0.4987452948557089,
      "grad_norm": 1.7153563499450684,
      "learning_rate": 4.668249119414692e-06,
      "loss": 1.4142,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 795
    },
    {
      "epoch": 0.4993726474278545,
      "grad_norm": 1.7995789051055908,
      "learning_rate": 4.66688558881928e-06,
      "loss": 1.2004,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 796
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.726151943206787,
      "learning_rate": 4.6655194618221385e-06,
      "loss": 1.2882,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 797
    },
    {
      "epoch": 0.5006273525721455,
      "grad_norm": 1.8986132144927979,
      "learning_rate": 4.664150740060186e-06,
      "loss": 1.4326,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 798
    },
    {
      "epoch": 0.501254705144291,
      "grad_norm": 1.6591218709945679,
      "learning_rate": 4.6627794251734485e-06,
      "loss": 1.251,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 799
    },
    {
      "epoch": 0.5018820577164367,
      "grad_norm": 1.7284833192825317,
      "learning_rate": 4.6614055188050615e-06,
      "loss": 1.3658,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 800
    },
    {
      "epoch": 0.5025094102885822,
      "grad_norm": 1.6566392183303833,
      "learning_rate": 4.660029022601264e-06,
      "loss": 1.2381,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 801
    },
    {
      "epoch": 0.5031367628607277,
      "grad_norm": 2.047153949737549,
      "learning_rate": 4.6586499382113985e-06,
      "loss": 1.5228,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 802
    },
    {
      "epoch": 0.5037641154328732,
      "grad_norm": 1.376027226448059,
      "learning_rate": 4.6572682672879095e-06,
      "loss": 1.1302,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 803
    },
    {
      "epoch": 0.5043914680050188,
      "grad_norm": 1.8712245225906372,
      "learning_rate": 4.655884011486341e-06,
      "loss": 1.3529,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 804
    },
    {
      "epoch": 0.5050188205771644,
      "grad_norm": 1.777405858039856,
      "learning_rate": 4.654497172465333e-06,
      "loss": 1.2575,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 805
    },
    {
      "epoch": 0.5056461731493099,
      "grad_norm": 1.7459986209869385,
      "learning_rate": 4.653107751886621e-06,
      "loss": 1.1989,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 806
    },
    {
      "epoch": 0.5062735257214555,
      "grad_norm": 1.6451423168182373,
      "learning_rate": 4.651715751415033e-06,
      "loss": 1.2338,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 807
    },
    {
      "epoch": 0.506900878293601,
      "grad_norm": 1.2586438655853271,
      "learning_rate": 4.65032117271849e-06,
      "loss": 1.2002,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 808
    },
    {
      "epoch": 0.5075282308657465,
      "grad_norm": 1.8858215808868408,
      "learning_rate": 4.648924017468003e-06,
      "loss": 1.1834,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 809
    },
    {
      "epoch": 0.5081555834378921,
      "grad_norm": 2.0493290424346924,
      "learning_rate": 4.647524287337668e-06,
      "loss": 1.3408,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 810
    },
    {
      "epoch": 0.5087829360100377,
      "grad_norm": 1.9480719566345215,
      "learning_rate": 4.646121984004666e-06,
      "loss": 1.2574,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 811
    },
    {
      "epoch": 0.5094102885821832,
      "grad_norm": 1.9408248662948608,
      "learning_rate": 4.644717109149262e-06,
      "loss": 1.3718,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 812
    },
    {
      "epoch": 0.5100376411543287,
      "grad_norm": 1.7562613487243652,
      "learning_rate": 4.643309664454806e-06,
      "loss": 1.3656,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 813
    },
    {
      "epoch": 0.5106649937264742,
      "grad_norm": 1.7359344959259033,
      "learning_rate": 4.641899651607721e-06,
      "loss": 1.3346,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 814
    },
    {
      "epoch": 0.5112923462986199,
      "grad_norm": 1.976855993270874,
      "learning_rate": 4.640487072297509e-06,
      "loss": 1.2468,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 815
    },
    {
      "epoch": 0.5119196988707654,
      "grad_norm": 1.5342873334884644,
      "learning_rate": 4.6390719282167515e-06,
      "loss": 1.1683,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 816
    },
    {
      "epoch": 0.5125470514429109,
      "grad_norm": 1.69839346408844,
      "learning_rate": 4.637654221061098e-06,
      "loss": 1.2034,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 817
    },
    {
      "epoch": 0.5131744040150564,
      "grad_norm": 2.431257486343384,
      "learning_rate": 4.636233952529271e-06,
      "loss": 1.3565,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 818
    },
    {
      "epoch": 0.513801756587202,
      "grad_norm": 1.39933180809021,
      "learning_rate": 4.634811124323062e-06,
      "loss": 1.3246,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 819
    },
    {
      "epoch": 0.5144291091593476,
      "grad_norm": 2.4654202461242676,
      "learning_rate": 4.6333857381473305e-06,
      "loss": 1.3884,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 820
    },
    {
      "epoch": 0.5150564617314931,
      "grad_norm": 1.9540578126907349,
      "learning_rate": 4.631957795709999e-06,
      "loss": 1.3104,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 821
    },
    {
      "epoch": 0.5156838143036386,
      "grad_norm": 1.9116793870925903,
      "learning_rate": 4.630527298722056e-06,
      "loss": 1.2827,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 822
    },
    {
      "epoch": 0.5163111668757842,
      "grad_norm": 2.0660154819488525,
      "learning_rate": 4.629094248897546e-06,
      "loss": 1.3751,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 823
    },
    {
      "epoch": 0.5169385194479298,
      "grad_norm": 1.8745118379592896,
      "learning_rate": 4.627658647953579e-06,
      "loss": 1.3202,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 824
    },
    {
      "epoch": 0.5175658720200753,
      "grad_norm": 1.8826225996017456,
      "learning_rate": 4.626220497610315e-06,
      "loss": 1.2711,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 825
    },
    {
      "epoch": 0.5181932245922208,
      "grad_norm": 2.0264394283294678,
      "learning_rate": 4.6247797995909735e-06,
      "loss": 1.2534,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 826
    },
    {
      "epoch": 0.5188205771643664,
      "grad_norm": 2.2216804027557373,
      "learning_rate": 4.623336555621825e-06,
      "loss": 1.3163,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 827
    },
    {
      "epoch": 0.5194479297365119,
      "grad_norm": 1.9490082263946533,
      "learning_rate": 4.621890767432189e-06,
      "loss": 1.3073,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 828
    },
    {
      "epoch": 0.5200752823086575,
      "grad_norm": 2.479620933532715,
      "learning_rate": 4.620442436754438e-06,
      "loss": 1.3627,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 829
    },
    {
      "epoch": 0.520702634880803,
      "grad_norm": 1.924384593963623,
      "learning_rate": 4.618991565323987e-06,
      "loss": 1.3137,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 830
    },
    {
      "epoch": 0.5213299874529486,
      "grad_norm": 1.8495651483535767,
      "learning_rate": 4.617538154879295e-06,
      "loss": 1.2852,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 831
    },
    {
      "epoch": 0.5219573400250941,
      "grad_norm": 1.8771755695343018,
      "learning_rate": 4.616082207161867e-06,
      "loss": 1.242,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 832
    },
    {
      "epoch": 0.5225846925972396,
      "grad_norm": 1.8403741121292114,
      "learning_rate": 4.614623723916245e-06,
      "loss": 1.2524,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 833
    },
    {
      "epoch": 0.5232120451693852,
      "grad_norm": 1.9191933870315552,
      "learning_rate": 4.613162706890011e-06,
      "loss": 1.4476,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 834
    },
    {
      "epoch": 0.5238393977415308,
      "grad_norm": 2.0666935443878174,
      "learning_rate": 4.611699157833781e-06,
      "loss": 1.3978,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 835
    },
    {
      "epoch": 0.5244667503136763,
      "grad_norm": 1.9643383026123047,
      "learning_rate": 4.610233078501207e-06,
      "loss": 1.1818,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 836
    },
    {
      "epoch": 0.5250941028858218,
      "grad_norm": 2.003880262374878,
      "learning_rate": 4.608764470648971e-06,
      "loss": 1.2073,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 837
    },
    {
      "epoch": 0.5257214554579673,
      "grad_norm": 1.6099451780319214,
      "learning_rate": 4.607293336036787e-06,
      "loss": 1.2718,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 838
    },
    {
      "epoch": 0.526348808030113,
      "grad_norm": 1.862188458442688,
      "learning_rate": 4.605819676427393e-06,
      "loss": 1.2728,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 839
    },
    {
      "epoch": 0.5269761606022585,
      "grad_norm": 2.5151021480560303,
      "learning_rate": 4.604343493586558e-06,
      "loss": 1.2652,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 840
    },
    {
      "epoch": 0.527603513174404,
      "grad_norm": 1.627556562423706,
      "learning_rate": 4.602864789283068e-06,
      "loss": 1.2863,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 841
    },
    {
      "epoch": 0.5282308657465495,
      "grad_norm": 1.7955687046051025,
      "learning_rate": 4.601383565288735e-06,
      "loss": 1.298,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 842
    },
    {
      "epoch": 0.5288582183186951,
      "grad_norm": 1.9044831991195679,
      "learning_rate": 4.599899823378389e-06,
      "loss": 1.2663,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 843
    },
    {
      "epoch": 0.5294855708908407,
      "grad_norm": 1.531988501548767,
      "learning_rate": 4.598413565329876e-06,
      "loss": 1.2011,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 844
    },
    {
      "epoch": 0.5301129234629862,
      "grad_norm": 2.0963053703308105,
      "learning_rate": 4.5969247929240555e-06,
      "loss": 1.4239,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 845
    },
    {
      "epoch": 0.5307402760351317,
      "grad_norm": 1.9998327493667603,
      "learning_rate": 4.595433507944805e-06,
      "loss": 1.3493,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 846
    },
    {
      "epoch": 0.5313676286072773,
      "grad_norm": 1.8099595308303833,
      "learning_rate": 4.593939712179007e-06,
      "loss": 1.3457,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 847
    },
    {
      "epoch": 0.5319949811794228,
      "grad_norm": 1.8672784566879272,
      "learning_rate": 4.5924434074165554e-06,
      "loss": 1.2896,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 848
    },
    {
      "epoch": 0.5326223337515684,
      "grad_norm": 1.7104904651641846,
      "learning_rate": 4.590944595450351e-06,
      "loss": 1.1611,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 849
    },
    {
      "epoch": 0.533249686323714,
      "grad_norm": 2.010240077972412,
      "learning_rate": 4.589443278076296e-06,
      "loss": 1.1827,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 850
    },
    {
      "epoch": 0.5338770388958595,
      "grad_norm": 2.0523414611816406,
      "learning_rate": 4.587939457093296e-06,
      "loss": 1.2599,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 851
    },
    {
      "epoch": 0.534504391468005,
      "grad_norm": 1.7473045587539673,
      "learning_rate": 4.586433134303257e-06,
      "loss": 1.1739,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 852
    },
    {
      "epoch": 0.5351317440401505,
      "grad_norm": 2.0589144229888916,
      "learning_rate": 4.584924311511083e-06,
      "loss": 1.304,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 853
    },
    {
      "epoch": 0.5357590966122961,
      "grad_norm": 1.6641743183135986,
      "learning_rate": 4.5834129905246725e-06,
      "loss": 1.3376,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 854
    },
    {
      "epoch": 0.5363864491844417,
      "grad_norm": 1.9077972173690796,
      "learning_rate": 4.5818991731549185e-06,
      "loss": 1.332,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 855
    },
    {
      "epoch": 0.5370138017565872,
      "grad_norm": 1.9205858707427979,
      "learning_rate": 4.5803828612157035e-06,
      "loss": 1.1763,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 856
    },
    {
      "epoch": 0.5376411543287327,
      "grad_norm": 1.5044771432876587,
      "learning_rate": 4.5788640565239e-06,
      "loss": 1.1735,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 857
    },
    {
      "epoch": 0.5382685069008782,
      "grad_norm": 1.7473088502883911,
      "learning_rate": 4.577342760899368e-06,
      "loss": 1.4267,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 858
    },
    {
      "epoch": 0.5388958594730239,
      "grad_norm": 1.798748254776001,
      "learning_rate": 4.575818976164952e-06,
      "loss": 1.3393,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 859
    },
    {
      "epoch": 0.5395232120451694,
      "grad_norm": 1.7975914478302002,
      "learning_rate": 4.574292704146478e-06,
      "loss": 1.1311,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 860
    },
    {
      "epoch": 0.5401505646173149,
      "grad_norm": 1.6357945203781128,
      "learning_rate": 4.5727639466727525e-06,
      "loss": 1.3367,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 861
    },
    {
      "epoch": 0.5407779171894604,
      "grad_norm": 1.763322353363037,
      "learning_rate": 4.5712327055755605e-06,
      "loss": 1.2515,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 862
    },
    {
      "epoch": 0.541405269761606,
      "grad_norm": 1.855243444442749,
      "learning_rate": 4.569698982689664e-06,
      "loss": 1.1587,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 863
    },
    {
      "epoch": 0.5420326223337516,
      "grad_norm": 1.829127550125122,
      "learning_rate": 4.5681627798527965e-06,
      "loss": 1.4049,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 864
    },
    {
      "epoch": 0.5426599749058971,
      "grad_norm": 1.9672813415527344,
      "learning_rate": 4.566624098905665e-06,
      "loss": 1.4499,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 865
    },
    {
      "epoch": 0.5432873274780426,
      "grad_norm": 1.776489496231079,
      "learning_rate": 4.565082941691944e-06,
      "loss": 1.2612,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 866
    },
    {
      "epoch": 0.5439146800501882,
      "grad_norm": 1.8452541828155518,
      "learning_rate": 4.5635393100582785e-06,
      "loss": 1.3581,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 867
    },
    {
      "epoch": 0.5445420326223338,
      "grad_norm": 2.2932517528533936,
      "learning_rate": 4.561993205854273e-06,
      "loss": 1.317,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 868
    },
    {
      "epoch": 0.5451693851944793,
      "grad_norm": 1.9137377738952637,
      "learning_rate": 4.560444630932499e-06,
      "loss": 1.2381,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 869
    },
    {
      "epoch": 0.5457967377666249,
      "grad_norm": 1.4997367858886719,
      "learning_rate": 4.558893587148488e-06,
      "loss": 1.2208,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 870
    },
    {
      "epoch": 0.5464240903387704,
      "grad_norm": 1.9404653310775757,
      "learning_rate": 4.5573400763607264e-06,
      "loss": 1.4,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 871
    },
    {
      "epoch": 0.5470514429109159,
      "grad_norm": 1.885493278503418,
      "learning_rate": 4.555784100430662e-06,
      "loss": 1.1995,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 872
    },
    {
      "epoch": 0.5476787954830615,
      "grad_norm": 1.9907666444778442,
      "learning_rate": 4.554225661222692e-06,
      "loss": 1.2612,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 873
    },
    {
      "epoch": 0.548306148055207,
      "grad_norm": 1.9186691045761108,
      "learning_rate": 4.552664760604167e-06,
      "loss": 1.3292,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 874
    },
    {
      "epoch": 0.5489335006273526,
      "grad_norm": 1.8157944679260254,
      "learning_rate": 4.5511014004453855e-06,
      "loss": 1.3547,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 875
    },
    {
      "epoch": 0.5495608531994981,
      "grad_norm": 1.961639165878296,
      "learning_rate": 4.549535582619595e-06,
      "loss": 1.2362,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 876
    },
    {
      "epoch": 0.5501882057716436,
      "grad_norm": 1.9759812355041504,
      "learning_rate": 4.547967309002988e-06,
      "loss": 1.2151,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 877
    },
    {
      "epoch": 0.5508155583437893,
      "grad_norm": 1.2574729919433594,
      "learning_rate": 4.5463965814746955e-06,
      "loss": 1.1432,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 878
    },
    {
      "epoch": 0.5514429109159348,
      "grad_norm": 2.019104242324829,
      "learning_rate": 4.544823401916794e-06,
      "loss": 1.2727,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 879
    },
    {
      "epoch": 0.5520702634880803,
      "grad_norm": 1.6454278230667114,
      "learning_rate": 4.543247772214297e-06,
      "loss": 1.2706,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 880
    },
    {
      "epoch": 0.5526976160602258,
      "grad_norm": 1.697099208831787,
      "learning_rate": 4.541669694255149e-06,
      "loss": 1.2914,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 881
    },
    {
      "epoch": 0.5533249686323714,
      "grad_norm": 1.8443635702133179,
      "learning_rate": 4.540089169930234e-06,
      "loss": 1.3881,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 882
    },
    {
      "epoch": 0.553952321204517,
      "grad_norm": 2.3314108848571777,
      "learning_rate": 4.538506201133366e-06,
      "loss": 1.5001,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 883
    },
    {
      "epoch": 0.5545796737766625,
      "grad_norm": 1.8084019422531128,
      "learning_rate": 4.536920789761286e-06,
      "loss": 1.3709,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 884
    },
    {
      "epoch": 0.555207026348808,
      "grad_norm": 1.7493175268173218,
      "learning_rate": 4.535332937713662e-06,
      "loss": 1.19,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 885
    },
    {
      "epoch": 0.5558343789209536,
      "grad_norm": 1.8383666276931763,
      "learning_rate": 4.533742646893086e-06,
      "loss": 1.259,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 886
    },
    {
      "epoch": 0.5564617314930991,
      "grad_norm": 2.549210786819458,
      "learning_rate": 4.5321499192050765e-06,
      "loss": 1.1625,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 887
    },
    {
      "epoch": 0.5570890840652447,
      "grad_norm": 1.8553715944290161,
      "learning_rate": 4.530554756558068e-06,
      "loss": 1.1967,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 888
    },
    {
      "epoch": 0.5577164366373902,
      "grad_norm": 2.863048553466797,
      "learning_rate": 4.528957160863412e-06,
      "loss": 1.1529,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 889
    },
    {
      "epoch": 0.5583437892095358,
      "grad_norm": 1.7921923398971558,
      "learning_rate": 4.527357134035379e-06,
      "loss": 1.3434,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 890
    },
    {
      "epoch": 0.5589711417816813,
      "grad_norm": 1.861366629600525,
      "learning_rate": 4.525754677991147e-06,
      "loss": 1.3188,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 891
    },
    {
      "epoch": 0.5595984943538268,
      "grad_norm": 1.7305035591125488,
      "learning_rate": 4.524149794650811e-06,
      "loss": 1.2743,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 892
    },
    {
      "epoch": 0.5602258469259724,
      "grad_norm": 1.790932059288025,
      "learning_rate": 4.522542485937369e-06,
      "loss": 1.2471,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 893
    },
    {
      "epoch": 0.560853199498118,
      "grad_norm": 1.6654341220855713,
      "learning_rate": 4.520932753776729e-06,
      "loss": 1.1872,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 894
    },
    {
      "epoch": 0.5614805520702635,
      "grad_norm": 1.4181909561157227,
      "learning_rate": 4.519320600097702e-06,
      "loss": 1.243,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 895
    },
    {
      "epoch": 0.562107904642409,
      "grad_norm": 1.785009741783142,
      "learning_rate": 4.5177060268319985e-06,
      "loss": 1.2753,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 896
    },
    {
      "epoch": 0.5627352572145545,
      "grad_norm": 1.4762773513793945,
      "learning_rate": 4.51608903591423e-06,
      "loss": 1.1472,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 897
    },
    {
      "epoch": 0.5633626097867002,
      "grad_norm": 1.899272084236145,
      "learning_rate": 4.514469629281905e-06,
      "loss": 1.2728,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 898
    },
    {
      "epoch": 0.5639899623588457,
      "grad_norm": 1.8159464597702026,
      "learning_rate": 4.512847808875424e-06,
      "loss": 1.3019,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 899
    },
    {
      "epoch": 0.5646173149309912,
      "grad_norm": 1.9190819263458252,
      "learning_rate": 4.511223576638084e-06,
      "loss": 1.3132,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 900
    },
    {
      "epoch": 0.5652446675031367,
      "grad_norm": 1.9054348468780518,
      "learning_rate": 4.5095969345160685e-06,
      "loss": 1.3536,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 901
    },
    {
      "epoch": 0.5658720200752823,
      "grad_norm": 1.8072419166564941,
      "learning_rate": 4.5079678844584505e-06,
      "loss": 1.3999,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 902
    },
    {
      "epoch": 0.5664993726474279,
      "grad_norm": 2.1858222484588623,
      "learning_rate": 4.5063364284171866e-06,
      "loss": 1.3366,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 903
    },
    {
      "epoch": 0.5671267252195734,
      "grad_norm": 1.7815707921981812,
      "learning_rate": 4.504702568347117e-06,
      "loss": 1.1586,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 904
    },
    {
      "epoch": 0.5677540777917189,
      "grad_norm": 1.7281264066696167,
      "learning_rate": 4.503066306205963e-06,
      "loss": 1.2503,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 905
    },
    {
      "epoch": 0.5683814303638645,
      "grad_norm": 1.8887368440628052,
      "learning_rate": 4.501427643954324e-06,
      "loss": 1.2161,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 906
    },
    {
      "epoch": 0.56900878293601,
      "grad_norm": 1.9003872871398926,
      "learning_rate": 4.499786583555675e-06,
      "loss": 1.2358,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 907
    },
    {
      "epoch": 0.5696361355081556,
      "grad_norm": 1.6889976263046265,
      "learning_rate": 4.498143126976362e-06,
      "loss": 1.2401,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 908
    },
    {
      "epoch": 0.5702634880803011,
      "grad_norm": 1.8518801927566528,
      "learning_rate": 4.4964972761856086e-06,
      "loss": 1.2482,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 909
    },
    {
      "epoch": 0.5708908406524467,
      "grad_norm": 1.7183051109313965,
      "learning_rate": 4.494849033155499e-06,
      "loss": 1.3633,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 910
    },
    {
      "epoch": 0.5715181932245922,
      "grad_norm": 1.5683485269546509,
      "learning_rate": 4.4931983998609915e-06,
      "loss": 1.2427,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 911
    },
    {
      "epoch": 0.5721455457967378,
      "grad_norm": 1.8224354982376099,
      "learning_rate": 4.491545378279903e-06,
      "loss": 1.2505,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 912
    },
    {
      "epoch": 0.5727728983688833,
      "grad_norm": 1.7461055517196655,
      "learning_rate": 4.489889970392915e-06,
      "loss": 1.1269,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 913
    },
    {
      "epoch": 0.5734002509410289,
      "grad_norm": 1.4676216840744019,
      "learning_rate": 4.4882321781835666e-06,
      "loss": 1.3263,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 914
    },
    {
      "epoch": 0.5740276035131744,
      "grad_norm": 1.9573190212249756,
      "learning_rate": 4.486572003638254e-06,
      "loss": 1.2915,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 915
    },
    {
      "epoch": 0.5746549560853199,
      "grad_norm": 1.5196433067321777,
      "learning_rate": 4.4849094487462305e-06,
      "loss": 1.236,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 916
    },
    {
      "epoch": 0.5752823086574655,
      "grad_norm": 1.8737750053405762,
      "learning_rate": 4.483244515499598e-06,
      "loss": 1.2711,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 917
    },
    {
      "epoch": 0.5759096612296111,
      "grad_norm": 1.7228468656539917,
      "learning_rate": 4.48157720589331e-06,
      "loss": 1.2358,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 918
    },
    {
      "epoch": 0.5765370138017566,
      "grad_norm": 2.034825325012207,
      "learning_rate": 4.479907521925168e-06,
      "loss": 1.2302,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 919
    },
    {
      "epoch": 0.5771643663739021,
      "grad_norm": 1.7857717275619507,
      "learning_rate": 4.478235465595817e-06,
      "loss": 1.4052,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 920
    },
    {
      "epoch": 0.5777917189460476,
      "grad_norm": 1.714264154434204,
      "learning_rate": 4.476561038908745e-06,
      "loss": 1.2664,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 921
    },
    {
      "epoch": 0.5784190715181933,
      "grad_norm": 1.6064285039901733,
      "learning_rate": 4.474884243870281e-06,
      "loss": 1.3298,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 922
    },
    {
      "epoch": 0.5790464240903388,
      "grad_norm": 1.78461492061615,
      "learning_rate": 4.473205082489592e-06,
      "loss": 1.3199,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 923
    },
    {
      "epoch": 0.5796737766624843,
      "grad_norm": 1.6482230424880981,
      "learning_rate": 4.471523556778679e-06,
      "loss": 1.204,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 924
    },
    {
      "epoch": 0.5803011292346298,
      "grad_norm": 1.351148247718811,
      "learning_rate": 4.4698396687523765e-06,
      "loss": 1.1804,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 925
    },
    {
      "epoch": 0.5809284818067754,
      "grad_norm": 1.809248685836792,
      "learning_rate": 4.46815342042835e-06,
      "loss": 1.1648,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 926
    },
    {
      "epoch": 0.581555834378921,
      "grad_norm": 1.9116644859313965,
      "learning_rate": 4.466464813827093e-06,
      "loss": 1.3359,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 927
    },
    {
      "epoch": 0.5821831869510665,
      "grad_norm": 1.755806565284729,
      "learning_rate": 4.464773850971924e-06,
      "loss": 1.2192,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 928
    },
    {
      "epoch": 0.582810539523212,
      "grad_norm": 1.5529253482818604,
      "learning_rate": 4.463080533888987e-06,
      "loss": 1.2275,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 929
    },
    {
      "epoch": 0.5834378920953576,
      "grad_norm": 1.8902771472930908,
      "learning_rate": 4.461384864607243e-06,
      "loss": 1.2347,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 930
    },
    {
      "epoch": 0.5840652446675031,
      "grad_norm": 1.9781829118728638,
      "learning_rate": 4.459686845158476e-06,
      "loss": 1.2602,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 931
    },
    {
      "epoch": 0.5846925972396487,
      "grad_norm": 1.968216896057129,
      "learning_rate": 4.457986477577283e-06,
      "loss": 1.1981,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 932
    },
    {
      "epoch": 0.5853199498117942,
      "grad_norm": 1.5119919776916504,
      "learning_rate": 4.456283763901075e-06,
      "loss": 1.2146,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 933
    },
    {
      "epoch": 0.5859473023839398,
      "grad_norm": 1.8399906158447266,
      "learning_rate": 4.454578706170075e-06,
      "loss": 1.3099,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 934
    },
    {
      "epoch": 0.5865746549560853,
      "grad_norm": 1.8300646543502808,
      "learning_rate": 4.452871306427314e-06,
      "loss": 1.1918,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 935
    },
    {
      "epoch": 0.5872020075282308,
      "grad_norm": 1.4192575216293335,
      "learning_rate": 4.451161566718629e-06,
      "loss": 1.1847,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 936
    },
    {
      "epoch": 0.5878293601003765,
      "grad_norm": 1.9011478424072266,
      "learning_rate": 4.4494494890926616e-06,
      "loss": 1.3157,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 937
    },
    {
      "epoch": 0.588456712672522,
      "grad_norm": 1.544992208480835,
      "learning_rate": 4.447735075600855e-06,
      "loss": 1.246,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 938
    },
    {
      "epoch": 0.5890840652446675,
      "grad_norm": 2.110203266143799,
      "learning_rate": 4.446018328297449e-06,
      "loss": 1.3751,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 939
    },
    {
      "epoch": 0.589711417816813,
      "grad_norm": 1.8133139610290527,
      "learning_rate": 4.444299249239483e-06,
      "loss": 1.2828,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 940
    },
    {
      "epoch": 0.5903387703889585,
      "grad_norm": 1.8554705381393433,
      "learning_rate": 4.442577840486789e-06,
      "loss": 1.4578,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 941
    },
    {
      "epoch": 0.5909661229611042,
      "grad_norm": 1.9675077199935913,
      "learning_rate": 4.440854104101988e-06,
      "loss": 1.1713,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 942
    },
    {
      "epoch": 0.5915934755332497,
      "grad_norm": 4.033464431762695,
      "learning_rate": 4.439128042150495e-06,
      "loss": 1.3743,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 943
    },
    {
      "epoch": 0.5922208281053952,
      "grad_norm": 1.724401593208313,
      "learning_rate": 4.437399656700507e-06,
      "loss": 1.2972,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 944
    },
    {
      "epoch": 0.5928481806775407,
      "grad_norm": 1.488004446029663,
      "learning_rate": 4.435668949823008e-06,
      "loss": 1.2991,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 945
    },
    {
      "epoch": 0.5934755332496863,
      "grad_norm": 1.8987462520599365,
      "learning_rate": 4.433935923591763e-06,
      "loss": 1.3099,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 946
    },
    {
      "epoch": 0.5941028858218319,
      "grad_norm": 1.8244067430496216,
      "learning_rate": 4.432200580083312e-06,
      "loss": 1.4052,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 947
    },
    {
      "epoch": 0.5947302383939774,
      "grad_norm": 1.6409897804260254,
      "learning_rate": 4.43046292137698e-06,
      "loss": 1.2372,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 948
    },
    {
      "epoch": 0.595357590966123,
      "grad_norm": 1.771867275238037,
      "learning_rate": 4.428722949554858e-06,
      "loss": 1.2957,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 949
    },
    {
      "epoch": 0.5959849435382685,
      "grad_norm": 1.882083773612976,
      "learning_rate": 4.426980666701813e-06,
      "loss": 1.3752,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 950
    },
    {
      "epoch": 0.596612296110414,
      "grad_norm": 1.818285584449768,
      "learning_rate": 4.42523607490548e-06,
      "loss": 1.3053,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 951
    },
    {
      "epoch": 0.5972396486825596,
      "grad_norm": 1.6514558792114258,
      "learning_rate": 4.42348917625626e-06,
      "loss": 1.1995,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 952
    },
    {
      "epoch": 0.5978670012547052,
      "grad_norm": 1.948164701461792,
      "learning_rate": 4.42173997284732e-06,
      "loss": 1.2561,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 953
    },
    {
      "epoch": 0.5984943538268507,
      "grad_norm": 1.9465723037719727,
      "learning_rate": 4.4199884667745866e-06,
      "loss": 1.3594,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 954
    },
    {
      "epoch": 0.5991217063989962,
      "grad_norm": 1.5567758083343506,
      "learning_rate": 4.418234660136745e-06,
      "loss": 1.1343,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 955
    },
    {
      "epoch": 0.5997490589711418,
      "grad_norm": 1.8369762897491455,
      "learning_rate": 4.416478555035241e-06,
      "loss": 1.2743,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 956
    },
    {
      "epoch": 0.6003764115432874,
      "grad_norm": 1.8079118728637695,
      "learning_rate": 4.414720153574269e-06,
      "loss": 1.4227,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 957
    },
    {
      "epoch": 0.6010037641154329,
      "grad_norm": 1.668028712272644,
      "learning_rate": 4.412959457860779e-06,
      "loss": 1.4489,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 958
    },
    {
      "epoch": 0.6016311166875784,
      "grad_norm": 1.7507413625717163,
      "learning_rate": 4.4111964700044684e-06,
      "loss": 1.2152,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 959
    },
    {
      "epoch": 0.6022584692597239,
      "grad_norm": 1.9011902809143066,
      "learning_rate": 4.409431192117782e-06,
      "loss": 1.2223,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 960
    },
    {
      "epoch": 0.6028858218318696,
      "grad_norm": 2.240939140319824,
      "learning_rate": 4.407663626315907e-06,
      "loss": 1.2813,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 961
    },
    {
      "epoch": 0.6035131744040151,
      "grad_norm": 1.7186036109924316,
      "learning_rate": 4.405893774716772e-06,
      "loss": 1.1469,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 962
    },
    {
      "epoch": 0.6041405269761606,
      "grad_norm": 1.5372388362884521,
      "learning_rate": 4.404121639441047e-06,
      "loss": 1.2067,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 963
    },
    {
      "epoch": 0.6047678795483061,
      "grad_norm": 1.7399442195892334,
      "learning_rate": 4.402347222612137e-06,
      "loss": 1.1696,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 964
    },
    {
      "epoch": 0.6053952321204517,
      "grad_norm": 1.8909478187561035,
      "learning_rate": 4.400570526356178e-06,
      "loss": 1.292,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 965
    },
    {
      "epoch": 0.6060225846925973,
      "grad_norm": 1.7780414819717407,
      "learning_rate": 4.398791552802043e-06,
      "loss": 1.5044,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 966
    },
    {
      "epoch": 0.6066499372647428,
      "grad_norm": 1.9680545330047607,
      "learning_rate": 4.397010304081328e-06,
      "loss": 1.3634,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 967
    },
    {
      "epoch": 0.6072772898368883,
      "grad_norm": 1.8577983379364014,
      "learning_rate": 4.39522678232836e-06,
      "loss": 1.2081,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 968
    },
    {
      "epoch": 0.6079046424090339,
      "grad_norm": 2.26708984375,
      "learning_rate": 4.393440989680184e-06,
      "loss": 1.3962,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 969
    },
    {
      "epoch": 0.6085319949811794,
      "grad_norm": 1.7874890565872192,
      "learning_rate": 4.391652928276572e-06,
      "loss": 1.2756,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 970
    },
    {
      "epoch": 0.609159347553325,
      "grad_norm": 2.3963043689727783,
      "learning_rate": 4.38986260026001e-06,
      "loss": 1.1627,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 971
    },
    {
      "epoch": 0.6097867001254705,
      "grad_norm": 1.408721685409546,
      "learning_rate": 4.388070007775703e-06,
      "loss": 1.1634,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 972
    },
    {
      "epoch": 0.6104140526976161,
      "grad_norm": 1.7340478897094727,
      "learning_rate": 4.3862751529715674e-06,
      "loss": 1.2605,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 973
    },
    {
      "epoch": 0.6110414052697616,
      "grad_norm": 1.6462897062301636,
      "learning_rate": 4.38447803799823e-06,
      "loss": 1.1403,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 974
    },
    {
      "epoch": 0.6116687578419071,
      "grad_norm": 1.8389941453933716,
      "learning_rate": 4.382678665009028e-06,
      "loss": 1.2517,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 975
    },
    {
      "epoch": 0.6122961104140527,
      "grad_norm": 1.944242238998413,
      "learning_rate": 4.380877036160002e-06,
      "loss": 1.2953,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 976
    },
    {
      "epoch": 0.6129234629861983,
      "grad_norm": 1.7658624649047852,
      "learning_rate": 4.379073153609896e-06,
      "loss": 1.2436,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 977
    },
    {
      "epoch": 0.6135508155583438,
      "grad_norm": 2.0299324989318848,
      "learning_rate": 4.3772670195201565e-06,
      "loss": 1.4657,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 978
    },
    {
      "epoch": 0.6141781681304893,
      "grad_norm": 1.6620166301727295,
      "learning_rate": 4.375458636054924e-06,
      "loss": 1.2166,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 979
    },
    {
      "epoch": 0.6148055207026348,
      "grad_norm": 2.2326536178588867,
      "learning_rate": 4.373648005381039e-06,
      "loss": 1.234,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 980
    },
    {
      "epoch": 0.6154328732747805,
      "grad_norm": 1.7606323957443237,
      "learning_rate": 4.37183512966803e-06,
      "loss": 1.3452,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 981
    },
    {
      "epoch": 0.616060225846926,
      "grad_norm": 1.1604722738265991,
      "learning_rate": 4.37002001108812e-06,
      "loss": 1.3058,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 982
    },
    {
      "epoch": 0.6166875784190715,
      "grad_norm": 1.894524097442627,
      "learning_rate": 4.368202651816213e-06,
      "loss": 1.3107,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 983
    },
    {
      "epoch": 0.617314930991217,
      "grad_norm": 1.7152777910232544,
      "learning_rate": 4.366383054029907e-06,
      "loss": 1.4284,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 984
    },
    {
      "epoch": 0.6179422835633626,
      "grad_norm": 1.7929563522338867,
      "learning_rate": 4.364561219909474e-06,
      "loss": 1.3094,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 985
    },
    {
      "epoch": 0.6185696361355082,
      "grad_norm": 2.654740571975708,
      "learning_rate": 4.3627371516378695e-06,
      "loss": 1.2025,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 986
    },
    {
      "epoch": 0.6191969887076537,
      "grad_norm": 1.9211310148239136,
      "learning_rate": 4.360910851400725e-06,
      "loss": 1.2593,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 987
    },
    {
      "epoch": 0.6198243412797992,
      "grad_norm": 1.8810856342315674,
      "learning_rate": 4.359082321386346e-06,
      "loss": 1.2612,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 988
    },
    {
      "epoch": 0.6204516938519448,
      "grad_norm": 1.736060380935669,
      "learning_rate": 4.357251563785712e-06,
      "loss": 1.2275,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 989
    },
    {
      "epoch": 0.6210790464240903,
      "grad_norm": 1.788694977760315,
      "learning_rate": 4.355418580792471e-06,
      "loss": 1.2669,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 990
    },
    {
      "epoch": 0.6217063989962359,
      "grad_norm": 1.666709542274475,
      "learning_rate": 4.3535833746029335e-06,
      "loss": 1.3095,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 991
    },
    {
      "epoch": 0.6223337515683814,
      "grad_norm": 1.8080217838287354,
      "learning_rate": 4.351745947416079e-06,
      "loss": 1.2086,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 992
    },
    {
      "epoch": 0.622961104140527,
      "grad_norm": 1.7072173357009888,
      "learning_rate": 4.349906301433545e-06,
      "loss": 1.2606,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 993
    },
    {
      "epoch": 0.6235884567126725,
      "grad_norm": 1.833138346672058,
      "learning_rate": 4.348064438859629e-06,
      "loss": 1.3743,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 994
    },
    {
      "epoch": 0.624215809284818,
      "grad_norm": 1.8404881954193115,
      "learning_rate": 4.346220361901286e-06,
      "loss": 1.1828,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 995
    },
    {
      "epoch": 0.6248431618569636,
      "grad_norm": 1.721911907196045,
      "learning_rate": 4.34437407276812e-06,
      "loss": 1.24,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 996
    },
    {
      "epoch": 0.6254705144291092,
      "grad_norm": 1.9910308122634888,
      "learning_rate": 4.342525573672391e-06,
      "loss": 1.2866,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 997
    },
    {
      "epoch": 0.6260978670012547,
      "grad_norm": 1.7354001998901367,
      "learning_rate": 4.340674866829001e-06,
      "loss": 1.2425,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 998
    },
    {
      "epoch": 0.6267252195734002,
      "grad_norm": 1.937098503112793,
      "learning_rate": 4.3388219544555035e-06,
      "loss": 1.2484,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 999
    },
    {
      "epoch": 0.6273525721455459,
      "grad_norm": 1.6897847652435303,
      "learning_rate": 4.3369668387720895e-06,
      "loss": 1.2683,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1000
    },
    {
      "epoch": 0.6279799247176914,
      "grad_norm": 1.848845362663269,
      "learning_rate": 4.335109522001594e-06,
      "loss": 1.3694,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1001
    },
    {
      "epoch": 0.6286072772898369,
      "grad_norm": 1.9161001443862915,
      "learning_rate": 4.333250006369487e-06,
      "loss": 1.3237,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1002
    },
    {
      "epoch": 0.6292346298619824,
      "grad_norm": 1.8927891254425049,
      "learning_rate": 4.331388294103874e-06,
      "loss": 1.3337,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1003
    },
    {
      "epoch": 0.6298619824341279,
      "grad_norm": 1.8695149421691895,
      "learning_rate": 4.329524387435493e-06,
      "loss": 1.2412,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1004
    },
    {
      "epoch": 0.6304893350062736,
      "grad_norm": 1.4502034187316895,
      "learning_rate": 4.32765828859771e-06,
      "loss": 1.2586,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1005
    },
    {
      "epoch": 0.6311166875784191,
      "grad_norm": 1.9272611141204834,
      "learning_rate": 4.3257899998265195e-06,
      "loss": 1.1892,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1006
    },
    {
      "epoch": 0.6317440401505646,
      "grad_norm": 1.830952763557434,
      "learning_rate": 4.323919523360539e-06,
      "loss": 1.408,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1007
    },
    {
      "epoch": 0.6323713927227101,
      "grad_norm": 1.6106399297714233,
      "learning_rate": 4.322046861441006e-06,
      "loss": 1.4208,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1008
    },
    {
      "epoch": 0.6329987452948557,
      "grad_norm": 1.7375121116638184,
      "learning_rate": 4.32017201631178e-06,
      "loss": 1.35,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1009
    },
    {
      "epoch": 0.6336260978670013,
      "grad_norm": 1.7395000457763672,
      "learning_rate": 4.318294990219334e-06,
      "loss": 1.3972,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1010
    },
    {
      "epoch": 0.6342534504391468,
      "grad_norm": 1.8794714212417603,
      "learning_rate": 4.316415785412754e-06,
      "loss": 1.232,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1011
    },
    {
      "epoch": 0.6348808030112923,
      "grad_norm": 1.9169164896011353,
      "learning_rate": 4.314534404143738e-06,
      "loss": 1.3077,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1012
    },
    {
      "epoch": 0.6355081555834379,
      "grad_norm": 1.6044981479644775,
      "learning_rate": 4.312650848666591e-06,
      "loss": 1.3272,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1013
    },
    {
      "epoch": 0.6361355081555834,
      "grad_norm": 1.8915094137191772,
      "learning_rate": 4.310765121238223e-06,
      "loss": 1.264,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1014
    },
    {
      "epoch": 0.636762860727729,
      "grad_norm": 1.7705018520355225,
      "learning_rate": 4.308877224118148e-06,
      "loss": 1.3933,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1015
    },
    {
      "epoch": 0.6373902132998746,
      "grad_norm": 1.880327582359314,
      "learning_rate": 4.3069871595684795e-06,
      "loss": 1.2974,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1016
    },
    {
      "epoch": 0.6380175658720201,
      "grad_norm": 1.5655988454818726,
      "learning_rate": 4.305094929853925e-06,
      "loss": 1.2287,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1017
    },
    {
      "epoch": 0.6386449184441656,
      "grad_norm": 1.868777871131897,
      "learning_rate": 4.303200537241789e-06,
      "loss": 1.3101,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1018
    },
    {
      "epoch": 0.6392722710163111,
      "grad_norm": 1.8370649814605713,
      "learning_rate": 4.3013039840019675e-06,
      "loss": 1.2703,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1019
    },
    {
      "epoch": 0.6398996235884568,
      "grad_norm": 1.7365206480026245,
      "learning_rate": 4.299405272406946e-06,
      "loss": 1.2183,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1020
    },
    {
      "epoch": 0.6405269761606023,
      "grad_norm": 1.7789405584335327,
      "learning_rate": 4.297504404731794e-06,
      "loss": 1.2166,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1021
    },
    {
      "epoch": 0.6411543287327478,
      "grad_norm": 1.79762601852417,
      "learning_rate": 4.295601383254166e-06,
      "loss": 1.2795,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1022
    },
    {
      "epoch": 0.6417816813048933,
      "grad_norm": 2.0368082523345947,
      "learning_rate": 4.293696210254299e-06,
      "loss": 1.1092,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1023
    },
    {
      "epoch": 0.6424090338770388,
      "grad_norm": 1.8230146169662476,
      "learning_rate": 4.291788888015002e-06,
      "loss": 1.3562,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1024
    },
    {
      "epoch": 0.6430363864491845,
      "grad_norm": 1.6714571714401245,
      "learning_rate": 4.289879418821667e-06,
      "loss": 1.2922,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1025
    },
    {
      "epoch": 0.64366373902133,
      "grad_norm": 1.827463150024414,
      "learning_rate": 4.287967804962252e-06,
      "loss": 1.1723,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1026
    },
    {
      "epoch": 0.6442910915934755,
      "grad_norm": 1.7430213689804077,
      "learning_rate": 4.286054048727289e-06,
      "loss": 1.3563,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1027
    },
    {
      "epoch": 0.644918444165621,
      "grad_norm": 1.9313528537750244,
      "learning_rate": 4.284138152409875e-06,
      "loss": 1.328,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1028
    },
    {
      "epoch": 0.6455457967377666,
      "grad_norm": 1.788396954536438,
      "learning_rate": 4.282220118305672e-06,
      "loss": 1.2578,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1029
    },
    {
      "epoch": 0.6461731493099122,
      "grad_norm": 1.9158381223678589,
      "learning_rate": 4.2802999487129025e-06,
      "loss": 1.2759,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1030
    },
    {
      "epoch": 0.6468005018820577,
      "grad_norm": 1.5709127187728882,
      "learning_rate": 4.278377645932351e-06,
      "loss": 1.2669,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1031
    },
    {
      "epoch": 0.6474278544542033,
      "grad_norm": 1.917354941368103,
      "learning_rate": 4.276453212267353e-06,
      "loss": 1.3398,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1032
    },
    {
      "epoch": 0.6480552070263488,
      "grad_norm": 1.9424017667770386,
      "learning_rate": 4.274526650023801e-06,
      "loss": 1.1625,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1033
    },
    {
      "epoch": 0.6486825595984943,
      "grad_norm": 1.6027320623397827,
      "learning_rate": 4.272597961510137e-06,
      "loss": 1.3137,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1034
    },
    {
      "epoch": 0.6493099121706399,
      "grad_norm": 1.4474650621414185,
      "learning_rate": 4.2706671490373505e-06,
      "loss": 1.2254,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1035
    },
    {
      "epoch": 0.6499372647427855,
      "grad_norm": 1.7298493385314941,
      "learning_rate": 4.268734214918977e-06,
      "loss": 1.2897,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1036
    },
    {
      "epoch": 0.650564617314931,
      "grad_norm": 1.4415090084075928,
      "learning_rate": 4.266799161471093e-06,
      "loss": 1.1688,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1037
    },
    {
      "epoch": 0.6511919698870765,
      "grad_norm": 2.0625054836273193,
      "learning_rate": 4.264861991012312e-06,
      "loss": 1.3359,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1038
    },
    {
      "epoch": 0.651819322459222,
      "grad_norm": 1.8560051918029785,
      "learning_rate": 4.262922705863791e-06,
      "loss": 1.3139,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1039
    },
    {
      "epoch": 0.6524466750313677,
      "grad_norm": 1.620300054550171,
      "learning_rate": 4.260981308349214e-06,
      "loss": 1.1617,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1040
    },
    {
      "epoch": 0.6530740276035132,
      "grad_norm": 1.65470552444458,
      "learning_rate": 4.2590378007948e-06,
      "loss": 1.1033,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1041
    },
    {
      "epoch": 0.6537013801756587,
      "grad_norm": 2.6354520320892334,
      "learning_rate": 4.257092185529295e-06,
      "loss": 1.3497,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1042
    },
    {
      "epoch": 0.6543287327478042,
      "grad_norm": 1.8058476448059082,
      "learning_rate": 4.25514446488397e-06,
      "loss": 1.304,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1043
    },
    {
      "epoch": 0.6549560853199499,
      "grad_norm": 2.0276644229888916,
      "learning_rate": 4.253194641192621e-06,
      "loss": 1.3475,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1044
    },
    {
      "epoch": 0.6555834378920954,
      "grad_norm": 1.907875895500183,
      "learning_rate": 4.2512427167915595e-06,
      "loss": 1.4468,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1045
    },
    {
      "epoch": 0.6562107904642409,
      "grad_norm": 1.8480969667434692,
      "learning_rate": 4.249288694019618e-06,
      "loss": 1.1998,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1046
    },
    {
      "epoch": 0.6568381430363864,
      "grad_norm": 1.6065888404846191,
      "learning_rate": 4.247332575218144e-06,
      "loss": 1.2323,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1047
    },
    {
      "epoch": 0.657465495608532,
      "grad_norm": 1.8290810585021973,
      "learning_rate": 4.245374362730992e-06,
      "loss": 1.2553,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1048
    },
    {
      "epoch": 0.6580928481806776,
      "grad_norm": 1.9927012920379639,
      "learning_rate": 4.2434140589045286e-06,
      "loss": 1.1855,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1049
    },
    {
      "epoch": 0.6587202007528231,
      "grad_norm": 1.6697808504104614,
      "learning_rate": 4.241451666087626e-06,
      "loss": 1.2958,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1050
    },
    {
      "epoch": 0.6593475533249686,
      "grad_norm": 1.8916678428649902,
      "learning_rate": 4.239487186631659e-06,
      "loss": 1.1373,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1051
    },
    {
      "epoch": 0.6599749058971142,
      "grad_norm": 1.8192722797393799,
      "learning_rate": 4.237520622890503e-06,
      "loss": 1.2844,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1052
    },
    {
      "epoch": 0.6606022584692597,
      "grad_norm": 1.9123562574386597,
      "learning_rate": 4.235551977220529e-06,
      "loss": 1.3009,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1053
    },
    {
      "epoch": 0.6612296110414053,
      "grad_norm": 1.4059549570083618,
      "learning_rate": 4.233581251980604e-06,
      "loss": 1.2611,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1054
    },
    {
      "epoch": 0.6618569636135508,
      "grad_norm": 1.9050414562225342,
      "learning_rate": 4.23160844953209e-06,
      "loss": 1.3481,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1055
    },
    {
      "epoch": 0.6624843161856964,
      "grad_norm": 1.5960050821304321,
      "learning_rate": 4.229633572238831e-06,
      "loss": 1.1204,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1056
    },
    {
      "epoch": 0.6631116687578419,
      "grad_norm": 1.4794005155563354,
      "learning_rate": 4.227656622467162e-06,
      "loss": 1.1655,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1057
    },
    {
      "epoch": 0.6637390213299874,
      "grad_norm": 1.9069862365722656,
      "learning_rate": 4.225677602585901e-06,
      "loss": 1.321,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1058
    },
    {
      "epoch": 0.664366373902133,
      "grad_norm": 1.4333158731460571,
      "learning_rate": 4.223696514966346e-06,
      "loss": 1.23,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1059
    },
    {
      "epoch": 0.6649937264742786,
      "grad_norm": 1.650801658630371,
      "learning_rate": 4.2217133619822705e-06,
      "loss": 1.2287,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1060
    },
    {
      "epoch": 0.6656210790464241,
      "grad_norm": 2.9155123233795166,
      "learning_rate": 4.2197281460099245e-06,
      "loss": 1.209,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1061
    },
    {
      "epoch": 0.6662484316185696,
      "grad_norm": 1.8751521110534668,
      "learning_rate": 4.217740869428032e-06,
      "loss": 1.3401,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1062
    },
    {
      "epoch": 0.6668757841907151,
      "grad_norm": 1.9066846370697021,
      "learning_rate": 4.215751534617783e-06,
      "loss": 1.28,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1063
    },
    {
      "epoch": 0.6675031367628608,
      "grad_norm": 1.853600263595581,
      "learning_rate": 4.213760143962834e-06,
      "loss": 1.36,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1064
    },
    {
      "epoch": 0.6681304893350063,
      "grad_norm": 1.5284377336502075,
      "learning_rate": 4.211766699849306e-06,
      "loss": 1.243,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1065
    },
    {
      "epoch": 0.6687578419071518,
      "grad_norm": 1.8650940656661987,
      "learning_rate": 4.209771204665779e-06,
      "loss": 1.2,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1066
    },
    {
      "epoch": 0.6693851944792973,
      "grad_norm": 1.7665506601333618,
      "learning_rate": 4.2077736608032925e-06,
      "loss": 1.1567,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1067
    },
    {
      "epoch": 0.6700125470514429,
      "grad_norm": 1.9517008066177368,
      "learning_rate": 4.2057740706553415e-06,
      "loss": 1.3147,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1068
    },
    {
      "epoch": 0.6706398996235885,
      "grad_norm": 1.7581521272659302,
      "learning_rate": 4.203772436617868e-06,
      "loss": 1.427,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1069
    },
    {
      "epoch": 0.671267252195734,
      "grad_norm": 1.9802473783493042,
      "learning_rate": 4.201768761089269e-06,
      "loss": 1.3791,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1070
    },
    {
      "epoch": 0.6718946047678795,
      "grad_norm": 1.914982795715332,
      "learning_rate": 4.199763046470384e-06,
      "loss": 1.3415,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1071
    },
    {
      "epoch": 0.6725219573400251,
      "grad_norm": 1.864111065864563,
      "learning_rate": 4.197755295164496e-06,
      "loss": 1.3969,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1072
    },
    {
      "epoch": 0.6731493099121706,
      "grad_norm": 1.913025975227356,
      "learning_rate": 4.19574550957733e-06,
      "loss": 1.2561,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1073
    },
    {
      "epoch": 0.6737766624843162,
      "grad_norm": 1.9390740394592285,
      "learning_rate": 4.193733692117048e-06,
      "loss": 1.2471,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1074
    },
    {
      "epoch": 0.6744040150564617,
      "grad_norm": 1.7494572401046753,
      "learning_rate": 4.191719845194246e-06,
      "loss": 1.1353,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1075
    },
    {
      "epoch": 0.6750313676286073,
      "grad_norm": 1.8655979633331299,
      "learning_rate": 4.1897039712219514e-06,
      "loss": 1.3104,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1076
    },
    {
      "epoch": 0.6756587202007528,
      "grad_norm": 1.9206185340881348,
      "learning_rate": 4.187686072615621e-06,
      "loss": 1.3032,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1077
    },
    {
      "epoch": 0.6762860727728983,
      "grad_norm": 1.896394968032837,
      "learning_rate": 4.185666151793139e-06,
      "loss": 1.4611,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1078
    },
    {
      "epoch": 0.676913425345044,
      "grad_norm": 1.8265060186386108,
      "learning_rate": 4.1836442111748086e-06,
      "loss": 1.2732,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1079
    },
    {
      "epoch": 0.6775407779171895,
      "grad_norm": 1.7321994304656982,
      "learning_rate": 4.181620253183359e-06,
      "loss": 1.233,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1080
    },
    {
      "epoch": 0.678168130489335,
      "grad_norm": 1.9845069646835327,
      "learning_rate": 4.179594280243932e-06,
      "loss": 1.2736,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1081
    },
    {
      "epoch": 0.6787954830614805,
      "grad_norm": 1.7132633924484253,
      "learning_rate": 4.177566294784085e-06,
      "loss": 1.4785,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1082
    },
    {
      "epoch": 0.679422835633626,
      "grad_norm": 1.8451826572418213,
      "learning_rate": 4.175536299233788e-06,
      "loss": 1.207,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1083
    },
    {
      "epoch": 0.6800501882057717,
      "grad_norm": 1.9522786140441895,
      "learning_rate": 4.173504296025417e-06,
      "loss": 1.1463,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1084
    },
    {
      "epoch": 0.6806775407779172,
      "grad_norm": 1.8040053844451904,
      "learning_rate": 4.171470287593757e-06,
      "loss": 1.3937,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1085
    },
    {
      "epoch": 0.6813048933500627,
      "grad_norm": 1.8072997331619263,
      "learning_rate": 4.169434276375992e-06,
      "loss": 1.258,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1086
    },
    {
      "epoch": 0.6819322459222082,
      "grad_norm": 1.5943018198013306,
      "learning_rate": 4.167396264811709e-06,
      "loss": 1.2709,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1087
    },
    {
      "epoch": 0.6825595984943539,
      "grad_norm": 1.5378663539886475,
      "learning_rate": 4.165356255342892e-06,
      "loss": 1.2698,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1088
    },
    {
      "epoch": 0.6831869510664994,
      "grad_norm": 1.8408434391021729,
      "learning_rate": 4.163314250413913e-06,
      "loss": 1.3284,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1089
    },
    {
      "epoch": 0.6838143036386449,
      "grad_norm": 2.0114541053771973,
      "learning_rate": 4.1612702524715445e-06,
      "loss": 1.2577,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1090
    },
    {
      "epoch": 0.6844416562107905,
      "grad_norm": 1.7099038362503052,
      "learning_rate": 4.159224263964939e-06,
      "loss": 1.2858,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1091
    },
    {
      "epoch": 0.685069008782936,
      "grad_norm": 1.7164630889892578,
      "learning_rate": 4.1571762873456376e-06,
      "loss": 1.3028,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1092
    },
    {
      "epoch": 0.6856963613550816,
      "grad_norm": 1.8126477003097534,
      "learning_rate": 4.1551263250675635e-06,
      "loss": 1.283,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1093
    },
    {
      "epoch": 0.6863237139272271,
      "grad_norm": 1.8203259706497192,
      "learning_rate": 4.153074379587018e-06,
      "loss": 1.4703,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1094
    },
    {
      "epoch": 0.6869510664993727,
      "grad_norm": 1.6437162160873413,
      "learning_rate": 4.151020453362681e-06,
      "loss": 1.3509,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1095
    },
    {
      "epoch": 0.6875784190715182,
      "grad_norm": 2.030787944793701,
      "learning_rate": 4.148964548855603e-06,
      "loss": 1.2699,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1096
    },
    {
      "epoch": 0.6882057716436637,
      "grad_norm": 1.8610453605651855,
      "learning_rate": 4.146906668529207e-06,
      "loss": 1.2309,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1097
    },
    {
      "epoch": 0.6888331242158093,
      "grad_norm": 1.8488279581069946,
      "learning_rate": 4.144846814849282e-06,
      "loss": 1.289,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1098
    },
    {
      "epoch": 0.6894604767879549,
      "grad_norm": 1.969459891319275,
      "learning_rate": 4.1427849902839826e-06,
      "loss": 1.2552,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1099
    },
    {
      "epoch": 0.6900878293601004,
      "grad_norm": 1.5772993564605713,
      "learning_rate": 4.1407211973038245e-06,
      "loss": 1.2722,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1100
    },
    {
      "epoch": 0.6907151819322459,
      "grad_norm": 1.722180962562561,
      "learning_rate": 4.138655438381681e-06,
      "loss": 1.3136,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1101
    },
    {
      "epoch": 0.6913425345043914,
      "grad_norm": 1.7175124883651733,
      "learning_rate": 4.1365877159927835e-06,
      "loss": 1.2704,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1102
    },
    {
      "epoch": 0.6919698870765371,
      "grad_norm": 1.7393913269042969,
      "learning_rate": 4.134518032614713e-06,
      "loss": 1.3046,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1103
    },
    {
      "epoch": 0.6925972396486826,
      "grad_norm": 1.930572509765625,
      "learning_rate": 4.1324463907274035e-06,
      "loss": 1.4181,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1104
    },
    {
      "epoch": 0.6932245922208281,
      "grad_norm": 1.9321399927139282,
      "learning_rate": 4.130372792813133e-06,
      "loss": 1.3217,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1105
    },
    {
      "epoch": 0.6938519447929736,
      "grad_norm": 1.7585630416870117,
      "learning_rate": 4.128297241356525e-06,
      "loss": 1.1922,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1106
    },
    {
      "epoch": 0.6944792973651192,
      "grad_norm": 1.8687766790390015,
      "learning_rate": 4.126219738844542e-06,
      "loss": 1.3053,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1107
    },
    {
      "epoch": 0.6951066499372648,
      "grad_norm": 1.7932958602905273,
      "learning_rate": 4.1241402877664845e-06,
      "loss": 1.3013,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1108
    },
    {
      "epoch": 0.6957340025094103,
      "grad_norm": 1.5376317501068115,
      "learning_rate": 4.122058890613991e-06,
      "loss": 1.3376,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1109
    },
    {
      "epoch": 0.6963613550815558,
      "grad_norm": 1.9468011856079102,
      "learning_rate": 4.119975549881029e-06,
      "loss": 1.3899,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1110
    },
    {
      "epoch": 0.6969887076537014,
      "grad_norm": 3.2425899505615234,
      "learning_rate": 4.117890268063894e-06,
      "loss": 1.2131,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1111
    },
    {
      "epoch": 0.6976160602258469,
      "grad_norm": 1.381169080734253,
      "learning_rate": 4.1158030476612085e-06,
      "loss": 1.2724,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1112
    },
    {
      "epoch": 0.6982434127979925,
      "grad_norm": 1.7302320003509521,
      "learning_rate": 4.1137138911739185e-06,
      "loss": 1.2668,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1113
    },
    {
      "epoch": 0.698870765370138,
      "grad_norm": 1.9618453979492188,
      "learning_rate": 4.11162280110529e-06,
      "loss": 1.4453,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1114
    },
    {
      "epoch": 0.6994981179422836,
      "grad_norm": 1.8300155401229858,
      "learning_rate": 4.109529779960905e-06,
      "loss": 1.3275,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1115
    },
    {
      "epoch": 0.7001254705144291,
      "grad_norm": 1.7978266477584839,
      "learning_rate": 4.107434830248658e-06,
      "loss": 1.2542,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1116
    },
    {
      "epoch": 0.7007528230865746,
      "grad_norm": 1.8951441049575806,
      "learning_rate": 4.105337954478756e-06,
      "loss": 1.1943,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1117
    },
    {
      "epoch": 0.7013801756587202,
      "grad_norm": 1.8222427368164062,
      "learning_rate": 4.103239155163718e-06,
      "loss": 1.191,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1118
    },
    {
      "epoch": 0.7020075282308658,
      "grad_norm": 1.821305513381958,
      "learning_rate": 4.1011384348183565e-06,
      "loss": 1.1811,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1119
    },
    {
      "epoch": 0.7026348808030113,
      "grad_norm": 1.656417965888977,
      "learning_rate": 4.099035795959798e-06,
      "loss": 1.2397,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1120
    },
    {
      "epoch": 0.7032622333751568,
      "grad_norm": 1.4519686698913574,
      "learning_rate": 4.096931241107461e-06,
      "loss": 1.2223,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1121
    },
    {
      "epoch": 0.7038895859473023,
      "grad_norm": 1.6120268106460571,
      "learning_rate": 4.094824772783062e-06,
      "loss": 1.2453,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1122
    },
    {
      "epoch": 0.704516938519448,
      "grad_norm": 1.7454333305358887,
      "learning_rate": 4.092716393510609e-06,
      "loss": 1.3674,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1123
    },
    {
      "epoch": 0.7051442910915935,
      "grad_norm": 1.7884482145309448,
      "learning_rate": 4.0906061058164e-06,
      "loss": 1.2876,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1124
    },
    {
      "epoch": 0.705771643663739,
      "grad_norm": 1.7549852132797241,
      "learning_rate": 4.0884939122290215e-06,
      "loss": 1.2451,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1125
    },
    {
      "epoch": 0.7063989962358845,
      "grad_norm": 1.5388882160186768,
      "learning_rate": 4.08637981527934e-06,
      "loss": 1.2364,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1126
    },
    {
      "epoch": 0.7070263488080301,
      "grad_norm": 1.7616428136825562,
      "learning_rate": 4.084263817500508e-06,
      "loss": 1.2736,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1127
    },
    {
      "epoch": 0.7076537013801757,
      "grad_norm": 1.576174259185791,
      "learning_rate": 4.082145921427949e-06,
      "loss": 1.2303,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1128
    },
    {
      "epoch": 0.7082810539523212,
      "grad_norm": 1.961186170578003,
      "learning_rate": 4.080026129599368e-06,
      "loss": 1.2455,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1129
    },
    {
      "epoch": 0.7089084065244667,
      "grad_norm": 1.8993115425109863,
      "learning_rate": 4.077904444554736e-06,
      "loss": 1.1639,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1130
    },
    {
      "epoch": 0.7095357590966123,
      "grad_norm": 1.7968101501464844,
      "learning_rate": 4.075780868836296e-06,
      "loss": 1.3066,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1131
    },
    {
      "epoch": 0.7101631116687579,
      "grad_norm": 2.5451998710632324,
      "learning_rate": 4.073655404988554e-06,
      "loss": 1.2309,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1132
    },
    {
      "epoch": 0.7107904642409034,
      "grad_norm": 1.7790062427520752,
      "learning_rate": 4.071528055558278e-06,
      "loss": 1.2954,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1133
    },
    {
      "epoch": 0.7114178168130489,
      "grad_norm": 1.8270624876022339,
      "learning_rate": 4.0693988230945e-06,
      "loss": 1.1384,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1134
    },
    {
      "epoch": 0.7120451693851945,
      "grad_norm": 1.8887920379638672,
      "learning_rate": 4.067267710148504e-06,
      "loss": 1.2272,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1135
    },
    {
      "epoch": 0.71267252195734,
      "grad_norm": 1.9454110860824585,
      "learning_rate": 4.065134719273828e-06,
      "loss": 1.3557,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1136
    },
    {
      "epoch": 0.7132998745294856,
      "grad_norm": 1.7404446601867676,
      "learning_rate": 4.062999853026259e-06,
      "loss": 1.2289,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1137
    },
    {
      "epoch": 0.7139272271016311,
      "grad_norm": 1.4533557891845703,
      "learning_rate": 4.060863113963835e-06,
      "loss": 1.2886,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1138
    },
    {
      "epoch": 0.7145545796737767,
      "grad_norm": 1.7258976697921753,
      "learning_rate": 4.058724504646834e-06,
      "loss": 1.2859,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1139
    },
    {
      "epoch": 0.7151819322459222,
      "grad_norm": 1.8206146955490112,
      "learning_rate": 4.056584027637778e-06,
      "loss": 1.3065,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1140
    },
    {
      "epoch": 0.7158092848180677,
      "grad_norm": 1.9629409313201904,
      "learning_rate": 4.054441685501423e-06,
      "loss": 1.3129,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1141
    },
    {
      "epoch": 0.7164366373902133,
      "grad_norm": 1.4002165794372559,
      "learning_rate": 4.0522974808047655e-06,
      "loss": 1.247,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1142
    },
    {
      "epoch": 0.7170639899623589,
      "grad_norm": 2.054499387741089,
      "learning_rate": 4.050151416117028e-06,
      "loss": 1.3253,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1143
    },
    {
      "epoch": 0.7176913425345044,
      "grad_norm": 1.6483572721481323,
      "learning_rate": 4.048003494009666e-06,
      "loss": 1.1892,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1144
    },
    {
      "epoch": 0.7183186951066499,
      "grad_norm": 1.6913396120071411,
      "learning_rate": 4.045853717056358e-06,
      "loss": 1.2009,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1145
    },
    {
      "epoch": 0.7189460476787954,
      "grad_norm": 1.7624170780181885,
      "learning_rate": 4.043702087833006e-06,
      "loss": 1.2018,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1146
    },
    {
      "epoch": 0.7195734002509411,
      "grad_norm": 2.329902410507202,
      "learning_rate": 4.04154860891773e-06,
      "loss": 1.161,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1147
    },
    {
      "epoch": 0.7202007528230866,
      "grad_norm": 1.9451358318328857,
      "learning_rate": 4.03939328289087e-06,
      "loss": 1.2845,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1148
    },
    {
      "epoch": 0.7208281053952321,
      "grad_norm": 2.0189058780670166,
      "learning_rate": 4.037236112334976e-06,
      "loss": 1.2569,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1149
    },
    {
      "epoch": 0.7214554579673776,
      "grad_norm": 1.3586088418960571,
      "learning_rate": 4.0350770998348075e-06,
      "loss": 1.1772,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1150
    },
    {
      "epoch": 0.7220828105395232,
      "grad_norm": 1.76394784450531,
      "learning_rate": 4.032916247977334e-06,
      "loss": 1.2178,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1151
    },
    {
      "epoch": 0.7227101631116688,
      "grad_norm": 1.484285593032837,
      "learning_rate": 4.030753559351728e-06,
      "loss": 1.1872,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1152
    },
    {
      "epoch": 0.7233375156838143,
      "grad_norm": 1.5376477241516113,
      "learning_rate": 4.028589036549361e-06,
      "loss": 1.2631,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1153
    },
    {
      "epoch": 0.7239648682559598,
      "grad_norm": 2.0866265296936035,
      "learning_rate": 4.026422682163804e-06,
      "loss": 1.2962,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1154
    },
    {
      "epoch": 0.7245922208281054,
      "grad_norm": 1.7734241485595703,
      "learning_rate": 4.0242544987908235e-06,
      "loss": 1.3571,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1155
    },
    {
      "epoch": 0.7252195734002509,
      "grad_norm": 1.6300272941589355,
      "learning_rate": 4.022084489028375e-06,
      "loss": 1.2431,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1156
    },
    {
      "epoch": 0.7258469259723965,
      "grad_norm": 1.834957480430603,
      "learning_rate": 4.019912655476603e-06,
      "loss": 1.3553,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1157
    },
    {
      "epoch": 0.726474278544542,
      "grad_norm": 1.8868942260742188,
      "learning_rate": 4.017739000737839e-06,
      "loss": 1.2373,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1158
    },
    {
      "epoch": 0.7271016311166876,
      "grad_norm": 1.9034756422042847,
      "learning_rate": 4.015563527416596e-06,
      "loss": 1.3091,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1159
    },
    {
      "epoch": 0.7277289836888331,
      "grad_norm": 1.6866358518600464,
      "learning_rate": 4.0133862381195645e-06,
      "loss": 1.1822,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1160
    },
    {
      "epoch": 0.7283563362609786,
      "grad_norm": 1.3602663278579712,
      "learning_rate": 4.011207135455612e-06,
      "loss": 1.1083,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1161
    },
    {
      "epoch": 0.7289836888331243,
      "grad_norm": 1.928031325340271,
      "learning_rate": 4.009026222035782e-06,
      "loss": 1.2096,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1162
    },
    {
      "epoch": 0.7296110414052698,
      "grad_norm": 1.8566958904266357,
      "learning_rate": 4.006843500473281e-06,
      "loss": 1.3159,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1163
    },
    {
      "epoch": 0.7302383939774153,
      "grad_norm": 1.9040616750717163,
      "learning_rate": 4.0046589733834875e-06,
      "loss": 1.3272,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1164
    },
    {
      "epoch": 0.7308657465495608,
      "grad_norm": 1.490311861038208,
      "learning_rate": 4.002472643383941e-06,
      "loss": 1.3552,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1165
    },
    {
      "epoch": 0.7314930991217063,
      "grad_norm": 1.6385480165481567,
      "learning_rate": 4.000284513094342e-06,
      "loss": 1.2069,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1166
    },
    {
      "epoch": 0.732120451693852,
      "grad_norm": 2.097144603729248,
      "learning_rate": 3.9980945851365485e-06,
      "loss": 1.2088,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1167
    },
    {
      "epoch": 0.7327478042659975,
      "grad_norm": 1.6262550354003906,
      "learning_rate": 3.99590286213457e-06,
      "loss": 1.4293,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1168
    },
    {
      "epoch": 0.733375156838143,
      "grad_norm": 1.6071372032165527,
      "learning_rate": 3.9937093467145725e-06,
      "loss": 1.0753,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1169
    },
    {
      "epoch": 0.7340025094102886,
      "grad_norm": 1.8095484972000122,
      "learning_rate": 3.991514041504863e-06,
      "loss": 1.3106,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1170
    },
    {
      "epoch": 0.7346298619824341,
      "grad_norm": 1.8286887407302856,
      "learning_rate": 3.989316949135898e-06,
      "loss": 1.2928,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1171
    },
    {
      "epoch": 0.7352572145545797,
      "grad_norm": 1.8071660995483398,
      "learning_rate": 3.987118072240272e-06,
      "loss": 1.2948,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1172
    },
    {
      "epoch": 0.7358845671267252,
      "grad_norm": 1.8113529682159424,
      "learning_rate": 3.984917413452721e-06,
      "loss": 1.3549,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1173
    },
    {
      "epoch": 0.7365119196988708,
      "grad_norm": 1.5373753309249878,
      "learning_rate": 3.982714975410111e-06,
      "loss": 1.1386,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1174
    },
    {
      "epoch": 0.7371392722710163,
      "grad_norm": 1.5793862342834473,
      "learning_rate": 3.980510760751447e-06,
      "loss": 1.2817,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1175
    },
    {
      "epoch": 0.7377666248431619,
      "grad_norm": 1.647489070892334,
      "learning_rate": 3.978304772117859e-06,
      "loss": 1.2604,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1176
    },
    {
      "epoch": 0.7383939774153074,
      "grad_norm": 1.910098910331726,
      "learning_rate": 3.9760970121526e-06,
      "loss": 1.2902,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1177
    },
    {
      "epoch": 0.739021329987453,
      "grad_norm": 2.1126773357391357,
      "learning_rate": 3.973887483501051e-06,
      "loss": 1.2406,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1178
    },
    {
      "epoch": 0.7396486825595985,
      "grad_norm": 1.8298122882843018,
      "learning_rate": 3.971676188810707e-06,
      "loss": 1.2546,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1179
    },
    {
      "epoch": 0.740276035131744,
      "grad_norm": 1.6926147937774658,
      "learning_rate": 3.969463130731183e-06,
      "loss": 1.2602,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1180
    },
    {
      "epoch": 0.7409033877038896,
      "grad_norm": 2.075894832611084,
      "learning_rate": 3.9672483119142055e-06,
      "loss": 1.2015,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1181
    },
    {
      "epoch": 0.7415307402760352,
      "grad_norm": 1.5787642002105713,
      "learning_rate": 3.96503173501361e-06,
      "loss": 1.2999,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1182
    },
    {
      "epoch": 0.7421580928481807,
      "grad_norm": 1.943341612815857,
      "learning_rate": 3.962813402685339e-06,
      "loss": 1.2898,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1183
    },
    {
      "epoch": 0.7427854454203262,
      "grad_norm": 1.6585906744003296,
      "learning_rate": 3.96059331758744e-06,
      "loss": 1.3,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1184
    },
    {
      "epoch": 0.7434127979924717,
      "grad_norm": 1.8577035665512085,
      "learning_rate": 3.9583714823800575e-06,
      "loss": 1.3885,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1185
    },
    {
      "epoch": 0.7440401505646174,
      "grad_norm": 1.5204012393951416,
      "learning_rate": 3.9561478997254375e-06,
      "loss": 1.1612,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1186
    },
    {
      "epoch": 0.7446675031367629,
      "grad_norm": 1.1847776174545288,
      "learning_rate": 3.953922572287915e-06,
      "loss": 1.1716,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1187
    },
    {
      "epoch": 0.7452948557089084,
      "grad_norm": 2.699477195739746,
      "learning_rate": 3.951695502733917e-06,
      "loss": 1.285,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1188
    },
    {
      "epoch": 0.7459222082810539,
      "grad_norm": 2.0467886924743652,
      "learning_rate": 3.949466693731962e-06,
      "loss": 1.3222,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1189
    },
    {
      "epoch": 0.7465495608531995,
      "grad_norm": 1.7042464017868042,
      "learning_rate": 3.947236147952647e-06,
      "loss": 1.1682,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1190
    },
    {
      "epoch": 0.7471769134253451,
      "grad_norm": 1.9766619205474854,
      "learning_rate": 3.945003868068653e-06,
      "loss": 1.3811,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1191
    },
    {
      "epoch": 0.7478042659974906,
      "grad_norm": 1.95023512840271,
      "learning_rate": 3.942769856754739e-06,
      "loss": 1.0653,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1192
    },
    {
      "epoch": 0.7484316185696361,
      "grad_norm": 2.0074973106384277,
      "learning_rate": 3.940534116687737e-06,
      "loss": 1.3036,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1193
    },
    {
      "epoch": 0.7490589711417817,
      "grad_norm": 1.8049767017364502,
      "learning_rate": 3.938296650546552e-06,
      "loss": 1.4716,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1194
    },
    {
      "epoch": 0.7496863237139272,
      "grad_norm": 1.9003103971481323,
      "learning_rate": 3.936057461012157e-06,
      "loss": 1.3468,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1195
    },
    {
      "epoch": 0.7503136762860728,
      "grad_norm": 1.6638295650482178,
      "learning_rate": 3.9338165507675885e-06,
      "loss": 1.2144,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1196
    },
    {
      "epoch": 0.7509410288582183,
      "grad_norm": 1.6862947940826416,
      "learning_rate": 3.931573922497947e-06,
      "loss": 1.2948,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1197
    },
    {
      "epoch": 0.7515683814303639,
      "grad_norm": 1.7455106973648071,
      "learning_rate": 3.92932957889039e-06,
      "loss": 1.388,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1198
    },
    {
      "epoch": 0.7521957340025094,
      "grad_norm": 1.7992185354232788,
      "learning_rate": 3.927083522634132e-06,
      "loss": 1.1686,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1199
    },
    {
      "epoch": 0.7528230865746549,
      "grad_norm": 2.1805531978607178,
      "learning_rate": 3.9248357564204364e-06,
      "loss": 1.3465,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1200
    },
    {
      "epoch": 0.7534504391468005,
      "grad_norm": 1.8067539930343628,
      "learning_rate": 3.9225862829426184e-06,
      "loss": 1.3312,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1201
    },
    {
      "epoch": 0.7540777917189461,
      "grad_norm": 1.7938944101333618,
      "learning_rate": 3.920335104896039e-06,
      "loss": 1.2371,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1202
    },
    {
      "epoch": 0.7547051442910916,
      "grad_norm": 1.7084840536117554,
      "learning_rate": 3.918082224978099e-06,
      "loss": 1.3185,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1203
    },
    {
      "epoch": 0.7553324968632371,
      "grad_norm": 1.7908164262771606,
      "learning_rate": 3.915827645888242e-06,
      "loss": 1.2634,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1204
    },
    {
      "epoch": 0.7559598494353826,
      "grad_norm": 1.709354281425476,
      "learning_rate": 3.913571370327944e-06,
      "loss": 1.1975,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1205
    },
    {
      "epoch": 0.7565872020075283,
      "grad_norm": 1.792067050933838,
      "learning_rate": 3.911313401000716e-06,
      "loss": 1.3193,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1206
    },
    {
      "epoch": 0.7572145545796738,
      "grad_norm": 1.76042640209198,
      "learning_rate": 3.909053740612099e-06,
      "loss": 1.226,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1207
    },
    {
      "epoch": 0.7578419071518193,
      "grad_norm": 1.750500202178955,
      "learning_rate": 3.906792391869657e-06,
      "loss": 1.4268,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1208
    },
    {
      "epoch": 0.7584692597239648,
      "grad_norm": 1.4998160600662231,
      "learning_rate": 3.904529357482981e-06,
      "loss": 1.1727,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1209
    },
    {
      "epoch": 0.7590966122961104,
      "grad_norm": 2.04630970954895,
      "learning_rate": 3.90226464016368e-06,
      "loss": 1.373,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1210
    },
    {
      "epoch": 0.759723964868256,
      "grad_norm": 2.0300686359405518,
      "learning_rate": 3.899998242625378e-06,
      "loss": 1.1817,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1211
    },
    {
      "epoch": 0.7603513174404015,
      "grad_norm": 1.8236775398254395,
      "learning_rate": 3.897730167583716e-06,
      "loss": 1.2345,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1212
    },
    {
      "epoch": 0.760978670012547,
      "grad_norm": 2.1852262020111084,
      "learning_rate": 3.8954604177563395e-06,
      "loss": 1.3007,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1213
    },
    {
      "epoch": 0.7616060225846926,
      "grad_norm": 1.4265241622924805,
      "learning_rate": 3.893188995862907e-06,
      "loss": 1.0932,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1214
    },
    {
      "epoch": 0.7622333751568381,
      "grad_norm": 1.7335155010223389,
      "learning_rate": 3.890915904625075e-06,
      "loss": 1.3525,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1215
    },
    {
      "epoch": 0.7628607277289837,
      "grad_norm": 1.8855781555175781,
      "learning_rate": 3.888641146766506e-06,
      "loss": 1.2274,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1216
    },
    {
      "epoch": 0.7634880803011292,
      "grad_norm": 1.8744924068450928,
      "learning_rate": 3.886364725012854e-06,
      "loss": 1.2034,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1217
    },
    {
      "epoch": 0.7641154328732748,
      "grad_norm": 1.7175720930099487,
      "learning_rate": 3.884086642091769e-06,
      "loss": 1.2869,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1218
    },
    {
      "epoch": 0.7647427854454203,
      "grad_norm": 1.662237286567688,
      "learning_rate": 3.881806900732893e-06,
      "loss": 1.1438,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1219
    },
    {
      "epoch": 0.7653701380175659,
      "grad_norm": 1.7514503002166748,
      "learning_rate": 3.879525503667851e-06,
      "loss": 1.1766,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1220
    },
    {
      "epoch": 0.7659974905897114,
      "grad_norm": 1.7529418468475342,
      "learning_rate": 3.8772424536302565e-06,
      "loss": 1.2273,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1221
    },
    {
      "epoch": 0.766624843161857,
      "grad_norm": 1.989951252937317,
      "learning_rate": 3.874957753355701e-06,
      "loss": 1.1705,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1222
    },
    {
      "epoch": 0.7672521957340025,
      "grad_norm": 1.5698517560958862,
      "learning_rate": 3.872671405581754e-06,
      "loss": 1.2493,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1223
    },
    {
      "epoch": 0.767879548306148,
      "grad_norm": 1.5441415309906006,
      "learning_rate": 3.870383413047959e-06,
      "loss": 1.1989,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1224
    },
    {
      "epoch": 0.7685069008782937,
      "grad_norm": 1.7378580570220947,
      "learning_rate": 3.86809377849583e-06,
      "loss": 1.1907,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1225
    },
    {
      "epoch": 0.7691342534504392,
      "grad_norm": 1.8756630420684814,
      "learning_rate": 3.865802504668849e-06,
      "loss": 1.2261,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1226
    },
    {
      "epoch": 0.7697616060225847,
      "grad_norm": 1.6204801797866821,
      "learning_rate": 3.863509594312461e-06,
      "loss": 1.163,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1227
    },
    {
      "epoch": 0.7703889585947302,
      "grad_norm": 1.9416941404342651,
      "learning_rate": 3.861215050174074e-06,
      "loss": 1.2847,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1228
    },
    {
      "epoch": 0.7710163111668757,
      "grad_norm": 1.780923843383789,
      "learning_rate": 3.858918875003053e-06,
      "loss": 1.4637,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1229
    },
    {
      "epoch": 0.7716436637390214,
      "grad_norm": 1.8576253652572632,
      "learning_rate": 3.856621071550716e-06,
      "loss": 1.1524,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1230
    },
    {
      "epoch": 0.7722710163111669,
      "grad_norm": 2.069793939590454,
      "learning_rate": 3.854321642570334e-06,
      "loss": 1.4231,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1231
    },
    {
      "epoch": 0.7728983688833124,
      "grad_norm": 1.6411324739456177,
      "learning_rate": 3.852020590817123e-06,
      "loss": 1.2096,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1232
    },
    {
      "epoch": 0.773525721455458,
      "grad_norm": 1.8884642124176025,
      "learning_rate": 3.8497179190482445e-06,
      "loss": 1.3151,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1233
    },
    {
      "epoch": 0.7741530740276035,
      "grad_norm": 2.067143678665161,
      "learning_rate": 3.847413630022804e-06,
      "loss": 1.2766,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1234
    },
    {
      "epoch": 0.7747804265997491,
      "grad_norm": 1.7415361404418945,
      "learning_rate": 3.8451077265018426e-06,
      "loss": 1.285,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1235
    },
    {
      "epoch": 0.7754077791718946,
      "grad_norm": 1.9397696256637573,
      "learning_rate": 3.842800211248333e-06,
      "loss": 1.299,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1236
    },
    {
      "epoch": 0.7760351317440402,
      "grad_norm": 2.149726390838623,
      "learning_rate": 3.840491087027184e-06,
      "loss": 1.2434,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1237
    },
    {
      "epoch": 0.7766624843161857,
      "grad_norm": 1.3199256658554077,
      "learning_rate": 3.83818035660523e-06,
      "loss": 1.2262,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1238
    },
    {
      "epoch": 0.7772898368883312,
      "grad_norm": 1.7844839096069336,
      "learning_rate": 3.835868022751231e-06,
      "loss": 1.146,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1239
    },
    {
      "epoch": 0.7779171894604768,
      "grad_norm": 3.71295166015625,
      "learning_rate": 3.833554088235866e-06,
      "loss": 1.3146,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1240
    },
    {
      "epoch": 0.7785445420326224,
      "grad_norm": 1.7507476806640625,
      "learning_rate": 3.831238555831735e-06,
      "loss": 1.1682,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1241
    },
    {
      "epoch": 0.7791718946047679,
      "grad_norm": 1.502406120300293,
      "learning_rate": 3.828921428313349e-06,
      "loss": 1.2024,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1242
    },
    {
      "epoch": 0.7797992471769134,
      "grad_norm": 1.2211238145828247,
      "learning_rate": 3.8266027084571335e-06,
      "loss": 1.253,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1243
    },
    {
      "epoch": 0.7804265997490589,
      "grad_norm": 1.661982774734497,
      "learning_rate": 3.824282399041421e-06,
      "loss": 1.3408,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1244
    },
    {
      "epoch": 0.7810539523212046,
      "grad_norm": 1.6972383260726929,
      "learning_rate": 3.821960502846449e-06,
      "loss": 1.3019,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1245
    },
    {
      "epoch": 0.7816813048933501,
      "grad_norm": 1.9191818237304688,
      "learning_rate": 3.819637022654355e-06,
      "loss": 1.2167,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1246
    },
    {
      "epoch": 0.7823086574654956,
      "grad_norm": 1.999821424484253,
      "learning_rate": 3.817311961249175e-06,
      "loss": 1.3381,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1247
    },
    {
      "epoch": 0.7829360100376411,
      "grad_norm": 1.9394023418426514,
      "learning_rate": 3.814985321416839e-06,
      "loss": 1.271,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1248
    },
    {
      "epoch": 0.7835633626097867,
      "grad_norm": 1.752686619758606,
      "learning_rate": 3.812657105945171e-06,
      "loss": 1.1284,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1249
    },
    {
      "epoch": 0.7841907151819323,
      "grad_norm": 1.853484034538269,
      "learning_rate": 3.810327317623881e-06,
      "loss": 1.1259,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1250
    },
    {
      "epoch": 0.7848180677540778,
      "grad_norm": 1.6975935697555542,
      "learning_rate": 3.8079959592445626e-06,
      "loss": 1.2346,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1251
    },
    {
      "epoch": 0.7854454203262233,
      "grad_norm": 1.5964429378509521,
      "learning_rate": 3.8056630336006915e-06,
      "loss": 1.2542,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1252
    },
    {
      "epoch": 0.7860727728983689,
      "grad_norm": 1.616788387298584,
      "learning_rate": 3.8033285434876245e-06,
      "loss": 1.3044,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1253
    },
    {
      "epoch": 0.7867001254705144,
      "grad_norm": 1.6051348447799683,
      "learning_rate": 3.8009924917025864e-06,
      "loss": 1.1955,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1254
    },
    {
      "epoch": 0.78732747804266,
      "grad_norm": 1.761275053024292,
      "learning_rate": 3.7986548810446796e-06,
      "loss": 1.2944,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1255
    },
    {
      "epoch": 0.7879548306148055,
      "grad_norm": 1.8812828063964844,
      "learning_rate": 3.7963157143148705e-06,
      "loss": 1.1681,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1256
    },
    {
      "epoch": 0.7885821831869511,
      "grad_norm": 1.7038171291351318,
      "learning_rate": 3.793974994315991e-06,
      "loss": 1.3871,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1257
    },
    {
      "epoch": 0.7892095357590966,
      "grad_norm": 2.0471980571746826,
      "learning_rate": 3.791632723852736e-06,
      "loss": 1.288,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1258
    },
    {
      "epoch": 0.7898368883312421,
      "grad_norm": 1.5858546495437622,
      "learning_rate": 3.789288905731655e-06,
      "loss": 1.2338,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1259
    },
    {
      "epoch": 0.7904642409033877,
      "grad_norm": 1.7153313159942627,
      "learning_rate": 3.7869435427611547e-06,
      "loss": 1.299,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1260
    },
    {
      "epoch": 0.7910915934755333,
      "grad_norm": 1.87894606590271,
      "learning_rate": 3.78459663775149e-06,
      "loss": 1.2089,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1261
    },
    {
      "epoch": 0.7917189460476788,
      "grad_norm": 2.0329928398132324,
      "learning_rate": 3.782248193514766e-06,
      "loss": 1.3064,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1262
    },
    {
      "epoch": 0.7923462986198243,
      "grad_norm": 2.049156427383423,
      "learning_rate": 3.7798982128649315e-06,
      "loss": 1.1991,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1263
    },
    {
      "epoch": 0.7929736511919699,
      "grad_norm": 1.8942718505859375,
      "learning_rate": 3.7775466986177763e-06,
      "loss": 1.2944,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1264
    },
    {
      "epoch": 0.7936010037641155,
      "grad_norm": 1.7939801216125488,
      "learning_rate": 3.775193653590927e-06,
      "loss": 1.3067,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1265
    },
    {
      "epoch": 0.794228356336261,
      "grad_norm": 1.5614609718322754,
      "learning_rate": 3.7728390806038456e-06,
      "loss": 1.1798,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1266
    },
    {
      "epoch": 0.7948557089084065,
      "grad_norm": 1.604327917098999,
      "learning_rate": 3.7704829824778245e-06,
      "loss": 1.166,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1267
    },
    {
      "epoch": 0.795483061480552,
      "grad_norm": 1.6718720197677612,
      "learning_rate": 3.7681253620359814e-06,
      "loss": 1.1947,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1268
    },
    {
      "epoch": 0.7961104140526977,
      "grad_norm": 1.7396267652511597,
      "learning_rate": 3.765766222103262e-06,
      "loss": 1.2189,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1269
    },
    {
      "epoch": 0.7967377666248432,
      "grad_norm": 1.5587661266326904,
      "learning_rate": 3.7634055655064296e-06,
      "loss": 1.3867,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1270
    },
    {
      "epoch": 0.7973651191969887,
      "grad_norm": 2.117201328277588,
      "learning_rate": 3.7610433950740667e-06,
      "loss": 1.2159,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1271
    },
    {
      "epoch": 0.7979924717691342,
      "grad_norm": 1.485001564025879,
      "learning_rate": 3.7586797136365682e-06,
      "loss": 1.278,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1272
    },
    {
      "epoch": 0.7986198243412798,
      "grad_norm": 1.8766347169876099,
      "learning_rate": 3.756314524026141e-06,
      "loss": 1.3496,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1273
    },
    {
      "epoch": 0.7992471769134254,
      "grad_norm": 1.7875092029571533,
      "learning_rate": 3.753947829076797e-06,
      "loss": 1.2092,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1274
    },
    {
      "epoch": 0.7998745294855709,
      "grad_norm": 1.8150584697723389,
      "learning_rate": 3.751579631624355e-06,
      "loss": 1.217,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1275
    },
    {
      "epoch": 0.8005018820577164,
      "grad_norm": 1.8528504371643066,
      "learning_rate": 3.74920993450643e-06,
      "loss": 1.3494,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1276
    },
    {
      "epoch": 0.801129234629862,
      "grad_norm": 1.870395541191101,
      "learning_rate": 3.7468387405624373e-06,
      "loss": 1.1856,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1277
    },
    {
      "epoch": 0.8017565872020075,
      "grad_norm": 1.8406163454055786,
      "learning_rate": 3.7444660526335853e-06,
      "loss": 1.1518,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1278
    },
    {
      "epoch": 0.8023839397741531,
      "grad_norm": 1.3434467315673828,
      "learning_rate": 3.7420918735628714e-06,
      "loss": 1.1693,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1279
    },
    {
      "epoch": 0.8030112923462986,
      "grad_norm": 1.5686734914779663,
      "learning_rate": 3.73971620619508e-06,
      "loss": 1.2969,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1280
    },
    {
      "epoch": 0.8036386449184442,
      "grad_norm": 1.5028196573257446,
      "learning_rate": 3.7373390533767783e-06,
      "loss": 1.2124,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1281
    },
    {
      "epoch": 0.8042659974905897,
      "grad_norm": 1.6024171113967896,
      "learning_rate": 3.7349604179563156e-06,
      "loss": 1.2056,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1282
    },
    {
      "epoch": 0.8048933500627352,
      "grad_norm": 1.773492455482483,
      "learning_rate": 3.7325803027838146e-06,
      "loss": 1.243,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1283
    },
    {
      "epoch": 0.8055207026348808,
      "grad_norm": 1.4282128810882568,
      "learning_rate": 3.730198710711173e-06,
      "loss": 1.2822,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1284
    },
    {
      "epoch": 0.8061480552070264,
      "grad_norm": 1.796323537826538,
      "learning_rate": 3.7278156445920584e-06,
      "loss": 1.3409,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1285
    },
    {
      "epoch": 0.8067754077791719,
      "grad_norm": 2.14421010017395,
      "learning_rate": 3.725431107281904e-06,
      "loss": 1.267,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1286
    },
    {
      "epoch": 0.8074027603513174,
      "grad_norm": 1.5279842615127563,
      "learning_rate": 3.7230451016379045e-06,
      "loss": 1.1884,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1287
    },
    {
      "epoch": 0.8080301129234629,
      "grad_norm": 1.8912734985351562,
      "learning_rate": 3.720657630519016e-06,
      "loss": 1.2774,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1288
    },
    {
      "epoch": 0.8086574654956086,
      "grad_norm": 1.9765260219573975,
      "learning_rate": 3.71826869678595e-06,
      "loss": 1.3237,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1289
    },
    {
      "epoch": 0.8092848180677541,
      "grad_norm": 1.8362704515457153,
      "learning_rate": 3.7158783033011702e-06,
      "loss": 1.1868,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1290
    },
    {
      "epoch": 0.8099121706398996,
      "grad_norm": 1.8150081634521484,
      "learning_rate": 3.7134864529288907e-06,
      "loss": 1.3989,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1291
    },
    {
      "epoch": 0.8105395232120451,
      "grad_norm": 1.8795225620269775,
      "learning_rate": 3.711093148535068e-06,
      "loss": 1.3224,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1292
    },
    {
      "epoch": 0.8111668757841907,
      "grad_norm": 2.3372652530670166,
      "learning_rate": 3.7086983929874044e-06,
      "loss": 1.3099,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1293
    },
    {
      "epoch": 0.8117942283563363,
      "grad_norm": 1.8464175462722778,
      "learning_rate": 3.7063021891553384e-06,
      "loss": 1.1561,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1294
    },
    {
      "epoch": 0.8124215809284818,
      "grad_norm": 1.796033263206482,
      "learning_rate": 3.703904539910047e-06,
      "loss": 1.4166,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1295
    },
    {
      "epoch": 0.8130489335006273,
      "grad_norm": 2.128056049346924,
      "learning_rate": 3.701505448124435e-06,
      "loss": 1.3366,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1296
    },
    {
      "epoch": 0.8136762860727729,
      "grad_norm": 2.16898512840271,
      "learning_rate": 3.6991049166731387e-06,
      "loss": 1.3019,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1297
    },
    {
      "epoch": 0.8143036386449184,
      "grad_norm": 1.7204794883728027,
      "learning_rate": 3.696702948432519e-06,
      "loss": 1.1802,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1298
    },
    {
      "epoch": 0.814930991217064,
      "grad_norm": 1.7379305362701416,
      "learning_rate": 3.6942995462806574e-06,
      "loss": 1.1881,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1299
    },
    {
      "epoch": 0.8155583437892095,
      "grad_norm": 2.0231478214263916,
      "learning_rate": 3.691894713097355e-06,
      "loss": 1.3784,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1300
    },
    {
      "epoch": 0.8161856963613551,
      "grad_norm": 1.7545968294143677,
      "learning_rate": 3.689488451764125e-06,
      "loss": 1.3519,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1301
    },
    {
      "epoch": 0.8168130489335006,
      "grad_norm": 1.6912481784820557,
      "learning_rate": 3.687080765164194e-06,
      "loss": 1.3577,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1302
    },
    {
      "epoch": 0.8174404015056461,
      "grad_norm": 1.7708526849746704,
      "learning_rate": 3.684671656182497e-06,
      "loss": 1.2427,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1303
    },
    {
      "epoch": 0.8180677540777918,
      "grad_norm": 1.7793676853179932,
      "learning_rate": 3.682261127705671e-06,
      "loss": 1.3287,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1304
    },
    {
      "epoch": 0.8186951066499373,
      "grad_norm": 1.8184021711349487,
      "learning_rate": 3.679849182622056e-06,
      "loss": 1.3435,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1305
    },
    {
      "epoch": 0.8193224592220828,
      "grad_norm": 1.8825409412384033,
      "learning_rate": 3.6774358238216878e-06,
      "loss": 1.3772,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1306
    },
    {
      "epoch": 0.8199498117942283,
      "grad_norm": 1.8324886560440063,
      "learning_rate": 3.6750210541962972e-06,
      "loss": 1.3492,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1307
    },
    {
      "epoch": 0.820577164366374,
      "grad_norm": 1.7730599641799927,
      "learning_rate": 3.6726048766393046e-06,
      "loss": 1.3669,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1308
    },
    {
      "epoch": 0.8212045169385195,
      "grad_norm": 1.756497859954834,
      "learning_rate": 3.670187294045819e-06,
      "loss": 1.4286,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1309
    },
    {
      "epoch": 0.821831869510665,
      "grad_norm": 1.1420440673828125,
      "learning_rate": 3.6677683093126297e-06,
      "loss": 1.1475,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1310
    },
    {
      "epoch": 0.8224592220828105,
      "grad_norm": 1.56114661693573,
      "learning_rate": 3.665347925338211e-06,
      "loss": 1.333,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1311
    },
    {
      "epoch": 0.823086574654956,
      "grad_norm": 1.8046437501907349,
      "learning_rate": 3.662926145022708e-06,
      "loss": 1.2138,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1312
    },
    {
      "epoch": 0.8237139272271017,
      "grad_norm": 1.684722900390625,
      "learning_rate": 3.660502971267945e-06,
      "loss": 1.254,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1313
    },
    {
      "epoch": 0.8243412797992472,
      "grad_norm": 1.9925870895385742,
      "learning_rate": 3.6580784069774104e-06,
      "loss": 1.2882,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1314
    },
    {
      "epoch": 0.8249686323713927,
      "grad_norm": 1.7439500093460083,
      "learning_rate": 3.655652455056263e-06,
      "loss": 1.1997,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1315
    },
    {
      "epoch": 0.8255959849435383,
      "grad_norm": 1.1201549768447876,
      "learning_rate": 3.653225118411321e-06,
      "loss": 1.2117,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1316
    },
    {
      "epoch": 0.8262233375156838,
      "grad_norm": 1.7333301305770874,
      "learning_rate": 3.6507963999510647e-06,
      "loss": 1.2691,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1317
    },
    {
      "epoch": 0.8268506900878294,
      "grad_norm": 1.7772564888000488,
      "learning_rate": 3.6483663025856276e-06,
      "loss": 1.3077,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1318
    },
    {
      "epoch": 0.8274780426599749,
      "grad_norm": 1.8323562145233154,
      "learning_rate": 3.645934829226797e-06,
      "loss": 1.3552,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1319
    },
    {
      "epoch": 0.8281053952321205,
      "grad_norm": 1.7378582954406738,
      "learning_rate": 3.6435019827880093e-06,
      "loss": 1.2976,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1320
    },
    {
      "epoch": 0.828732747804266,
      "grad_norm": 1.983419418334961,
      "learning_rate": 3.641067766184344e-06,
      "loss": 1.3097,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1321
    },
    {
      "epoch": 0.8293601003764115,
      "grad_norm": 2.4208033084869385,
      "learning_rate": 3.6386321823325242e-06,
      "loss": 1.1557,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1322
    },
    {
      "epoch": 0.8299874529485571,
      "grad_norm": 1.7591361999511719,
      "learning_rate": 3.636195234150911e-06,
      "loss": 1.3991,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1323
    },
    {
      "epoch": 0.8306148055207027,
      "grad_norm": 1.894167423248291,
      "learning_rate": 3.6337569245595007e-06,
      "loss": 1.2257,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1324
    },
    {
      "epoch": 0.8312421580928482,
      "grad_norm": 1.8774477243423462,
      "learning_rate": 3.6313172564799193e-06,
      "loss": 1.2091,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1325
    },
    {
      "epoch": 0.8318695106649937,
      "grad_norm": 1.8875328302383423,
      "learning_rate": 3.628876232835421e-06,
      "loss": 1.3331,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1326
    },
    {
      "epoch": 0.8324968632371392,
      "grad_norm": 1.873525857925415,
      "learning_rate": 3.626433856550886e-06,
      "loss": 1.2277,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1327
    },
    {
      "epoch": 0.8331242158092849,
      "grad_norm": 1.8272095918655396,
      "learning_rate": 3.623990130552813e-06,
      "loss": 1.3521,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1328
    },
    {
      "epoch": 0.8337515683814304,
      "grad_norm": 2.8294026851654053,
      "learning_rate": 3.6215450577693196e-06,
      "loss": 1.2214,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1329
    },
    {
      "epoch": 0.8343789209535759,
      "grad_norm": 1.6993393898010254,
      "learning_rate": 3.6190986411301354e-06,
      "loss": 1.1745,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1330
    },
    {
      "epoch": 0.8350062735257214,
      "grad_norm": 1.560639500617981,
      "learning_rate": 3.6166508835666026e-06,
      "loss": 1.2045,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1331
    },
    {
      "epoch": 0.835633626097867,
      "grad_norm": 1.559907078742981,
      "learning_rate": 3.6142017880116685e-06,
      "loss": 1.1456,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1332
    },
    {
      "epoch": 0.8362609786700126,
      "grad_norm": 1.6818002462387085,
      "learning_rate": 3.611751357399884e-06,
      "loss": 1.1618,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1333
    },
    {
      "epoch": 0.8368883312421581,
      "grad_norm": 1.8955743312835693,
      "learning_rate": 3.6092995946673996e-06,
      "loss": 1.291,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1334
    },
    {
      "epoch": 0.8375156838143036,
      "grad_norm": 1.8564447164535522,
      "learning_rate": 3.606846502751962e-06,
      "loss": 1.3731,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1335
    },
    {
      "epoch": 0.8381430363864492,
      "grad_norm": 1.6093361377716064,
      "learning_rate": 3.6043920845929093e-06,
      "loss": 1.1482,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1336
    },
    {
      "epoch": 0.8387703889585947,
      "grad_norm": 1.5696303844451904,
      "learning_rate": 3.6019363431311715e-06,
      "loss": 1.2315,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1337
    },
    {
      "epoch": 0.8393977415307403,
      "grad_norm": 1.8272252082824707,
      "learning_rate": 3.599479281309263e-06,
      "loss": 1.2891,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1338
    },
    {
      "epoch": 0.8400250941028858,
      "grad_norm": 1.5957363843917847,
      "learning_rate": 3.597020902071278e-06,
      "loss": 1.2716,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1339
    },
    {
      "epoch": 0.8406524466750314,
      "grad_norm": 2.0031421184539795,
      "learning_rate": 3.594561208362894e-06,
      "loss": 1.1314,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1340
    },
    {
      "epoch": 0.8412797992471769,
      "grad_norm": 1.8888684511184692,
      "learning_rate": 3.5921002031313586e-06,
      "loss": 1.3124,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1341
    },
    {
      "epoch": 0.8419071518193224,
      "grad_norm": 1.83388352394104,
      "learning_rate": 3.5896378893254936e-06,
      "loss": 1.2654,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1342
    },
    {
      "epoch": 0.842534504391468,
      "grad_norm": 1.7649545669555664,
      "learning_rate": 3.587174269895688e-06,
      "loss": 1.2907,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1343
    },
    {
      "epoch": 0.8431618569636136,
      "grad_norm": 1.8308320045471191,
      "learning_rate": 3.5847093477938955e-06,
      "loss": 1.2235,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1344
    },
    {
      "epoch": 0.8437892095357591,
      "grad_norm": 1.8020336627960205,
      "learning_rate": 3.582243125973631e-06,
      "loss": 1.2072,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1345
    },
    {
      "epoch": 0.8444165621079046,
      "grad_norm": 2.188171148300171,
      "learning_rate": 3.5797756073899665e-06,
      "loss": 1.2954,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1346
    },
    {
      "epoch": 0.8450439146800501,
      "grad_norm": 2.0063283443450928,
      "learning_rate": 3.577306794999527e-06,
      "loss": 1.2451,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1347
    },
    {
      "epoch": 0.8456712672521958,
      "grad_norm": 1.9352948665618896,
      "learning_rate": 3.574836691760489e-06,
      "loss": 1.2155,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1348
    },
    {
      "epoch": 0.8462986198243413,
      "grad_norm": 1.7398347854614258,
      "learning_rate": 3.572365300632574e-06,
      "loss": 1.338,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1349
    },
    {
      "epoch": 0.8469259723964868,
      "grad_norm": 1.6711233854293823,
      "learning_rate": 3.5698926245770495e-06,
      "loss": 1.4667,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1350
    },
    {
      "epoch": 0.8475533249686323,
      "grad_norm": 1.8236483335494995,
      "learning_rate": 3.5674186665567197e-06,
      "loss": 1.379,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1351
    },
    {
      "epoch": 0.848180677540778,
      "grad_norm": 1.9310721158981323,
      "learning_rate": 3.5649434295359265e-06,
      "loss": 1.1314,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1352
    },
    {
      "epoch": 0.8488080301129235,
      "grad_norm": 1.820559024810791,
      "learning_rate": 3.562466916480544e-06,
      "loss": 1.3131,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1353
    },
    {
      "epoch": 0.849435382685069,
      "grad_norm": 1.7749238014221191,
      "learning_rate": 3.5599891303579747e-06,
      "loss": 1.3218,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1354
    },
    {
      "epoch": 0.8500627352572145,
      "grad_norm": 1.6481010913848877,
      "learning_rate": 3.557510074137147e-06,
      "loss": 1.1834,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1355
    },
    {
      "epoch": 0.8506900878293601,
      "grad_norm": 1.7191163301467896,
      "learning_rate": 3.5550297507885123e-06,
      "loss": 1.3883,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1356
    },
    {
      "epoch": 0.8513174404015057,
      "grad_norm": 1.464750051498413,
      "learning_rate": 3.552548163284038e-06,
      "loss": 1.2309,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1357
    },
    {
      "epoch": 0.8519447929736512,
      "grad_norm": 1.6320300102233887,
      "learning_rate": 3.5500653145972085e-06,
      "loss": 1.2576,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1358
    },
    {
      "epoch": 0.8525721455457967,
      "grad_norm": 1.6275529861450195,
      "learning_rate": 3.547581207703017e-06,
      "loss": 1.1981,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1359
    },
    {
      "epoch": 0.8531994981179423,
      "grad_norm": 1.5294578075408936,
      "learning_rate": 3.545095845577967e-06,
      "loss": 1.2997,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1360
    },
    {
      "epoch": 0.8538268506900878,
      "grad_norm": 1.5798944234848022,
      "learning_rate": 3.5426092312000627e-06,
      "loss": 1.3639,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1361
    },
    {
      "epoch": 0.8544542032622334,
      "grad_norm": 1.7874319553375244,
      "learning_rate": 3.540121367548811e-06,
      "loss": 1.3735,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1362
    },
    {
      "epoch": 0.855081555834379,
      "grad_norm": 1.8202686309814453,
      "learning_rate": 3.5376322576052176e-06,
      "loss": 1.235,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1363
    },
    {
      "epoch": 0.8557089084065245,
      "grad_norm": 1.403754711151123,
      "learning_rate": 3.5351419043517764e-06,
      "loss": 1.14,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1364
    },
    {
      "epoch": 0.85633626097867,
      "grad_norm": 1.901134967803955,
      "learning_rate": 3.5326503107724763e-06,
      "loss": 1.2498,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1365
    },
    {
      "epoch": 0.8569636135508155,
      "grad_norm": 1.5691863298416138,
      "learning_rate": 3.5301574798527895e-06,
      "loss": 1.3065,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1366
    },
    {
      "epoch": 0.8575909661229612,
      "grad_norm": 1.2738248109817505,
      "learning_rate": 3.5276634145796706e-06,
      "loss": 1.1509,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1367
    },
    {
      "epoch": 0.8582183186951067,
      "grad_norm": 1.6212133169174194,
      "learning_rate": 3.525168117941554e-06,
      "loss": 1.2146,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1368
    },
    {
      "epoch": 0.8588456712672522,
      "grad_norm": 1.5654613971710205,
      "learning_rate": 3.5226715929283507e-06,
      "loss": 1.2141,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1369
    },
    {
      "epoch": 0.8594730238393977,
      "grad_norm": 1.9199938774108887,
      "learning_rate": 3.5201738425314403e-06,
      "loss": 1.1485,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1370
    },
    {
      "epoch": 0.8601003764115432,
      "grad_norm": 1.919198751449585,
      "learning_rate": 3.5176748697436746e-06,
      "loss": 1.3639,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1371
    },
    {
      "epoch": 0.8607277289836889,
      "grad_norm": 1.6670496463775635,
      "learning_rate": 3.5151746775593676e-06,
      "loss": 1.0675,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1372
    },
    {
      "epoch": 0.8613550815558344,
      "grad_norm": 1.9802881479263306,
      "learning_rate": 3.5126732689742953e-06,
      "loss": 1.2979,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1373
    },
    {
      "epoch": 0.8619824341279799,
      "grad_norm": 1.9427763223648071,
      "learning_rate": 3.5101706469856913e-06,
      "loss": 1.2531,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1374
    },
    {
      "epoch": 0.8626097867001254,
      "grad_norm": 1.8517464399337769,
      "learning_rate": 3.507666814592242e-06,
      "loss": 1.3706,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1375
    },
    {
      "epoch": 0.863237139272271,
      "grad_norm": 1.6945019960403442,
      "learning_rate": 3.505161774794085e-06,
      "loss": 1.2589,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1376
    },
    {
      "epoch": 0.8638644918444166,
      "grad_norm": 1.1320945024490356,
      "learning_rate": 3.5026555305928047e-06,
      "loss": 1.1627,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1377
    },
    {
      "epoch": 0.8644918444165621,
      "grad_norm": 1.643018364906311,
      "learning_rate": 3.500148084991428e-06,
      "loss": 1.2658,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1378
    },
    {
      "epoch": 0.8651191969887077,
      "grad_norm": 1.681472659111023,
      "learning_rate": 3.4976394409944236e-06,
      "loss": 1.3512,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1379
    },
    {
      "epoch": 0.8657465495608532,
      "grad_norm": 1.6602472066879272,
      "learning_rate": 3.495129601607692e-06,
      "loss": 1.3452,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1380
    },
    {
      "epoch": 0.8663739021329987,
      "grad_norm": 1.923701286315918,
      "learning_rate": 3.492618569838571e-06,
      "loss": 1.3233,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1381
    },
    {
      "epoch": 0.8670012547051443,
      "grad_norm": 1.7094676494598389,
      "learning_rate": 3.490106348695822e-06,
      "loss": 1.264,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1382
    },
    {
      "epoch": 0.8676286072772899,
      "grad_norm": 1.6848981380462646,
      "learning_rate": 3.487592941189636e-06,
      "loss": 1.1531,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1383
    },
    {
      "epoch": 0.8682559598494354,
      "grad_norm": 1.67879319190979,
      "learning_rate": 3.485078350331622e-06,
      "loss": 1.1951,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1384
    },
    {
      "epoch": 0.8688833124215809,
      "grad_norm": 1.8846187591552734,
      "learning_rate": 3.4825625791348093e-06,
      "loss": 1.3897,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1385
    },
    {
      "epoch": 0.8695106649937264,
      "grad_norm": 1.9970777034759521,
      "learning_rate": 3.480045630613641e-06,
      "loss": 1.3452,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1386
    },
    {
      "epoch": 0.8701380175658721,
      "grad_norm": 1.727866768836975,
      "learning_rate": 3.477527507783971e-06,
      "loss": 1.2737,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1387
    },
    {
      "epoch": 0.8707653701380176,
      "grad_norm": 2.935303211212158,
      "learning_rate": 3.475008213663058e-06,
      "loss": 1.1305,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1388
    },
    {
      "epoch": 0.8713927227101631,
      "grad_norm": 1.5516635179519653,
      "learning_rate": 3.4724877512695677e-06,
      "loss": 1.2315,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1389
    },
    {
      "epoch": 0.8720200752823086,
      "grad_norm": 1.6491432189941406,
      "learning_rate": 3.469966123623563e-06,
      "loss": 1.2025,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1390
    },
    {
      "epoch": 0.8726474278544541,
      "grad_norm": 1.6983126401901245,
      "learning_rate": 3.467443333746506e-06,
      "loss": 1.4585,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1391
    },
    {
      "epoch": 0.8732747804265998,
      "grad_norm": 2.029611110687256,
      "learning_rate": 3.4649193846612465e-06,
      "loss": 1.283,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1392
    },
    {
      "epoch": 0.8739021329987453,
      "grad_norm": 1.6874094009399414,
      "learning_rate": 3.4623942793920286e-06,
      "loss": 1.3645,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1393
    },
    {
      "epoch": 0.8745294855708908,
      "grad_norm": 1.797200083732605,
      "learning_rate": 3.459868020964478e-06,
      "loss": 1.1996,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1394
    },
    {
      "epoch": 0.8751568381430364,
      "grad_norm": 1.703599452972412,
      "learning_rate": 3.457340612405604e-06,
      "loss": 1.3993,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1395
    },
    {
      "epoch": 0.875784190715182,
      "grad_norm": 1.7407021522521973,
      "learning_rate": 3.4548120567437932e-06,
      "loss": 1.1895,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1396
    },
    {
      "epoch": 0.8764115432873275,
      "grad_norm": 1.3256012201309204,
      "learning_rate": 3.4522823570088073e-06,
      "loss": 1.1549,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1397
    },
    {
      "epoch": 0.877038895859473,
      "grad_norm": 2.297656774520874,
      "learning_rate": 3.4497515162317786e-06,
      "loss": 1.3253,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1398
    },
    {
      "epoch": 0.8776662484316186,
      "grad_norm": 1.6978338956832886,
      "learning_rate": 3.4472195374452067e-06,
      "loss": 1.2676,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1399
    },
    {
      "epoch": 0.8782936010037641,
      "grad_norm": 1.733985424041748,
      "learning_rate": 3.444686423682954e-06,
      "loss": 1.197,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1400
    },
    {
      "epoch": 0.8789209535759097,
      "grad_norm": 2.119873285293579,
      "learning_rate": 3.442152177980245e-06,
      "loss": 1.1963,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1401
    },
    {
      "epoch": 0.8795483061480552,
      "grad_norm": 1.4531352519989014,
      "learning_rate": 3.4396168033736564e-06,
      "loss": 1.1837,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1402
    },
    {
      "epoch": 0.8801756587202008,
      "grad_norm": 1.9602431058883667,
      "learning_rate": 3.437080302901122e-06,
      "loss": 1.2665,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1403
    },
    {
      "epoch": 0.8808030112923463,
      "grad_norm": 1.8748997449874878,
      "learning_rate": 3.434542679601922e-06,
      "loss": 1.2853,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1404
    },
    {
      "epoch": 0.8814303638644918,
      "grad_norm": 1.863858938217163,
      "learning_rate": 3.432003936516684e-06,
      "loss": 1.2804,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1405
    },
    {
      "epoch": 0.8820577164366374,
      "grad_norm": 1.690569519996643,
      "learning_rate": 3.429464076687375e-06,
      "loss": 1.3098,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1406
    },
    {
      "epoch": 0.882685069008783,
      "grad_norm": 1.9281234741210938,
      "learning_rate": 3.4269231031573023e-06,
      "loss": 1.2294,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1407
    },
    {
      "epoch": 0.8833124215809285,
      "grad_norm": 1.8205302953720093,
      "learning_rate": 3.4243810189711052e-06,
      "loss": 1.3322,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1408
    },
    {
      "epoch": 0.883939774153074,
      "grad_norm": 1.8361729383468628,
      "learning_rate": 3.4218378271747566e-06,
      "loss": 1.1492,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1409
    },
    {
      "epoch": 0.8845671267252195,
      "grad_norm": 1.9009759426116943,
      "learning_rate": 3.419293530815554e-06,
      "loss": 1.2921,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1410
    },
    {
      "epoch": 0.8851944792973652,
      "grad_norm": 1.9854016304016113,
      "learning_rate": 3.4167481329421204e-06,
      "loss": 1.2964,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1411
    },
    {
      "epoch": 0.8858218318695107,
      "grad_norm": 2.017411470413208,
      "learning_rate": 3.4142016366043973e-06,
      "loss": 1.323,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1412
    },
    {
      "epoch": 0.8864491844416562,
      "grad_norm": 1.6297651529312134,
      "learning_rate": 3.4116540448536432e-06,
      "loss": 1.1715,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1413
    },
    {
      "epoch": 0.8870765370138017,
      "grad_norm": 1.628180742263794,
      "learning_rate": 3.4091053607424295e-06,
      "loss": 1.2592,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1414
    },
    {
      "epoch": 0.8877038895859473,
      "grad_norm": 1.8852691650390625,
      "learning_rate": 3.4065555873246348e-06,
      "loss": 1.2048,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1415
    },
    {
      "epoch": 0.8883312421580929,
      "grad_norm": 2.035797357559204,
      "learning_rate": 3.404004727655444e-06,
      "loss": 1.3187,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1416
    },
    {
      "epoch": 0.8889585947302384,
      "grad_norm": 1.75394606590271,
      "learning_rate": 3.401452784791345e-06,
      "loss": 1.1598,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1417
    },
    {
      "epoch": 0.8895859473023839,
      "grad_norm": 1.621436595916748,
      "learning_rate": 3.39889976179012e-06,
      "loss": 1.2028,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1418
    },
    {
      "epoch": 0.8902132998745295,
      "grad_norm": 2.014482021331787,
      "learning_rate": 3.396345661710849e-06,
      "loss": 1.3148,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1419
    },
    {
      "epoch": 0.890840652446675,
      "grad_norm": 1.7939558029174805,
      "learning_rate": 3.393790487613901e-06,
      "loss": 1.2393,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1420
    },
    {
      "epoch": 0.8914680050188206,
      "grad_norm": 1.7533239126205444,
      "learning_rate": 3.3912342425609325e-06,
      "loss": 1.246,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1421
    },
    {
      "epoch": 0.8920953575909661,
      "grad_norm": 1.6140942573547363,
      "learning_rate": 3.388676929614882e-06,
      "loss": 1.2078,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1422
    },
    {
      "epoch": 0.8927227101631117,
      "grad_norm": 1.6957402229309082,
      "learning_rate": 3.386118551839969e-06,
      "loss": 1.2466,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1423
    },
    {
      "epoch": 0.8933500627352572,
      "grad_norm": 1.8988665342330933,
      "learning_rate": 3.383559112301687e-06,
      "loss": 1.2817,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1424
    },
    {
      "epoch": 0.8939774153074027,
      "grad_norm": 2.0279648303985596,
      "learning_rate": 3.380998614066805e-06,
      "loss": 1.3186,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1425
    },
    {
      "epoch": 0.8946047678795483,
      "grad_norm": 1.6259845495224,
      "learning_rate": 3.3784370602033572e-06,
      "loss": 1.1776,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1426
    },
    {
      "epoch": 0.8952321204516939,
      "grad_norm": 1.4232373237609863,
      "learning_rate": 3.3758744537806443e-06,
      "loss": 1.1832,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1427
    },
    {
      "epoch": 0.8958594730238394,
      "grad_norm": 1.6897512674331665,
      "learning_rate": 3.3733107978692287e-06,
      "loss": 1.1143,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1428
    },
    {
      "epoch": 0.8964868255959849,
      "grad_norm": 1.6576286554336548,
      "learning_rate": 3.370746095540928e-06,
      "loss": 1.3167,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1429
    },
    {
      "epoch": 0.8971141781681304,
      "grad_norm": 1.9939161539077759,
      "learning_rate": 3.3681803498688165e-06,
      "loss": 1.3256,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1430
    },
    {
      "epoch": 0.8977415307402761,
      "grad_norm": 1.6561397314071655,
      "learning_rate": 3.365613563927217e-06,
      "loss": 1.4733,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1431
    },
    {
      "epoch": 0.8983688833124216,
      "grad_norm": 1.4825150966644287,
      "learning_rate": 3.363045740791698e-06,
      "loss": 1.1664,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1432
    },
    {
      "epoch": 0.8989962358845671,
      "grad_norm": 1.7675479650497437,
      "learning_rate": 3.3604768835390746e-06,
      "loss": 1.2684,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1433
    },
    {
      "epoch": 0.8996235884567126,
      "grad_norm": 1.8060680627822876,
      "learning_rate": 3.3579069952473964e-06,
      "loss": 1.4176,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1434
    },
    {
      "epoch": 0.9002509410288582,
      "grad_norm": 2.16556978225708,
      "learning_rate": 3.35533607899595e-06,
      "loss": 1.222,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1435
    },
    {
      "epoch": 0.9008782936010038,
      "grad_norm": 1.771340250968933,
      "learning_rate": 3.352764137865254e-06,
      "loss": 1.2995,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1436
    },
    {
      "epoch": 0.9015056461731493,
      "grad_norm": 2.0096311569213867,
      "learning_rate": 3.3501911749370557e-06,
      "loss": 1.1834,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1437
    },
    {
      "epoch": 0.9021329987452948,
      "grad_norm": 1.8969783782958984,
      "learning_rate": 3.3476171932943265e-06,
      "loss": 1.2817,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1438
    },
    {
      "epoch": 0.9027603513174404,
      "grad_norm": 1.6149653196334839,
      "learning_rate": 3.345042196021257e-06,
      "loss": 1.2378,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1439
    },
    {
      "epoch": 0.903387703889586,
      "grad_norm": 1.2029558420181274,
      "learning_rate": 3.3424661862032563e-06,
      "loss": 1.2035,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1440
    },
    {
      "epoch": 0.9040150564617315,
      "grad_norm": 1.5148495435714722,
      "learning_rate": 3.3398891669269466e-06,
      "loss": 1.2292,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1441
    },
    {
      "epoch": 0.904642409033877,
      "grad_norm": 1.7740296125411987,
      "learning_rate": 3.3373111412801594e-06,
      "loss": 1.3184,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1442
    },
    {
      "epoch": 0.9052697616060226,
      "grad_norm": 1.9513882398605347,
      "learning_rate": 3.3347321123519304e-06,
      "loss": 1.2391,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1443
    },
    {
      "epoch": 0.9058971141781681,
      "grad_norm": 1.9446839094161987,
      "learning_rate": 3.3321520832325e-06,
      "loss": 1.2979,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1444
    },
    {
      "epoch": 0.9065244667503137,
      "grad_norm": 1.897504448890686,
      "learning_rate": 3.3295710570133063e-06,
      "loss": 1.4606,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1445
    },
    {
      "epoch": 0.9071518193224593,
      "grad_norm": 1.7566696405410767,
      "learning_rate": 3.326989036786981e-06,
      "loss": 1.3677,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1446
    },
    {
      "epoch": 0.9077791718946048,
      "grad_norm": 1.8443048000335693,
      "learning_rate": 3.3244060256473488e-06,
      "loss": 1.2227,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1447
    },
    {
      "epoch": 0.9084065244667503,
      "grad_norm": 2.029710292816162,
      "learning_rate": 3.32182202668942e-06,
      "loss": 1.2015,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1448
    },
    {
      "epoch": 0.9090338770388958,
      "grad_norm": 1.9194732904434204,
      "learning_rate": 3.319237043009389e-06,
      "loss": 1.2434,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1449
    },
    {
      "epoch": 0.9096612296110415,
      "grad_norm": 1.645271897315979,
      "learning_rate": 3.3166510777046313e-06,
      "loss": 1.2735,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1450
    },
    {
      "epoch": 0.910288582183187,
      "grad_norm": 1.6370049715042114,
      "learning_rate": 3.314064133873696e-06,
      "loss": 1.1844,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1451
    },
    {
      "epoch": 0.9109159347553325,
      "grad_norm": 1.9140526056289673,
      "learning_rate": 3.311476214616307e-06,
      "loss": 1.2394,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1452
    },
    {
      "epoch": 0.911543287327478,
      "grad_norm": 1.867573857307434,
      "learning_rate": 3.3088873230333562e-06,
      "loss": 1.2614,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1453
    },
    {
      "epoch": 0.9121706398996235,
      "grad_norm": 1.5740164518356323,
      "learning_rate": 3.3062974622269006e-06,
      "loss": 1.3641,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1454
    },
    {
      "epoch": 0.9127979924717692,
      "grad_norm": 1.985310673713684,
      "learning_rate": 3.3037066353001595e-06,
      "loss": 1.2517,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1455
    },
    {
      "epoch": 0.9134253450439147,
      "grad_norm": 1.9480928182601929,
      "learning_rate": 3.301114845357507e-06,
      "loss": 1.2349,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1456
    },
    {
      "epoch": 0.9140526976160602,
      "grad_norm": 1.7110068798065186,
      "learning_rate": 3.2985220955044742e-06,
      "loss": 1.3385,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1457
    },
    {
      "epoch": 0.9146800501882058,
      "grad_norm": 1.584576964378357,
      "learning_rate": 3.2959283888477416e-06,
      "loss": 1.1978,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1458
    },
    {
      "epoch": 0.9153074027603513,
      "grad_norm": 1.6054190397262573,
      "learning_rate": 3.2933337284951338e-06,
      "loss": 1.4712,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1459
    },
    {
      "epoch": 0.9159347553324969,
      "grad_norm": 1.9523940086364746,
      "learning_rate": 3.290738117555622e-06,
      "loss": 1.3496,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1460
    },
    {
      "epoch": 0.9165621079046424,
      "grad_norm": 1.800445318222046,
      "learning_rate": 3.2881415591393128e-06,
      "loss": 1.3048,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1461
    },
    {
      "epoch": 0.917189460476788,
      "grad_norm": 1.6035422086715698,
      "learning_rate": 3.2855440563574516e-06,
      "loss": 1.4824,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1462
    },
    {
      "epoch": 0.9178168130489335,
      "grad_norm": 2.0005290508270264,
      "learning_rate": 3.2829456123224115e-06,
      "loss": 1.2858,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1463
    },
    {
      "epoch": 0.918444165621079,
      "grad_norm": 1.987656593322754,
      "learning_rate": 3.2803462301476962e-06,
      "loss": 1.3675,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1464
    },
    {
      "epoch": 0.9190715181932246,
      "grad_norm": 1.3656636476516724,
      "learning_rate": 3.277745912947933e-06,
      "loss": 1.1349,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1465
    },
    {
      "epoch": 0.9196988707653702,
      "grad_norm": 1.7897242307662964,
      "learning_rate": 3.2751446638388694e-06,
      "loss": 1.3293,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1466
    },
    {
      "epoch": 0.9203262233375157,
      "grad_norm": 1.4551517963409424,
      "learning_rate": 3.272542485937369e-06,
      "loss": 1.2862,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1467
    },
    {
      "epoch": 0.9209535759096612,
      "grad_norm": 1.6863459348678589,
      "learning_rate": 3.2699393823614088e-06,
      "loss": 1.2697,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1468
    },
    {
      "epoch": 0.9215809284818067,
      "grad_norm": 1.777715802192688,
      "learning_rate": 3.2673353562300753e-06,
      "loss": 1.1703,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1469
    },
    {
      "epoch": 0.9222082810539524,
      "grad_norm": 1.701059103012085,
      "learning_rate": 3.2647304106635587e-06,
      "loss": 1.2239,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1470
    },
    {
      "epoch": 0.9228356336260979,
      "grad_norm": 2.0280165672302246,
      "learning_rate": 3.2621245487831533e-06,
      "loss": 1.3686,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1471
    },
    {
      "epoch": 0.9234629861982434,
      "grad_norm": 1.660024881362915,
      "learning_rate": 3.25951777371125e-06,
      "loss": 1.3431,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1472
    },
    {
      "epoch": 0.9240903387703889,
      "grad_norm": 1.6660288572311401,
      "learning_rate": 3.2569100885713333e-06,
      "loss": 1.5176,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1473
    },
    {
      "epoch": 0.9247176913425345,
      "grad_norm": 1.8647363185882568,
      "learning_rate": 3.2543014964879814e-06,
      "loss": 1.2584,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1474
    },
    {
      "epoch": 0.9253450439146801,
      "grad_norm": 1.8955433368682861,
      "learning_rate": 3.2516920005868536e-06,
      "loss": 1.1977,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1475
    },
    {
      "epoch": 0.9259723964868256,
      "grad_norm": 1.7556467056274414,
      "learning_rate": 3.249081603994698e-06,
      "loss": 1.3804,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1476
    },
    {
      "epoch": 0.9265997490589711,
      "grad_norm": 1.8404650688171387,
      "learning_rate": 3.2464703098393376e-06,
      "loss": 1.1758,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1477
    },
    {
      "epoch": 0.9272271016311167,
      "grad_norm": 1.8023796081542969,
      "learning_rate": 3.2438581212496723e-06,
      "loss": 1.2302,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1478
    },
    {
      "epoch": 0.9278544542032622,
      "grad_norm": 1.2507985830307007,
      "learning_rate": 3.2412450413556753e-06,
      "loss": 1.239,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1479
    },
    {
      "epoch": 0.9284818067754078,
      "grad_norm": 1.8706785440444946,
      "learning_rate": 3.2386310732883865e-06,
      "loss": 1.1204,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1480
    },
    {
      "epoch": 0.9291091593475533,
      "grad_norm": 1.7828782796859741,
      "learning_rate": 3.2360162201799085e-06,
      "loss": 1.2433,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1481
    },
    {
      "epoch": 0.9297365119196989,
      "grad_norm": 1.877037763595581,
      "learning_rate": 3.2334004851634073e-06,
      "loss": 1.1678,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1482
    },
    {
      "epoch": 0.9303638644918444,
      "grad_norm": 1.7916743755340576,
      "learning_rate": 3.2307838713731033e-06,
      "loss": 1.3769,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1483
    },
    {
      "epoch": 0.93099121706399,
      "grad_norm": 1.0694576501846313,
      "learning_rate": 3.228166381944272e-06,
      "loss": 1.0804,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1484
    },
    {
      "epoch": 0.9316185696361355,
      "grad_norm": 1.9489041566848755,
      "learning_rate": 3.225548020013235e-06,
      "loss": 1.1571,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1485
    },
    {
      "epoch": 0.9322459222082811,
      "grad_norm": 1.5911556482315063,
      "learning_rate": 3.2229287887173623e-06,
      "loss": 1.3908,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1486
    },
    {
      "epoch": 0.9328732747804266,
      "grad_norm": 1.844138741493225,
      "learning_rate": 3.2203086911950643e-06,
      "loss": 1.1133,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1487
    },
    {
      "epoch": 0.9335006273525721,
      "grad_norm": 1.7832318544387817,
      "learning_rate": 3.21768773058579e-06,
      "loss": 1.3063,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1488
    },
    {
      "epoch": 0.9341279799247177,
      "grad_norm": 1.8222450017929077,
      "learning_rate": 3.2150659100300215e-06,
      "loss": 1.2093,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1489
    },
    {
      "epoch": 0.9347553324968633,
      "grad_norm": 1.9622671604156494,
      "learning_rate": 3.2124432326692717e-06,
      "loss": 1.2801,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1490
    },
    {
      "epoch": 0.9353826850690088,
      "grad_norm": 2.0116896629333496,
      "learning_rate": 3.2098197016460807e-06,
      "loss": 1.2931,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1491
    },
    {
      "epoch": 0.9360100376411543,
      "grad_norm": 1.770477056503296,
      "learning_rate": 3.2071953201040113e-06,
      "loss": 1.2336,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1492
    },
    {
      "epoch": 0.9366373902132998,
      "grad_norm": 1.954681396484375,
      "learning_rate": 3.2045700911876443e-06,
      "loss": 1.2612,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1493
    },
    {
      "epoch": 0.9372647427854455,
      "grad_norm": 1.5379961729049683,
      "learning_rate": 3.2019440180425774e-06,
      "loss": 1.209,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1494
    },
    {
      "epoch": 0.937892095357591,
      "grad_norm": 1.8084098100662231,
      "learning_rate": 3.1993171038154203e-06,
      "loss": 1.4448,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1495
    },
    {
      "epoch": 0.9385194479297365,
      "grad_norm": 1.5959023237228394,
      "learning_rate": 3.1966893516537877e-06,
      "loss": 1.2557,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1496
    },
    {
      "epoch": 0.939146800501882,
      "grad_norm": 1.8980833292007446,
      "learning_rate": 3.1940607647062997e-06,
      "loss": 1.2172,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1497
    },
    {
      "epoch": 0.9397741530740276,
      "grad_norm": 1.6335502862930298,
      "learning_rate": 3.191431346122579e-06,
      "loss": 1.1199,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1498
    },
    {
      "epoch": 0.9404015056461732,
      "grad_norm": 1.936535358428955,
      "learning_rate": 3.1888010990532412e-06,
      "loss": 1.2016,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1499
    },
    {
      "epoch": 0.9410288582183187,
      "grad_norm": 2.046355962753296,
      "learning_rate": 3.1861700266498975e-06,
      "loss": 1.2847,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1500
    },
    {
      "epoch": 0.9416562107904642,
      "grad_norm": 1.6399589776992798,
      "learning_rate": 3.183538132065145e-06,
      "loss": 1.1248,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1501
    },
    {
      "epoch": 0.9422835633626098,
      "grad_norm": 2.1181039810180664,
      "learning_rate": 3.180905418452569e-06,
      "loss": 1.2885,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1502
    },
    {
      "epoch": 0.9429109159347553,
      "grad_norm": 1.5633800029754639,
      "learning_rate": 3.178271888966734e-06,
      "loss": 1.1638,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1503
    },
    {
      "epoch": 0.9435382685069009,
      "grad_norm": 2.5201916694641113,
      "learning_rate": 3.1756375467631832e-06,
      "loss": 1.2562,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1504
    },
    {
      "epoch": 0.9441656210790464,
      "grad_norm": 1.6307491064071655,
      "learning_rate": 3.173002394998434e-06,
      "loss": 1.2839,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1505
    },
    {
      "epoch": 0.944792973651192,
      "grad_norm": 1.7592899799346924,
      "learning_rate": 3.1703664368299724e-06,
      "loss": 1.2898,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1506
    },
    {
      "epoch": 0.9454203262233375,
      "grad_norm": 1.8422400951385498,
      "learning_rate": 3.167729675416252e-06,
      "loss": 1.1932,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1507
    },
    {
      "epoch": 0.946047678795483,
      "grad_norm": 1.5810227394104004,
      "learning_rate": 3.165092113916688e-06,
      "loss": 1.0733,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1508
    },
    {
      "epoch": 0.9466750313676286,
      "grad_norm": 1.9519747495651245,
      "learning_rate": 3.162453755491655e-06,
      "loss": 1.3441,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1509
    },
    {
      "epoch": 0.9473023839397742,
      "grad_norm": 1.8681365251541138,
      "learning_rate": 3.1598146033024805e-06,
      "loss": 1.2143,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1510
    },
    {
      "epoch": 0.9479297365119197,
      "grad_norm": 1.7927610874176025,
      "learning_rate": 3.1571746605114466e-06,
      "loss": 1.2172,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1511
    },
    {
      "epoch": 0.9485570890840652,
      "grad_norm": 1.7754586935043335,
      "learning_rate": 3.1545339302817796e-06,
      "loss": 1.1724,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1512
    },
    {
      "epoch": 0.9491844416562107,
      "grad_norm": 1.8852242231369019,
      "learning_rate": 3.1518924157776507e-06,
      "loss": 1.3607,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1513
    },
    {
      "epoch": 0.9498117942283564,
      "grad_norm": 1.8877124786376953,
      "learning_rate": 3.149250120164171e-06,
      "loss": 1.3618,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1514
    },
    {
      "epoch": 0.9504391468005019,
      "grad_norm": 1.8442416191101074,
      "learning_rate": 3.1466070466073874e-06,
      "loss": 1.2654,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1515
    },
    {
      "epoch": 0.9510664993726474,
      "grad_norm": 1.7711665630340576,
      "learning_rate": 3.143963198274278e-06,
      "loss": 1.272,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1516
    },
    {
      "epoch": 0.951693851944793,
      "grad_norm": 2.4715964794158936,
      "learning_rate": 3.1413185783327503e-06,
      "loss": 1.2734,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1517
    },
    {
      "epoch": 0.9523212045169385,
      "grad_norm": 1.8979958295822144,
      "learning_rate": 3.1386731899516353e-06,
      "loss": 1.3359,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1518
    },
    {
      "epoch": 0.9529485570890841,
      "grad_norm": 2.25026273727417,
      "learning_rate": 3.136027036300687e-06,
      "loss": 1.3138,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1519
    },
    {
      "epoch": 0.9535759096612296,
      "grad_norm": 2.0142345428466797,
      "learning_rate": 3.1333801205505734e-06,
      "loss": 1.3283,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1520
    },
    {
      "epoch": 0.9542032622333751,
      "grad_norm": 1.6834921836853027,
      "learning_rate": 3.130732445872877e-06,
      "loss": 1.1996,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1521
    },
    {
      "epoch": 0.9548306148055207,
      "grad_norm": 1.9240096807479858,
      "learning_rate": 3.128084015440092e-06,
      "loss": 1.1892,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1522
    },
    {
      "epoch": 0.9554579673776662,
      "grad_norm": 1.3459506034851074,
      "learning_rate": 3.125434832425613e-06,
      "loss": 1.12,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1523
    },
    {
      "epoch": 0.9560853199498118,
      "grad_norm": 1.0296326875686646,
      "learning_rate": 3.122784900003742e-06,
      "loss": 1.1246,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1524
    },
    {
      "epoch": 0.9567126725219574,
      "grad_norm": 1.813663125038147,
      "learning_rate": 3.120134221349674e-06,
      "loss": 1.2281,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1525
    },
    {
      "epoch": 0.9573400250941029,
      "grad_norm": 1.8600475788116455,
      "learning_rate": 3.117482799639502e-06,
      "loss": 1.2991,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1526
    },
    {
      "epoch": 0.9579673776662484,
      "grad_norm": 1.82486891746521,
      "learning_rate": 3.114830638050206e-06,
      "loss": 1.2133,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1527
    },
    {
      "epoch": 0.958594730238394,
      "grad_norm": 1.6651502847671509,
      "learning_rate": 3.112177739759656e-06,
      "loss": 1.2577,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1528
    },
    {
      "epoch": 0.9592220828105396,
      "grad_norm": 1.9689431190490723,
      "learning_rate": 3.109524107946602e-06,
      "loss": 1.3149,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1529
    },
    {
      "epoch": 0.9598494353826851,
      "grad_norm": 1.5160422325134277,
      "learning_rate": 3.1068697457906736e-06,
      "loss": 1.2537,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1530
    },
    {
      "epoch": 0.9604767879548306,
      "grad_norm": 1.9524043798446655,
      "learning_rate": 3.104214656472377e-06,
      "loss": 1.3181,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1531
    },
    {
      "epoch": 0.9611041405269761,
      "grad_norm": 1.8175241947174072,
      "learning_rate": 3.1015588431730873e-06,
      "loss": 1.23,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1532
    },
    {
      "epoch": 0.9617314930991218,
      "grad_norm": 1.801020622253418,
      "learning_rate": 3.0989023090750496e-06,
      "loss": 1.2502,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1533
    },
    {
      "epoch": 0.9623588456712673,
      "grad_norm": 1.8386653661727905,
      "learning_rate": 3.0962450573613705e-06,
      "loss": 1.2601,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1534
    },
    {
      "epoch": 0.9629861982434128,
      "grad_norm": 3.8357949256896973,
      "learning_rate": 3.093587091216017e-06,
      "loss": 1.234,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1535
    },
    {
      "epoch": 0.9636135508155583,
      "grad_norm": 1.52729070186615,
      "learning_rate": 3.0909284138238136e-06,
      "loss": 1.1171,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1536
    },
    {
      "epoch": 0.9642409033877039,
      "grad_norm": 1.551767349243164,
      "learning_rate": 3.0882690283704355e-06,
      "loss": 1.0957,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1537
    },
    {
      "epoch": 0.9648682559598495,
      "grad_norm": 1.573925256729126,
      "learning_rate": 3.085608938042406e-06,
      "loss": 1.3389,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1538
    },
    {
      "epoch": 0.965495608531995,
      "grad_norm": 1.9069485664367676,
      "learning_rate": 3.0829481460270937e-06,
      "loss": 1.3142,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1539
    },
    {
      "epoch": 0.9661229611041405,
      "grad_norm": 1.6924388408660889,
      "learning_rate": 3.0802866555127094e-06,
      "loss": 1.2382,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1540
    },
    {
      "epoch": 0.966750313676286,
      "grad_norm": 1.9501606225967407,
      "learning_rate": 3.077624469688299e-06,
      "loss": 1.3925,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1541
    },
    {
      "epoch": 0.9673776662484316,
      "grad_norm": 1.637292742729187,
      "learning_rate": 3.074961591743741e-06,
      "loss": 1.1863,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1542
    },
    {
      "epoch": 0.9680050188205772,
      "grad_norm": 1.2631224393844604,
      "learning_rate": 3.0722980248697447e-06,
      "loss": 1.2935,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1543
    },
    {
      "epoch": 0.9686323713927227,
      "grad_norm": 1.5776970386505127,
      "learning_rate": 3.0696337722578444e-06,
      "loss": 1.3131,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1544
    },
    {
      "epoch": 0.9692597239648683,
      "grad_norm": 1.8164136409759521,
      "learning_rate": 3.0669688371003962e-06,
      "loss": 1.2615,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1545
    },
    {
      "epoch": 0.9698870765370138,
      "grad_norm": 1.9003275632858276,
      "learning_rate": 3.064303222590574e-06,
      "loss": 1.2848,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1546
    },
    {
      "epoch": 0.9705144291091593,
      "grad_norm": 1.5986956357955933,
      "learning_rate": 3.061636931922365e-06,
      "loss": 1.1825,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1547
    },
    {
      "epoch": 0.9711417816813049,
      "grad_norm": 1.9041141271591187,
      "learning_rate": 3.058969968290567e-06,
      "loss": 1.242,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1548
    },
    {
      "epoch": 0.9717691342534505,
      "grad_norm": 1.947312831878662,
      "learning_rate": 3.056302334890786e-06,
      "loss": 1.1776,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1549
    },
    {
      "epoch": 0.972396486825596,
      "grad_norm": 1.5843112468719482,
      "learning_rate": 3.053634034919428e-06,
      "loss": 1.1799,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1550
    },
    {
      "epoch": 0.9730238393977415,
      "grad_norm": 1.5609612464904785,
      "learning_rate": 3.0509650715736977e-06,
      "loss": 1.303,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1551
    },
    {
      "epoch": 0.973651191969887,
      "grad_norm": 1.6556776762008667,
      "learning_rate": 3.0482954480515963e-06,
      "loss": 1.1567,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1552
    },
    {
      "epoch": 0.9742785445420327,
      "grad_norm": 1.801066517829895,
      "learning_rate": 3.045625167551915e-06,
      "loss": 1.2528,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1553
    },
    {
      "epoch": 0.9749058971141782,
      "grad_norm": 1.9137632846832275,
      "learning_rate": 3.0429542332742322e-06,
      "loss": 1.2109,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1554
    },
    {
      "epoch": 0.9755332496863237,
      "grad_norm": 2.0313117504119873,
      "learning_rate": 3.040282648418911e-06,
      "loss": 1.226,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1555
    },
    {
      "epoch": 0.9761606022584692,
      "grad_norm": 1.9553405046463013,
      "learning_rate": 3.0376104161870922e-06,
      "loss": 1.3244,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1556
    },
    {
      "epoch": 0.9767879548306148,
      "grad_norm": 1.9397183656692505,
      "learning_rate": 3.0349375397806928e-06,
      "loss": 1.2911,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1557
    },
    {
      "epoch": 0.9774153074027604,
      "grad_norm": 1.5607104301452637,
      "learning_rate": 3.0322640224024024e-06,
      "loss": 1.3396,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1558
    },
    {
      "epoch": 0.9780426599749059,
      "grad_norm": 1.742556095123291,
      "learning_rate": 3.0295898672556785e-06,
      "loss": 1.2465,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1559
    },
    {
      "epoch": 0.9786700125470514,
      "grad_norm": 1.6819292306900024,
      "learning_rate": 3.0269150775447414e-06,
      "loss": 1.1875,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1560
    },
    {
      "epoch": 0.979297365119197,
      "grad_norm": 1.7650625705718994,
      "learning_rate": 3.0242396564745736e-06,
      "loss": 1.2395,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1561
    },
    {
      "epoch": 0.9799247176913425,
      "grad_norm": 1.3795490264892578,
      "learning_rate": 3.0215636072509134e-06,
      "loss": 1.1491,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1562
    },
    {
      "epoch": 0.9805520702634881,
      "grad_norm": 1.8588756322860718,
      "learning_rate": 3.0188869330802512e-06,
      "loss": 1.3543,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1563
    },
    {
      "epoch": 0.9811794228356336,
      "grad_norm": 2.0225319862365723,
      "learning_rate": 3.016209637169827e-06,
      "loss": 1.3106,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1564
    },
    {
      "epoch": 0.9818067754077792,
      "grad_norm": 1.8785102367401123,
      "learning_rate": 3.0135317227276247e-06,
      "loss": 1.3376,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1565
    },
    {
      "epoch": 0.9824341279799247,
      "grad_norm": 1.3855329751968384,
      "learning_rate": 3.0108531929623708e-06,
      "loss": 1.1439,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1566
    },
    {
      "epoch": 0.9830614805520702,
      "grad_norm": 1.591722011566162,
      "learning_rate": 3.0081740510835287e-06,
      "loss": 1.1241,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1567
    },
    {
      "epoch": 0.9836888331242158,
      "grad_norm": 1.98593008518219,
      "learning_rate": 3.005494300301294e-06,
      "loss": 1.1092,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1568
    },
    {
      "epoch": 0.9843161856963614,
      "grad_norm": 1.7478628158569336,
      "learning_rate": 3.0028139438265946e-06,
      "loss": 1.3299,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1569
    },
    {
      "epoch": 0.9849435382685069,
      "grad_norm": 1.8850418329238892,
      "learning_rate": 3.0001329848710803e-06,
      "loss": 1.2433,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1570
    },
    {
      "epoch": 0.9855708908406524,
      "grad_norm": 1.8818042278289795,
      "learning_rate": 2.9974514266471254e-06,
      "loss": 1.3072,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1571
    },
    {
      "epoch": 0.986198243412798,
      "grad_norm": 1.5135859251022339,
      "learning_rate": 2.994769272367822e-06,
      "loss": 1.4054,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1572
    },
    {
      "epoch": 0.9868255959849436,
      "grad_norm": 1.7623196840286255,
      "learning_rate": 2.9920865252469764e-06,
      "loss": 1.3091,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1573
    },
    {
      "epoch": 0.9874529485570891,
      "grad_norm": 1.6980853080749512,
      "learning_rate": 2.9894031884991047e-06,
      "loss": 1.2509,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1574
    },
    {
      "epoch": 0.9880803011292346,
      "grad_norm": 1.5661550760269165,
      "learning_rate": 2.9867192653394307e-06,
      "loss": 1.3949,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1575
    },
    {
      "epoch": 0.9887076537013801,
      "grad_norm": 1.6995539665222168,
      "learning_rate": 2.9840347589838786e-06,
      "loss": 1.3404,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1576
    },
    {
      "epoch": 0.9893350062735258,
      "grad_norm": 2.477058172225952,
      "learning_rate": 2.9813496726490732e-06,
      "loss": 1.1899,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1577
    },
    {
      "epoch": 0.9899623588456713,
      "grad_norm": 2.359335422515869,
      "learning_rate": 2.978664009552334e-06,
      "loss": 1.3828,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1578
    },
    {
      "epoch": 0.9905897114178168,
      "grad_norm": 1.6977730989456177,
      "learning_rate": 2.975977772911671e-06,
      "loss": 1.1621,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1579
    },
    {
      "epoch": 0.9912170639899623,
      "grad_norm": 1.6472805738449097,
      "learning_rate": 2.9732909659457827e-06,
      "loss": 1.3421,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1580
    },
    {
      "epoch": 0.9918444165621079,
      "grad_norm": 1.8243848085403442,
      "learning_rate": 2.970603591874049e-06,
      "loss": 1.3114,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1581
    },
    {
      "epoch": 0.9924717691342535,
      "grad_norm": 1.5195167064666748,
      "learning_rate": 2.967915653916531e-06,
      "loss": 1.1619,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1582
    },
    {
      "epoch": 0.993099121706399,
      "grad_norm": 1.8628544807434082,
      "learning_rate": 2.965227155293965e-06,
      "loss": 1.3999,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1583
    },
    {
      "epoch": 0.9937264742785445,
      "grad_norm": 1.9833089113235474,
      "learning_rate": 2.9625380992277585e-06,
      "loss": 1.4334,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1584
    },
    {
      "epoch": 0.9943538268506901,
      "grad_norm": 1.512103796005249,
      "learning_rate": 2.9598484889399873e-06,
      "loss": 1.2211,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1585
    },
    {
      "epoch": 0.9949811794228356,
      "grad_norm": 1.7887911796569824,
      "learning_rate": 2.9571583276533923e-06,
      "loss": 1.2617,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1586
    },
    {
      "epoch": 0.9956085319949812,
      "grad_norm": 1.7717543840408325,
      "learning_rate": 2.9544676185913725e-06,
      "loss": 1.3387,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1587
    },
    {
      "epoch": 0.9962358845671268,
      "grad_norm": 1.646287202835083,
      "learning_rate": 2.951776364977985e-06,
      "loss": 1.2494,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1588
    },
    {
      "epoch": 0.9968632371392723,
      "grad_norm": 1.721743106842041,
      "learning_rate": 2.949084570037939e-06,
      "loss": 1.402,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1589
    },
    {
      "epoch": 0.9974905897114178,
      "grad_norm": 1.685555338859558,
      "learning_rate": 2.946392236996592e-06,
      "loss": 1.2543,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1590
    },
    {
      "epoch": 0.9981179422835633,
      "grad_norm": 1.7904672622680664,
      "learning_rate": 2.9436993690799464e-06,
      "loss": 1.2328,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1591
    },
    {
      "epoch": 0.998745294855709,
      "grad_norm": 1.5652368068695068,
      "learning_rate": 2.941005969514646e-06,
      "loss": 1.2202,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1592
    },
    {
      "epoch": 0.9993726474278545,
      "grad_norm": 1.5595532655715942,
      "learning_rate": 2.93831204152797e-06,
      "loss": 1.2724,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1593
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.7482446432113647,
      "learning_rate": 2.935617588347832e-06,
      "loss": 1.1597,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1594
    },
    {
      "epoch": 1.0006273525721456,
      "grad_norm": 1.098419189453125,
      "learning_rate": 2.932922613202777e-06,
      "loss": 1.273,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1595
    },
    {
      "epoch": 1.001254705144291,
      "grad_norm": 1.6278892755508423,
      "learning_rate": 2.9302271193219705e-06,
      "loss": 1.1618,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1596
    },
    {
      "epoch": 1.0018820577164367,
      "grad_norm": 1.8631848096847534,
      "learning_rate": 2.927531109935204e-06,
      "loss": 1.248,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1597
    },
    {
      "epoch": 1.002509410288582,
      "grad_norm": 1.707324743270874,
      "learning_rate": 2.924834588272884e-06,
      "loss": 1.1259,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1598
    },
    {
      "epoch": 1.0031367628607277,
      "grad_norm": 1.3775336742401123,
      "learning_rate": 2.922137557566032e-06,
      "loss": 1.2228,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1599
    },
    {
      "epoch": 1.0037641154328734,
      "grad_norm": 1.8195487260818481,
      "learning_rate": 2.9194400210462808e-06,
      "loss": 1.3825,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1600
    },
    {
      "epoch": 1.0043914680050188,
      "grad_norm": 1.2769285440444946,
      "learning_rate": 2.9167419819458653e-06,
      "loss": 1.2026,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1601
    },
    {
      "epoch": 1.0050188205771644,
      "grad_norm": 1.8686038255691528,
      "learning_rate": 2.914043443497627e-06,
      "loss": 1.1631,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1602
    },
    {
      "epoch": 1.0056461731493098,
      "grad_norm": 1.7626304626464844,
      "learning_rate": 2.9113444089350025e-06,
      "loss": 1.234,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1603
    },
    {
      "epoch": 1.0062735257214555,
      "grad_norm": 1.8479125499725342,
      "learning_rate": 2.908644881492024e-06,
      "loss": 1.2123,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1604
    },
    {
      "epoch": 1.006900878293601,
      "grad_norm": 1.9348218441009521,
      "learning_rate": 2.9059448644033157e-06,
      "loss": 1.1826,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1605
    },
    {
      "epoch": 1.0075282308657465,
      "grad_norm": 1.7956023216247559,
      "learning_rate": 2.903244360904086e-06,
      "loss": 1.1822,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1606
    },
    {
      "epoch": 1.0081555834378921,
      "grad_norm": 2.6866087913513184,
      "learning_rate": 2.9005433742301274e-06,
      "loss": 1.2719,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1607
    },
    {
      "epoch": 1.0087829360100375,
      "grad_norm": 1.9031373262405396,
      "learning_rate": 2.897841907617813e-06,
      "loss": 1.3587,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1608
    },
    {
      "epoch": 1.0094102885821832,
      "grad_norm": 1.678466796875,
      "learning_rate": 2.8951399643040867e-06,
      "loss": 1.3873,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1609
    },
    {
      "epoch": 1.0100376411543288,
      "grad_norm": 1.8993548154830933,
      "learning_rate": 2.892437547526468e-06,
      "loss": 1.3635,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1610
    },
    {
      "epoch": 1.0106649937264742,
      "grad_norm": 1.76960027217865,
      "learning_rate": 2.8897346605230403e-06,
      "loss": 1.2992,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1611
    },
    {
      "epoch": 1.0112923462986199,
      "grad_norm": 1.396877408027649,
      "learning_rate": 2.887031306532453e-06,
      "loss": 1.2675,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1612
    },
    {
      "epoch": 1.0119196988707653,
      "grad_norm": 1.9383916854858398,
      "learning_rate": 2.8843274887939136e-06,
      "loss": 1.3098,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1613
    },
    {
      "epoch": 1.012547051442911,
      "grad_norm": 1.8831377029418945,
      "learning_rate": 2.8816232105471864e-06,
      "loss": 1.2467,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1614
    },
    {
      "epoch": 1.0131744040150565,
      "grad_norm": 1.6349576711654663,
      "learning_rate": 2.878918475032586e-06,
      "loss": 1.2172,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1615
    },
    {
      "epoch": 1.013801756587202,
      "grad_norm": 1.781866431236267,
      "learning_rate": 2.876213285490976e-06,
      "loss": 1.2,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1616
    },
    {
      "epoch": 1.0144291091593476,
      "grad_norm": 1.9570376873016357,
      "learning_rate": 2.873507645163765e-06,
      "loss": 1.3187,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1617
    },
    {
      "epoch": 1.015056461731493,
      "grad_norm": 1.9935839176177979,
      "learning_rate": 2.8708015572928978e-06,
      "loss": 1.3265,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1618
    },
    {
      "epoch": 1.0156838143036386,
      "grad_norm": 6.492181777954102,
      "learning_rate": 2.8680950251208595e-06,
      "loss": 1.3201,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1619
    },
    {
      "epoch": 1.0163111668757843,
      "grad_norm": 1.873011589050293,
      "learning_rate": 2.8653880518906662e-06,
      "loss": 1.3138,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1620
    },
    {
      "epoch": 1.0169385194479297,
      "grad_norm": 1.8839685916900635,
      "learning_rate": 2.8626806408458626e-06,
      "loss": 1.2164,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1621
    },
    {
      "epoch": 1.0175658720200753,
      "grad_norm": 1.667243242263794,
      "learning_rate": 2.8599727952305173e-06,
      "loss": 1.4182,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1622
    },
    {
      "epoch": 1.0181932245922207,
      "grad_norm": 1.6265138387680054,
      "learning_rate": 2.8572645182892215e-06,
      "loss": 1.1193,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1623
    },
    {
      "epoch": 1.0188205771643664,
      "grad_norm": 1.573209524154663,
      "learning_rate": 2.8545558132670804e-06,
      "loss": 1.2317,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1624
    },
    {
      "epoch": 1.019447929736512,
      "grad_norm": 1.759530782699585,
      "learning_rate": 2.851846683409715e-06,
      "loss": 1.2441,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1625
    },
    {
      "epoch": 1.0200752823086574,
      "grad_norm": 1.9403883218765259,
      "learning_rate": 2.8491371319632515e-06,
      "loss": 1.288,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1626
    },
    {
      "epoch": 1.020702634880803,
      "grad_norm": 1.501510500907898,
      "learning_rate": 2.8464271621743266e-06,
      "loss": 1.2533,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1627
    },
    {
      "epoch": 1.0213299874529485,
      "grad_norm": 1.815144658088684,
      "learning_rate": 2.843716777290074e-06,
      "loss": 1.352,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1628
    },
    {
      "epoch": 1.021957340025094,
      "grad_norm": 1.590133547782898,
      "learning_rate": 2.8410059805581258e-06,
      "loss": 1.3409,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1629
    },
    {
      "epoch": 1.0225846925972397,
      "grad_norm": 1.507057547569275,
      "learning_rate": 2.8382947752266095e-06,
      "loss": 1.2165,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1630
    },
    {
      "epoch": 1.0232120451693851,
      "grad_norm": 1.6463191509246826,
      "learning_rate": 2.835583164544139e-06,
      "loss": 1.1584,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1631
    },
    {
      "epoch": 1.0238393977415308,
      "grad_norm": 1.7527943849563599,
      "learning_rate": 2.832871151759816e-06,
      "loss": 1.2905,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1632
    },
    {
      "epoch": 1.0244667503136762,
      "grad_norm": 1.900711178779602,
      "learning_rate": 2.830158740123225e-06,
      "loss": 1.3619,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1633
    },
    {
      "epoch": 1.0250941028858218,
      "grad_norm": 1.7869784832000732,
      "learning_rate": 2.827445932884425e-06,
      "loss": 1.3931,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1634
    },
    {
      "epoch": 1.0257214554579674,
      "grad_norm": 1.4502146244049072,
      "learning_rate": 2.8247327332939512e-06,
      "loss": 1.1494,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1635
    },
    {
      "epoch": 1.0263488080301129,
      "grad_norm": 2.0032737255096436,
      "learning_rate": 2.82201914460281e-06,
      "loss": 1.2799,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1636
    },
    {
      "epoch": 1.0269761606022585,
      "grad_norm": 1.9300953149795532,
      "learning_rate": 2.819305170062472e-06,
      "loss": 1.3849,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1637
    },
    {
      "epoch": 1.027603513174404,
      "grad_norm": 1.5697492361068726,
      "learning_rate": 2.81659081292487e-06,
      "loss": 1.2505,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1638
    },
    {
      "epoch": 1.0282308657465495,
      "grad_norm": 1.7658262252807617,
      "learning_rate": 2.813876076442397e-06,
      "loss": 1.2168,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1639
    },
    {
      "epoch": 1.0288582183186952,
      "grad_norm": 1.8125656843185425,
      "learning_rate": 2.8111609638679007e-06,
      "loss": 1.3612,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1640
    },
    {
      "epoch": 1.0294855708908406,
      "grad_norm": 1.9421466588974,
      "learning_rate": 2.808445478454677e-06,
      "loss": 1.2607,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1641
    },
    {
      "epoch": 1.0301129234629862,
      "grad_norm": 1.39006769657135,
      "learning_rate": 2.805729623456469e-06,
      "loss": 1.1229,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1642
    },
    {
      "epoch": 1.0307402760351319,
      "grad_norm": 1.8550273180007935,
      "learning_rate": 2.8030134021274656e-06,
      "loss": 1.1465,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1643
    },
    {
      "epoch": 1.0313676286072773,
      "grad_norm": 1.9449824094772339,
      "learning_rate": 2.8002968177222916e-06,
      "loss": 1.1624,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1644
    },
    {
      "epoch": 1.031994981179423,
      "grad_norm": 1.8409748077392578,
      "learning_rate": 2.7975798734960075e-06,
      "loss": 1.3491,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1645
    },
    {
      "epoch": 1.0326223337515683,
      "grad_norm": 1.7642316818237305,
      "learning_rate": 2.794862572704106e-06,
      "loss": 1.1807,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1646
    },
    {
      "epoch": 1.033249686323714,
      "grad_norm": 1.9541568756103516,
      "learning_rate": 2.7921449186025064e-06,
      "loss": 1.2875,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1647
    },
    {
      "epoch": 1.0338770388958596,
      "grad_norm": 1.905375599861145,
      "learning_rate": 2.7894269144475505e-06,
      "loss": 1.3454,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1648
    },
    {
      "epoch": 1.034504391468005,
      "grad_norm": 1.6375961303710938,
      "learning_rate": 2.786708563496002e-06,
      "loss": 1.4185,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1649
    },
    {
      "epoch": 1.0351317440401506,
      "grad_norm": 1.8116945028305054,
      "learning_rate": 2.783989869005036e-06,
      "loss": 1.3068,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1650
    },
    {
      "epoch": 1.035759096612296,
      "grad_norm": 1.7843427658081055,
      "learning_rate": 2.7812708342322435e-06,
      "loss": 1.2902,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1651
    },
    {
      "epoch": 1.0363864491844417,
      "grad_norm": 2.144293785095215,
      "learning_rate": 2.778551462435621e-06,
      "loss": 1.2698,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1652
    },
    {
      "epoch": 1.0370138017565873,
      "grad_norm": 1.6094316244125366,
      "learning_rate": 2.775831756873568e-06,
      "loss": 1.1651,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1653
    },
    {
      "epoch": 1.0376411543287327,
      "grad_norm": 1.9586728811264038,
      "learning_rate": 2.7731117208048875e-06,
      "loss": 1.3646,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1654
    },
    {
      "epoch": 1.0382685069008784,
      "grad_norm": 1.6644067764282227,
      "learning_rate": 2.770391357488775e-06,
      "loss": 1.2683,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1655
    },
    {
      "epoch": 1.0388958594730238,
      "grad_norm": 1.9723621606826782,
      "learning_rate": 2.7676706701848187e-06,
      "loss": 1.3013,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1656
    },
    {
      "epoch": 1.0395232120451694,
      "grad_norm": 1.8578646183013916,
      "learning_rate": 2.764949662152997e-06,
      "loss": 1.2815,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1657
    },
    {
      "epoch": 1.040150564617315,
      "grad_norm": 1.8444366455078125,
      "learning_rate": 2.76222833665367e-06,
      "loss": 1.2592,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1658
    },
    {
      "epoch": 1.0407779171894604,
      "grad_norm": 1.4603006839752197,
      "learning_rate": 2.75950669694758e-06,
      "loss": 1.2188,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1659
    },
    {
      "epoch": 1.041405269761606,
      "grad_norm": 1.6441221237182617,
      "learning_rate": 2.7567847462958453e-06,
      "loss": 1.2206,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1660
    },
    {
      "epoch": 1.0420326223337515,
      "grad_norm": 1.8188725709915161,
      "learning_rate": 2.754062487959956e-06,
      "loss": 1.1723,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1661
    },
    {
      "epoch": 1.0426599749058971,
      "grad_norm": 2.0048978328704834,
      "learning_rate": 2.751339925201772e-06,
      "loss": 1.3691,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1662
    },
    {
      "epoch": 1.0432873274780428,
      "grad_norm": 1.4602092504501343,
      "learning_rate": 2.748617061283518e-06,
      "loss": 1.2087,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1663
    },
    {
      "epoch": 1.0439146800501882,
      "grad_norm": 1.841833472251892,
      "learning_rate": 2.7458938994677784e-06,
      "loss": 1.2076,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1664
    },
    {
      "epoch": 1.0445420326223338,
      "grad_norm": 1.5950837135314941,
      "learning_rate": 2.7431704430174953e-06,
      "loss": 1.171,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1665
    },
    {
      "epoch": 1.0451693851944792,
      "grad_norm": 1.6101592779159546,
      "learning_rate": 2.740446695195964e-06,
      "loss": 1.2858,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1666
    },
    {
      "epoch": 1.0457967377666249,
      "grad_norm": 1.9419490098953247,
      "learning_rate": 2.737722659266829e-06,
      "loss": 1.3255,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1667
    },
    {
      "epoch": 1.0464240903387705,
      "grad_norm": 1.777923822402954,
      "learning_rate": 2.734998338494079e-06,
      "loss": 1.1948,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1668
    },
    {
      "epoch": 1.047051442910916,
      "grad_norm": 1.7209194898605347,
      "learning_rate": 2.7322737361420454e-06,
      "loss": 1.169,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1669
    },
    {
      "epoch": 1.0476787954830615,
      "grad_norm": 2.2611277103424072,
      "learning_rate": 2.7295488554753957e-06,
      "loss": 1.2104,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1670
    },
    {
      "epoch": 1.048306148055207,
      "grad_norm": 1.8065840005874634,
      "learning_rate": 2.7268236997591312e-06,
      "loss": 1.2578,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1671
    },
    {
      "epoch": 1.0489335006273526,
      "grad_norm": 1.9297280311584473,
      "learning_rate": 2.724098272258584e-06,
      "loss": 1.2406,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1672
    },
    {
      "epoch": 1.0495608531994982,
      "grad_norm": 1.824525237083435,
      "learning_rate": 2.7213725762394104e-06,
      "loss": 1.3392,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1673
    },
    {
      "epoch": 1.0501882057716436,
      "grad_norm": 1.7247780561447144,
      "learning_rate": 2.718646614967589e-06,
      "loss": 1.1674,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1674
    },
    {
      "epoch": 1.0508155583437893,
      "grad_norm": 1.902340054512024,
      "learning_rate": 2.7159203917094164e-06,
      "loss": 1.2874,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1675
    },
    {
      "epoch": 1.0514429109159347,
      "grad_norm": 1.7102277278900146,
      "learning_rate": 2.7131939097315023e-06,
      "loss": 1.2785,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1676
    },
    {
      "epoch": 1.0520702634880803,
      "grad_norm": 1.567852258682251,
      "learning_rate": 2.710467172300768e-06,
      "loss": 1.3513,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1677
    },
    {
      "epoch": 1.052697616060226,
      "grad_norm": 1.7976839542388916,
      "learning_rate": 2.7077401826844384e-06,
      "loss": 1.1563,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1678
    },
    {
      "epoch": 1.0533249686323714,
      "grad_norm": 1.5595132112503052,
      "learning_rate": 2.7050129441500437e-06,
      "loss": 1.0841,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1679
    },
    {
      "epoch": 1.053952321204517,
      "grad_norm": 1.9650431871414185,
      "learning_rate": 2.70228545996541e-06,
      "loss": 1.234,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1680
    },
    {
      "epoch": 1.0545796737766624,
      "grad_norm": 2.415435552597046,
      "learning_rate": 2.699557733398657e-06,
      "loss": 1.2938,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1681
    },
    {
      "epoch": 1.055207026348808,
      "grad_norm": 1.3821427822113037,
      "learning_rate": 2.696829767718199e-06,
      "loss": 1.2192,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1682
    },
    {
      "epoch": 1.0558343789209537,
      "grad_norm": 1.729189395904541,
      "learning_rate": 2.694101566192733e-06,
      "loss": 1.2821,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1683
    },
    {
      "epoch": 1.056461731493099,
      "grad_norm": 1.5959374904632568,
      "learning_rate": 2.69137313209124e-06,
      "loss": 1.1041,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1684
    },
    {
      "epoch": 1.0570890840652447,
      "grad_norm": 1.8131731748580933,
      "learning_rate": 2.688644468682978e-06,
      "loss": 1.3402,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1685
    },
    {
      "epoch": 1.0577164366373901,
      "grad_norm": 1.918359637260437,
      "learning_rate": 2.685915579237482e-06,
      "loss": 1.3129,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1686
    },
    {
      "epoch": 1.0583437892095358,
      "grad_norm": 1.7511471509933472,
      "learning_rate": 2.6831864670245573e-06,
      "loss": 1.2815,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1687
    },
    {
      "epoch": 1.0589711417816814,
      "grad_norm": 1.9037708044052124,
      "learning_rate": 2.680457135314275e-06,
      "loss": 1.2278,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1688
    },
    {
      "epoch": 1.0595984943538268,
      "grad_norm": 1.8454982042312622,
      "learning_rate": 2.6777275873769703e-06,
      "loss": 1.233,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1689
    },
    {
      "epoch": 1.0602258469259724,
      "grad_norm": 1.8991624116897583,
      "learning_rate": 2.674997826483239e-06,
      "loss": 1.3211,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1690
    },
    {
      "epoch": 1.0608531994981178,
      "grad_norm": 1.03837251663208,
      "learning_rate": 2.672267855903927e-06,
      "loss": 1.0668,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1691
    },
    {
      "epoch": 1.0614805520702635,
      "grad_norm": 1.8049486875534058,
      "learning_rate": 2.669537678910138e-06,
      "loss": 1.2028,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1692
    },
    {
      "epoch": 1.0621079046424091,
      "grad_norm": 1.9821325540542603,
      "learning_rate": 2.666807298773217e-06,
      "loss": 1.2478,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1693
    },
    {
      "epoch": 1.0627352572145545,
      "grad_norm": 1.344751000404358,
      "learning_rate": 2.664076718764756e-06,
      "loss": 1.115,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1694
    },
    {
      "epoch": 1.0633626097867002,
      "grad_norm": 2.035903215408325,
      "learning_rate": 2.661345942156586e-06,
      "loss": 1.2145,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1695
    },
    {
      "epoch": 1.0639899623588456,
      "grad_norm": 1.738403558731079,
      "learning_rate": 2.658614972220773e-06,
      "loss": 1.1725,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1696
    },
    {
      "epoch": 1.0646173149309912,
      "grad_norm": 1.6733629703521729,
      "learning_rate": 2.655883812229616e-06,
      "loss": 1.262,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1697
    },
    {
      "epoch": 1.0652446675031368,
      "grad_norm": 1.5742219686508179,
      "learning_rate": 2.653152465455639e-06,
      "loss": 1.06,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1698
    },
    {
      "epoch": 1.0658720200752823,
      "grad_norm": 1.934395670890808,
      "learning_rate": 2.6504209351715914e-06,
      "loss": 1.2199,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1699
    },
    {
      "epoch": 1.066499372647428,
      "grad_norm": 1.454143762588501,
      "learning_rate": 2.6476892246504438e-06,
      "loss": 1.0309,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1700
    },
    {
      "epoch": 1.0671267252195733,
      "grad_norm": 1.5390369892120361,
      "learning_rate": 2.6449573371653804e-06,
      "loss": 1.3156,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1701
    },
    {
      "epoch": 1.067754077791719,
      "grad_norm": 1.9746370315551758,
      "learning_rate": 2.642225275989798e-06,
      "loss": 1.2621,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1702
    },
    {
      "epoch": 1.0683814303638646,
      "grad_norm": 1.5100734233856201,
      "learning_rate": 2.6394930443973034e-06,
      "loss": 1.1465,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1703
    },
    {
      "epoch": 1.06900878293601,
      "grad_norm": 1.6106586456298828,
      "learning_rate": 2.6367606456617057e-06,
      "loss": 1.4355,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1704
    },
    {
      "epoch": 1.0696361355081556,
      "grad_norm": 1.8604997396469116,
      "learning_rate": 2.6340280830570142e-06,
      "loss": 1.2849,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1705
    },
    {
      "epoch": 1.070263488080301,
      "grad_norm": 1.9282605648040771,
      "learning_rate": 2.6312953598574353e-06,
      "loss": 1.238,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1706
    },
    {
      "epoch": 1.0708908406524467,
      "grad_norm": 2.088406801223755,
      "learning_rate": 2.6285624793373682e-06,
      "loss": 1.1732,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1707
    },
    {
      "epoch": 1.0715181932245923,
      "grad_norm": 1.6809911727905273,
      "learning_rate": 2.6258294447714e-06,
      "loss": 1.202,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1708
    },
    {
      "epoch": 1.0721455457967377,
      "grad_norm": 1.7849010229110718,
      "learning_rate": 2.6230962594343018e-06,
      "loss": 1.1261,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1709
    },
    {
      "epoch": 1.0727728983688833,
      "grad_norm": 1.7734084129333496,
      "learning_rate": 2.620362926601026e-06,
      "loss": 1.2661,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1710
    },
    {
      "epoch": 1.0734002509410288,
      "grad_norm": 1.7300902605056763,
      "learning_rate": 2.617629449546703e-06,
      "loss": 1.2376,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1711
    },
    {
      "epoch": 1.0740276035131744,
      "grad_norm": 1.3210912942886353,
      "learning_rate": 2.614895831546633e-06,
      "loss": 1.2699,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1712
    },
    {
      "epoch": 1.07465495608532,
      "grad_norm": 1.6365119218826294,
      "learning_rate": 2.6121620758762877e-06,
      "loss": 1.3032,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1713
    },
    {
      "epoch": 1.0752823086574654,
      "grad_norm": 1.5156607627868652,
      "learning_rate": 2.6094281858113026e-06,
      "loss": 1.233,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1714
    },
    {
      "epoch": 1.075909661229611,
      "grad_norm": 1.562544822692871,
      "learning_rate": 2.6066941646274748e-06,
      "loss": 1.2535,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1715
    },
    {
      "epoch": 1.0765370138017567,
      "grad_norm": 1.8988306522369385,
      "learning_rate": 2.603960015600759e-06,
      "loss": 1.2751,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1716
    },
    {
      "epoch": 1.0771643663739021,
      "grad_norm": 1.7881109714508057,
      "learning_rate": 2.60122574200726e-06,
      "loss": 1.166,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1717
    },
    {
      "epoch": 1.0777917189460477,
      "grad_norm": 1.5304796695709229,
      "learning_rate": 2.5984913471232363e-06,
      "loss": 1.1873,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1718
    },
    {
      "epoch": 1.0784190715181932,
      "grad_norm": 1.8655920028686523,
      "learning_rate": 2.595756834225089e-06,
      "loss": 1.3397,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1719
    },
    {
      "epoch": 1.0790464240903388,
      "grad_norm": 1.7003823518753052,
      "learning_rate": 2.5930222065893607e-06,
      "loss": 1.218,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1720
    },
    {
      "epoch": 1.0796737766624842,
      "grad_norm": 1.684619426727295,
      "learning_rate": 2.590287467492732e-06,
      "loss": 1.3297,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1721
    },
    {
      "epoch": 1.0803011292346298,
      "grad_norm": 2.0093002319335938,
      "learning_rate": 2.5875526202120175e-06,
      "loss": 1.1254,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1722
    },
    {
      "epoch": 1.0809284818067755,
      "grad_norm": 1.9766197204589844,
      "learning_rate": 2.584817668024161e-06,
      "loss": 1.2286,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1723
    },
    {
      "epoch": 1.0815558343789209,
      "grad_norm": 1.893913745880127,
      "learning_rate": 2.5820826142062323e-06,
      "loss": 1.2263,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1724
    },
    {
      "epoch": 1.0821831869510665,
      "grad_norm": 1.637317180633545,
      "learning_rate": 2.5793474620354213e-06,
      "loss": 1.2383,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1725
    },
    {
      "epoch": 1.0828105395232122,
      "grad_norm": 1.6308879852294922,
      "learning_rate": 2.576612214789039e-06,
      "loss": 1.2169,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1726
    },
    {
      "epoch": 1.0834378920953576,
      "grad_norm": 1.8625541925430298,
      "learning_rate": 2.573876875744506e-06,
      "loss": 1.392,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1727
    },
    {
      "epoch": 1.0840652446675032,
      "grad_norm": 1.7185273170471191,
      "learning_rate": 2.5711414481793574e-06,
      "loss": 1.1853,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1728
    },
    {
      "epoch": 1.0846925972396486,
      "grad_norm": 1.9088993072509766,
      "learning_rate": 2.568405935371231e-06,
      "loss": 1.1485,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1729
    },
    {
      "epoch": 1.0853199498117942,
      "grad_norm": 1.4835342168807983,
      "learning_rate": 2.5656703405978683e-06,
      "loss": 1.2852,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1730
    },
    {
      "epoch": 1.0859473023839397,
      "grad_norm": 1.8482460975646973,
      "learning_rate": 2.5629346671371098e-06,
      "loss": 1.2489,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1731
    },
    {
      "epoch": 1.0865746549560853,
      "grad_norm": 1.4996140003204346,
      "learning_rate": 2.5601989182668875e-06,
      "loss": 1.1504,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1732
    },
    {
      "epoch": 1.087202007528231,
      "grad_norm": 1.9409292936325073,
      "learning_rate": 2.5574630972652263e-06,
      "loss": 1.1184,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1733
    },
    {
      "epoch": 1.0878293601003763,
      "grad_norm": 1.8671555519104004,
      "learning_rate": 2.5547272074102375e-06,
      "loss": 1.2446,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1734
    },
    {
      "epoch": 1.088456712672522,
      "grad_norm": 1.867174506187439,
      "learning_rate": 2.551991251980112e-06,
      "loss": 1.1564,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1735
    },
    {
      "epoch": 1.0890840652446676,
      "grad_norm": 1.884575366973877,
      "learning_rate": 2.5492552342531234e-06,
      "loss": 1.2545,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1736
    },
    {
      "epoch": 1.089711417816813,
      "grad_norm": 1.6578123569488525,
      "learning_rate": 2.546519157507617e-06,
      "loss": 1.1447,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1737
    },
    {
      "epoch": 1.0903387703889587,
      "grad_norm": 1.6877325773239136,
      "learning_rate": 2.5437830250220104e-06,
      "loss": 1.2048,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1738
    },
    {
      "epoch": 1.090966122961104,
      "grad_norm": 1.7903735637664795,
      "learning_rate": 2.5410468400747858e-06,
      "loss": 1.4386,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1739
    },
    {
      "epoch": 1.0915934755332497,
      "grad_norm": 1.7975722551345825,
      "learning_rate": 2.538310605944491e-06,
      "loss": 1.2991,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1740
    },
    {
      "epoch": 1.0922208281053953,
      "grad_norm": 1.7802077531814575,
      "learning_rate": 2.5355743259097314e-06,
      "loss": 1.1507,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1741
    },
    {
      "epoch": 1.0928481806775407,
      "grad_norm": 1.5844789743423462,
      "learning_rate": 2.532838003249168e-06,
      "loss": 1.1518,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1742
    },
    {
      "epoch": 1.0934755332496864,
      "grad_norm": 1.7955514192581177,
      "learning_rate": 2.5301016412415107e-06,
      "loss": 1.1548,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1743
    },
    {
      "epoch": 1.0941028858218318,
      "grad_norm": 2.168617010116577,
      "learning_rate": 2.5273652431655204e-06,
      "loss": 1.2472,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1744
    },
    {
      "epoch": 1.0947302383939774,
      "grad_norm": 1.5748448371887207,
      "learning_rate": 2.524628812299997e-06,
      "loss": 1.3269,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1745
    },
    {
      "epoch": 1.095357590966123,
      "grad_norm": 1.7519515752792358,
      "learning_rate": 2.5218923519237824e-06,
      "loss": 1.1221,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1746
    },
    {
      "epoch": 1.0959849435382685,
      "grad_norm": 1.8812686204910278,
      "learning_rate": 2.5191558653157542e-06,
      "loss": 1.1561,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1747
    },
    {
      "epoch": 1.096612296110414,
      "grad_norm": 1.7000455856323242,
      "learning_rate": 2.5164193557548196e-06,
      "loss": 1.2085,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1748
    },
    {
      "epoch": 1.0972396486825595,
      "grad_norm": 1.8226401805877686,
      "learning_rate": 2.5136828265199143e-06,
      "loss": 1.2332,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1749
    },
    {
      "epoch": 1.0978670012547052,
      "grad_norm": 1.5405991077423096,
      "learning_rate": 2.5109462808899976e-06,
      "loss": 1.3213,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1750
    },
    {
      "epoch": 1.0984943538268508,
      "grad_norm": 1.548893928527832,
      "learning_rate": 2.5082097221440484e-06,
      "loss": 1.2114,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1751
    },
    {
      "epoch": 1.0991217063989962,
      "grad_norm": 1.9334415197372437,
      "learning_rate": 2.5054731535610593e-06,
      "loss": 1.3025,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1752
    },
    {
      "epoch": 1.0997490589711418,
      "grad_norm": 1.8565415143966675,
      "learning_rate": 2.502736578420039e-06,
      "loss": 1.2197,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1753
    },
    {
      "epoch": 1.1003764115432872,
      "grad_norm": 1.8411093950271606,
      "learning_rate": 2.5e-06,
      "loss": 1.3657,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1754
    },
    {
      "epoch": 1.1010037641154329,
      "grad_norm": 1.4905067682266235,
      "learning_rate": 2.497263421579962e-06,
      "loss": 1.1542,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1755
    },
    {
      "epoch": 1.1016311166875785,
      "grad_norm": 1.9230446815490723,
      "learning_rate": 2.494526846438941e-06,
      "loss": 1.3883,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1756
    },
    {
      "epoch": 1.102258469259724,
      "grad_norm": 1.8002456426620483,
      "learning_rate": 2.4917902778559533e-06,
      "loss": 1.2106,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1757
    },
    {
      "epoch": 1.1028858218318696,
      "grad_norm": 1.7526447772979736,
      "learning_rate": 2.4890537191100033e-06,
      "loss": 1.2373,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1758
    },
    {
      "epoch": 1.103513174404015,
      "grad_norm": 2.3764398097991943,
      "learning_rate": 2.4863171734800866e-06,
      "loss": 1.0688,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1759
    },
    {
      "epoch": 1.1041405269761606,
      "grad_norm": 1.8072677850723267,
      "learning_rate": 2.4835806442451804e-06,
      "loss": 1.436,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1760
    },
    {
      "epoch": 1.1047678795483062,
      "grad_norm": 2.030597686767578,
      "learning_rate": 2.480844134684246e-06,
      "loss": 1.2645,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1761
    },
    {
      "epoch": 1.1053952321204517,
      "grad_norm": 1.8442529439926147,
      "learning_rate": 2.4781076480762184e-06,
      "loss": 1.1999,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1762
    },
    {
      "epoch": 1.1060225846925973,
      "grad_norm": 1.9745938777923584,
      "learning_rate": 2.4753711877000035e-06,
      "loss": 1.2658,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1763
    },
    {
      "epoch": 1.1066499372647427,
      "grad_norm": 1.66330087184906,
      "learning_rate": 2.472634756834481e-06,
      "loss": 1.2338,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1764
    },
    {
      "epoch": 1.1072772898368883,
      "grad_norm": 1.705795168876648,
      "learning_rate": 2.4698983587584892e-06,
      "loss": 1.2796,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1765
    },
    {
      "epoch": 1.107904642409034,
      "grad_norm": 1.7829371690750122,
      "learning_rate": 2.467161996750833e-06,
      "loss": 1.3402,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1766
    },
    {
      "epoch": 1.1085319949811794,
      "grad_norm": 1.7036097049713135,
      "learning_rate": 2.464425674090269e-06,
      "loss": 1.2326,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1767
    },
    {
      "epoch": 1.109159347553325,
      "grad_norm": 1.9317264556884766,
      "learning_rate": 2.4616893940555094e-06,
      "loss": 1.4099,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1768
    },
    {
      "epoch": 1.1097867001254704,
      "grad_norm": 1.7943556308746338,
      "learning_rate": 2.4589531599252155e-06,
      "loss": 1.2372,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1769
    },
    {
      "epoch": 1.110414052697616,
      "grad_norm": 1.7894842624664307,
      "learning_rate": 2.4562169749779904e-06,
      "loss": 1.3813,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1770
    },
    {
      "epoch": 1.1110414052697617,
      "grad_norm": 1.698136568069458,
      "learning_rate": 2.4534808424923837e-06,
      "loss": 1.3245,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1771
    },
    {
      "epoch": 1.111668757841907,
      "grad_norm": 1.8109681606292725,
      "learning_rate": 2.4507447657468766e-06,
      "loss": 1.2317,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1772
    },
    {
      "epoch": 1.1122961104140527,
      "grad_norm": 1.8442978858947754,
      "learning_rate": 2.4480087480198884e-06,
      "loss": 1.2614,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1773
    },
    {
      "epoch": 1.1129234629861982,
      "grad_norm": 1.7043442726135254,
      "learning_rate": 2.4452727925897633e-06,
      "loss": 1.132,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1774
    },
    {
      "epoch": 1.1135508155583438,
      "grad_norm": 2.0383293628692627,
      "learning_rate": 2.4425369027347746e-06,
      "loss": 1.4807,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1775
    },
    {
      "epoch": 1.1141781681304894,
      "grad_norm": 1.718584418296814,
      "learning_rate": 2.4398010817331133e-06,
      "loss": 1.1544,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1776
    },
    {
      "epoch": 1.1148055207026348,
      "grad_norm": 1.7924760580062866,
      "learning_rate": 2.4370653328628915e-06,
      "loss": 1.4051,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1777
    },
    {
      "epoch": 1.1154328732747805,
      "grad_norm": 1.5274145603179932,
      "learning_rate": 2.4343296594021325e-06,
      "loss": 1.242,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1778
    },
    {
      "epoch": 1.1160602258469259,
      "grad_norm": 2.0505731105804443,
      "learning_rate": 2.4315940646287693e-06,
      "loss": 1.326,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1779
    },
    {
      "epoch": 1.1166875784190715,
      "grad_norm": 1.7353205680847168,
      "learning_rate": 2.4288585518206435e-06,
      "loss": 1.3196,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1780
    },
    {
      "epoch": 1.1173149309912171,
      "grad_norm": 1.5183053016662598,
      "learning_rate": 2.426123124255495e-06,
      "loss": 1.3235,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1781
    },
    {
      "epoch": 1.1179422835633626,
      "grad_norm": 1.6555873155593872,
      "learning_rate": 2.423387785210962e-06,
      "loss": 1.2756,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1782
    },
    {
      "epoch": 1.1185696361355082,
      "grad_norm": 1.7769240140914917,
      "learning_rate": 2.420652537964579e-06,
      "loss": 1.301,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1783
    },
    {
      "epoch": 1.1191969887076536,
      "grad_norm": 1.6840736865997314,
      "learning_rate": 2.4179173857937686e-06,
      "loss": 1.2422,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1784
    },
    {
      "epoch": 1.1198243412797992,
      "grad_norm": 1.6172568798065186,
      "learning_rate": 2.4151823319758397e-06,
      "loss": 1.2463,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1785
    },
    {
      "epoch": 1.1204516938519449,
      "grad_norm": 1.7393953800201416,
      "learning_rate": 2.4124473797879825e-06,
      "loss": 1.2352,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1786
    },
    {
      "epoch": 1.1210790464240903,
      "grad_norm": 1.824742317199707,
      "learning_rate": 2.4097125325072687e-06,
      "loss": 1.212,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1787
    },
    {
      "epoch": 1.121706398996236,
      "grad_norm": 1.6635518074035645,
      "learning_rate": 2.406977793410641e-06,
      "loss": 1.2657,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1788
    },
    {
      "epoch": 1.1223337515683813,
      "grad_norm": 1.4325731992721558,
      "learning_rate": 2.404243165774912e-06,
      "loss": 1.1163,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1789
    },
    {
      "epoch": 1.122961104140527,
      "grad_norm": 1.8488574028015137,
      "learning_rate": 2.4015086528767645e-06,
      "loss": 1.2844,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1790
    },
    {
      "epoch": 1.1235884567126726,
      "grad_norm": 1.8103573322296143,
      "learning_rate": 2.3987742579927407e-06,
      "loss": 1.3157,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1791
    },
    {
      "epoch": 1.124215809284818,
      "grad_norm": 1.4980573654174805,
      "learning_rate": 2.396039984399242e-06,
      "loss": 1.162,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1792
    },
    {
      "epoch": 1.1248431618569636,
      "grad_norm": 1.831099271774292,
      "learning_rate": 2.393305835372525e-06,
      "loss": 1.2391,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1793
    },
    {
      "epoch": 1.125470514429109,
      "grad_norm": 1.9267017841339111,
      "learning_rate": 2.390571814188698e-06,
      "loss": 1.3137,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1794
    },
    {
      "epoch": 1.1260978670012547,
      "grad_norm": 1.8979977369308472,
      "learning_rate": 2.3878379241237136e-06,
      "loss": 1.2354,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1795
    },
    {
      "epoch": 1.1267252195734003,
      "grad_norm": 2.1855428218841553,
      "learning_rate": 2.3851041684533677e-06,
      "loss": 1.1774,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1796
    },
    {
      "epoch": 1.1273525721455457,
      "grad_norm": 2.137327194213867,
      "learning_rate": 2.382370550453298e-06,
      "loss": 1.2398,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1797
    },
    {
      "epoch": 1.1279799247176914,
      "grad_norm": 1.9005476236343384,
      "learning_rate": 2.379637073398974e-06,
      "loss": 1.2664,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1798
    },
    {
      "epoch": 1.128607277289837,
      "grad_norm": 1.5104117393493652,
      "learning_rate": 2.376903740565699e-06,
      "loss": 1.0942,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1799
    },
    {
      "epoch": 1.1292346298619824,
      "grad_norm": 1.8482791185379028,
      "learning_rate": 2.3741705552286007e-06,
      "loss": 1.2907,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1800
    },
    {
      "epoch": 1.129861982434128,
      "grad_norm": 1.5305474996566772,
      "learning_rate": 2.3714375206626326e-06,
      "loss": 1.184,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1801
    },
    {
      "epoch": 1.1304893350062735,
      "grad_norm": 1.593786358833313,
      "learning_rate": 2.3687046401425656e-06,
      "loss": 1.3543,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1802
    },
    {
      "epoch": 1.131116687578419,
      "grad_norm": 1.7399150133132935,
      "learning_rate": 2.3659719169429866e-06,
      "loss": 1.2495,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1803
    },
    {
      "epoch": 1.1317440401505645,
      "grad_norm": 1.9040228128433228,
      "learning_rate": 2.363239354338295e-06,
      "loss": 1.2781,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1804
    },
    {
      "epoch": 1.1323713927227101,
      "grad_norm": 3.0083515644073486,
      "learning_rate": 2.3605069556026966e-06,
      "loss": 1.1329,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1805
    },
    {
      "epoch": 1.1329987452948558,
      "grad_norm": 1.74836003780365,
      "learning_rate": 2.3577747240102024e-06,
      "loss": 1.2894,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1806
    },
    {
      "epoch": 1.1336260978670012,
      "grad_norm": 1.5725035667419434,
      "learning_rate": 2.3550426628346204e-06,
      "loss": 1.2281,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1807
    },
    {
      "epoch": 1.1342534504391468,
      "grad_norm": 1.735289216041565,
      "learning_rate": 2.352310775349557e-06,
      "loss": 1.2272,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1808
    },
    {
      "epoch": 1.1348808030112925,
      "grad_norm": 1.5815335512161255,
      "learning_rate": 2.349579064828409e-06,
      "loss": 1.225,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1809
    },
    {
      "epoch": 1.1355081555834379,
      "grad_norm": 1.7958688735961914,
      "learning_rate": 2.346847534544362e-06,
      "loss": 1.2673,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1810
    },
    {
      "epoch": 1.1361355081555835,
      "grad_norm": 1.7950382232666016,
      "learning_rate": 2.3441161877703855e-06,
      "loss": 1.2082,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1811
    },
    {
      "epoch": 1.136762860727729,
      "grad_norm": 1.9151309728622437,
      "learning_rate": 2.341385027779227e-06,
      "loss": 1.2292,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1812
    },
    {
      "epoch": 1.1373902132998746,
      "grad_norm": 1.782690405845642,
      "learning_rate": 2.3386540578434147e-06,
      "loss": 1.3128,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1813
    },
    {
      "epoch": 1.13801756587202,
      "grad_norm": 1.810426950454712,
      "learning_rate": 2.3359232812352444e-06,
      "loss": 1.1123,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1814
    },
    {
      "epoch": 1.1386449184441656,
      "grad_norm": 1.3380026817321777,
      "learning_rate": 2.3331927012267842e-06,
      "loss": 1.1458,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1815
    },
    {
      "epoch": 1.1392722710163112,
      "grad_norm": 1.7149876356124878,
      "learning_rate": 2.3304623210898634e-06,
      "loss": 1.2083,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1816
    },
    {
      "epoch": 1.1398996235884566,
      "grad_norm": 2.052325963973999,
      "learning_rate": 2.3277321440960733e-06,
      "loss": 1.1781,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1817
    },
    {
      "epoch": 1.1405269761606023,
      "grad_norm": 1.4271595478057861,
      "learning_rate": 2.3250021735167624e-06,
      "loss": 1.1893,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1818
    },
    {
      "epoch": 1.141154328732748,
      "grad_norm": 1.9418270587921143,
      "learning_rate": 2.3222724126230296e-06,
      "loss": 1.2215,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1819
    },
    {
      "epoch": 1.1417816813048933,
      "grad_norm": 1.9351154565811157,
      "learning_rate": 2.3195428646857257e-06,
      "loss": 1.2943,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1820
    },
    {
      "epoch": 1.142409033877039,
      "grad_norm": 1.6721124649047852,
      "learning_rate": 2.316813532975443e-06,
      "loss": 1.1828,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1821
    },
    {
      "epoch": 1.1430363864491844,
      "grad_norm": 1.749272346496582,
      "learning_rate": 2.3140844207625188e-06,
      "loss": 1.2076,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1822
    },
    {
      "epoch": 1.14366373902133,
      "grad_norm": 1.8967795372009277,
      "learning_rate": 2.3113555313170234e-06,
      "loss": 1.3525,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1823
    },
    {
      "epoch": 1.1442910915934754,
      "grad_norm": 1.8196319341659546,
      "learning_rate": 2.308626867908761e-06,
      "loss": 1.1991,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1824
    },
    {
      "epoch": 1.144918444165621,
      "grad_norm": 1.7723875045776367,
      "learning_rate": 2.3058984338072672e-06,
      "loss": 1.3093,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1825
    },
    {
      "epoch": 1.1455457967377667,
      "grad_norm": 1.8782674074172974,
      "learning_rate": 2.3031702322818013e-06,
      "loss": 1.2351,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1826
    },
    {
      "epoch": 1.146173149309912,
      "grad_norm": 2.256856679916382,
      "learning_rate": 2.3004422666013432e-06,
      "loss": 1.2479,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1827
    },
    {
      "epoch": 1.1468005018820577,
      "grad_norm": 1.7046806812286377,
      "learning_rate": 2.2977145400345906e-06,
      "loss": 1.3064,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1828
    },
    {
      "epoch": 1.1474278544542034,
      "grad_norm": 1.9272924661636353,
      "learning_rate": 2.294987055849957e-06,
      "loss": 1.3769,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1829
    },
    {
      "epoch": 1.1480552070263488,
      "grad_norm": 1.8157531023025513,
      "learning_rate": 2.2922598173155624e-06,
      "loss": 1.2539,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1830
    },
    {
      "epoch": 1.1486825595984944,
      "grad_norm": 1.9448930025100708,
      "learning_rate": 2.2895328276992325e-06,
      "loss": 1.2955,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1831
    },
    {
      "epoch": 1.1493099121706398,
      "grad_norm": 1.8604695796966553,
      "learning_rate": 2.2868060902684986e-06,
      "loss": 1.2383,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1832
    },
    {
      "epoch": 1.1499372647427855,
      "grad_norm": 1.7356843948364258,
      "learning_rate": 2.284079608290584e-06,
      "loss": 1.2334,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1833
    },
    {
      "epoch": 1.1505646173149309,
      "grad_norm": 1.9711636304855347,
      "learning_rate": 2.281353385032412e-06,
      "loss": 1.1848,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1834
    },
    {
      "epoch": 1.1511919698870765,
      "grad_norm": 2.1687605381011963,
      "learning_rate": 2.2786274237605904e-06,
      "loss": 1.2311,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1835
    },
    {
      "epoch": 1.1518193224592221,
      "grad_norm": 1.8387099504470825,
      "learning_rate": 2.2759017277414165e-06,
      "loss": 1.2324,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1836
    },
    {
      "epoch": 1.1524466750313676,
      "grad_norm": 1.8152990341186523,
      "learning_rate": 2.2731763002408696e-06,
      "loss": 1.2585,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1837
    },
    {
      "epoch": 1.1530740276035132,
      "grad_norm": 1.7368495464324951,
      "learning_rate": 2.270451144524605e-06,
      "loss": 1.1628,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1838
    },
    {
      "epoch": 1.1537013801756588,
      "grad_norm": 1.7837342023849487,
      "learning_rate": 2.2677262638579554e-06,
      "loss": 1.3714,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1839
    },
    {
      "epoch": 1.1543287327478042,
      "grad_norm": 1.7108197212219238,
      "learning_rate": 2.265001661505921e-06,
      "loss": 1.2377,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1840
    },
    {
      "epoch": 1.1549560853199499,
      "grad_norm": 1.7582619190216064,
      "learning_rate": 2.2622773407331713e-06,
      "loss": 1.1935,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1841
    },
    {
      "epoch": 1.1555834378920953,
      "grad_norm": 1.199617624282837,
      "learning_rate": 2.2595533048040362e-06,
      "loss": 1.134,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1842
    },
    {
      "epoch": 1.156210790464241,
      "grad_norm": 1.9170664548873901,
      "learning_rate": 2.256829556982505e-06,
      "loss": 1.3591,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1843
    },
    {
      "epoch": 1.1568381430363865,
      "grad_norm": 1.8503596782684326,
      "learning_rate": 2.254106100532223e-06,
      "loss": 1.1394,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1844
    },
    {
      "epoch": 1.157465495608532,
      "grad_norm": 1.9844796657562256,
      "learning_rate": 2.251382938716482e-06,
      "loss": 1.2423,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1845
    },
    {
      "epoch": 1.1580928481806776,
      "grad_norm": 1.7377749681472778,
      "learning_rate": 2.2486600747982286e-06,
      "loss": 1.1889,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1846
    },
    {
      "epoch": 1.158720200752823,
      "grad_norm": 1.7416249513626099,
      "learning_rate": 2.2459375120400444e-06,
      "loss": 1.2454,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1847
    },
    {
      "epoch": 1.1593475533249686,
      "grad_norm": 1.5563774108886719,
      "learning_rate": 2.2432152537041556e-06,
      "loss": 1.242,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1848
    },
    {
      "epoch": 1.1599749058971143,
      "grad_norm": 2.0742273330688477,
      "learning_rate": 2.240493303052421e-06,
      "loss": 1.2014,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1849
    },
    {
      "epoch": 1.1606022584692597,
      "grad_norm": 1.8404746055603027,
      "learning_rate": 2.237771663346331e-06,
      "loss": 1.2473,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1850
    },
    {
      "epoch": 1.1612296110414053,
      "grad_norm": 1.564107060432434,
      "learning_rate": 2.235050337847004e-06,
      "loss": 1.1768,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1851
    },
    {
      "epoch": 1.1618569636135507,
      "grad_norm": 1.5945340394973755,
      "learning_rate": 2.2323293298151817e-06,
      "loss": 1.1811,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1852
    },
    {
      "epoch": 1.1624843161856964,
      "grad_norm": 1.763053059577942,
      "learning_rate": 2.229608642511226e-06,
      "loss": 1.1327,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1853
    },
    {
      "epoch": 1.163111668757842,
      "grad_norm": 1.7052210569381714,
      "learning_rate": 2.2268882791951125e-06,
      "loss": 1.2935,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1854
    },
    {
      "epoch": 1.1637390213299874,
      "grad_norm": 2.095878839492798,
      "learning_rate": 2.2241682431264323e-06,
      "loss": 1.2472,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1855
    },
    {
      "epoch": 1.164366373902133,
      "grad_norm": 2.009671926498413,
      "learning_rate": 2.2214485375643804e-06,
      "loss": 1.4277,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1856
    },
    {
      "epoch": 1.1649937264742785,
      "grad_norm": 1.7650864124298096,
      "learning_rate": 2.218729165767757e-06,
      "loss": 1.3452,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1857
    },
    {
      "epoch": 1.165621079046424,
      "grad_norm": 1.6924291849136353,
      "learning_rate": 2.216010130994965e-06,
      "loss": 1.0748,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1858
    },
    {
      "epoch": 1.1662484316185697,
      "grad_norm": 1.9703468084335327,
      "learning_rate": 2.2132914365039993e-06,
      "loss": 1.3066,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1859
    },
    {
      "epoch": 1.1668757841907151,
      "grad_norm": 1.3924521207809448,
      "learning_rate": 2.2105730855524503e-06,
      "loss": 1.3009,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1860
    },
    {
      "epoch": 1.1675031367628608,
      "grad_norm": 1.778253197669983,
      "learning_rate": 2.207855081397494e-06,
      "loss": 1.1961,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1861
    },
    {
      "epoch": 1.1681304893350062,
      "grad_norm": 1.7997092008590698,
      "learning_rate": 2.2051374272958946e-06,
      "loss": 1.2506,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1862
    },
    {
      "epoch": 1.1687578419071518,
      "grad_norm": 1.8622349500656128,
      "learning_rate": 2.2024201265039933e-06,
      "loss": 1.2358,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1863
    },
    {
      "epoch": 1.1693851944792975,
      "grad_norm": 1.6936147212982178,
      "learning_rate": 2.1997031822777093e-06,
      "loss": 1.3494,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1864
    },
    {
      "epoch": 1.1700125470514429,
      "grad_norm": 1.8487699031829834,
      "learning_rate": 2.1969865978725353e-06,
      "loss": 1.2655,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1865
    },
    {
      "epoch": 1.1706398996235885,
      "grad_norm": 1.7117902040481567,
      "learning_rate": 2.1942703765435317e-06,
      "loss": 1.1697,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1866
    },
    {
      "epoch": 1.171267252195734,
      "grad_norm": 1.899256944656372,
      "learning_rate": 2.1915545215453245e-06,
      "loss": 1.3501,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1867
    },
    {
      "epoch": 1.1718946047678795,
      "grad_norm": 1.7100666761398315,
      "learning_rate": 2.1888390361320997e-06,
      "loss": 1.187,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1868
    },
    {
      "epoch": 1.1725219573400252,
      "grad_norm": 1.6679086685180664,
      "learning_rate": 2.1861239235576033e-06,
      "loss": 1.1233,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1869
    },
    {
      "epoch": 1.1731493099121706,
      "grad_norm": 1.8351097106933594,
      "learning_rate": 2.183409187075131e-06,
      "loss": 1.2701,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1870
    },
    {
      "epoch": 1.1737766624843162,
      "grad_norm": 1.9863923788070679,
      "learning_rate": 2.180694829937529e-06,
      "loss": 1.3121,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1871
    },
    {
      "epoch": 1.1744040150564616,
      "grad_norm": 1.8615024089813232,
      "learning_rate": 2.177980855397191e-06,
      "loss": 1.1982,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1872
    },
    {
      "epoch": 1.1750313676286073,
      "grad_norm": 1.921243667602539,
      "learning_rate": 2.1752672667060488e-06,
      "loss": 1.244,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1873
    },
    {
      "epoch": 1.175658720200753,
      "grad_norm": 1.8630694150924683,
      "learning_rate": 2.172554067115576e-06,
      "loss": 1.3146,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1874
    },
    {
      "epoch": 1.1762860727728983,
      "grad_norm": 1.8418269157409668,
      "learning_rate": 2.169841259876776e-06,
      "loss": 1.2563,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1875
    },
    {
      "epoch": 1.176913425345044,
      "grad_norm": 1.811339259147644,
      "learning_rate": 2.1671288482401842e-06,
      "loss": 1.2523,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1876
    },
    {
      "epoch": 1.1775407779171894,
      "grad_norm": 1.922473430633545,
      "learning_rate": 2.1644168354558623e-06,
      "loss": 1.1695,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1877
    },
    {
      "epoch": 1.178168130489335,
      "grad_norm": 1.6754716634750366,
      "learning_rate": 2.161705224773391e-06,
      "loss": 1.1444,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1878
    },
    {
      "epoch": 1.1787954830614806,
      "grad_norm": 2.2326772212982178,
      "learning_rate": 2.158994019441875e-06,
      "loss": 1.1346,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1879
    },
    {
      "epoch": 1.179422835633626,
      "grad_norm": 1.7165148258209229,
      "learning_rate": 2.1562832227099266e-06,
      "loss": 1.1899,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1880
    },
    {
      "epoch": 1.1800501882057717,
      "grad_norm": 1.8452473878860474,
      "learning_rate": 2.1535728378256742e-06,
      "loss": 1.1952,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1881
    },
    {
      "epoch": 1.1806775407779173,
      "grad_norm": 1.6404380798339844,
      "learning_rate": 2.150862868036749e-06,
      "loss": 1.2802,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1882
    },
    {
      "epoch": 1.1813048933500627,
      "grad_norm": 2.0825002193450928,
      "learning_rate": 2.1481533165902863e-06,
      "loss": 1.2548,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1883
    },
    {
      "epoch": 1.1819322459222084,
      "grad_norm": 1.647680401802063,
      "learning_rate": 2.1454441867329205e-06,
      "loss": 1.1099,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1884
    },
    {
      "epoch": 1.1825595984943538,
      "grad_norm": 1.8102099895477295,
      "learning_rate": 2.1427354817107793e-06,
      "loss": 1.2962,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1885
    },
    {
      "epoch": 1.1831869510664994,
      "grad_norm": 1.9231828451156616,
      "learning_rate": 2.1400272047694835e-06,
      "loss": 1.2266,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1886
    },
    {
      "epoch": 1.1838143036386448,
      "grad_norm": 1.8153226375579834,
      "learning_rate": 2.137319359154138e-06,
      "loss": 1.2095,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1887
    },
    {
      "epoch": 1.1844416562107905,
      "grad_norm": 1.8013314008712769,
      "learning_rate": 2.1346119481093346e-06,
      "loss": 1.1613,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1888
    },
    {
      "epoch": 1.185069008782936,
      "grad_norm": 1.8556623458862305,
      "learning_rate": 2.1319049748791418e-06,
      "loss": 1.2125,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1889
    },
    {
      "epoch": 1.1856963613550815,
      "grad_norm": 1.7150856256484985,
      "learning_rate": 2.129198442707103e-06,
      "loss": 1.2205,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1890
    },
    {
      "epoch": 1.1863237139272271,
      "grad_norm": 1.7657842636108398,
      "learning_rate": 2.1264923548362366e-06,
      "loss": 1.3468,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1891
    },
    {
      "epoch": 1.1869510664993728,
      "grad_norm": 1.9392738342285156,
      "learning_rate": 2.1237867145090242e-06,
      "loss": 1.3777,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1892
    },
    {
      "epoch": 1.1875784190715182,
      "grad_norm": 1.9274314641952515,
      "learning_rate": 2.1210815249674148e-06,
      "loss": 1.2644,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1893
    },
    {
      "epoch": 1.1882057716436638,
      "grad_norm": 1.9306237697601318,
      "learning_rate": 2.1183767894528135e-06,
      "loss": 1.3857,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1894
    },
    {
      "epoch": 1.1888331242158092,
      "grad_norm": 1.8561091423034668,
      "learning_rate": 2.1156725112060868e-06,
      "loss": 1.203,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1895
    },
    {
      "epoch": 1.1894604767879549,
      "grad_norm": 2.7982521057128906,
      "learning_rate": 2.112968693467548e-06,
      "loss": 1.2022,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1896
    },
    {
      "epoch": 1.1900878293601003,
      "grad_norm": 1.6578912734985352,
      "learning_rate": 2.1102653394769605e-06,
      "loss": 1.2315,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1897
    },
    {
      "epoch": 1.190715181932246,
      "grad_norm": 1.6461844444274902,
      "learning_rate": 2.1075624524735335e-06,
      "loss": 1.2014,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1898
    },
    {
      "epoch": 1.1913425345043915,
      "grad_norm": 1.4624578952789307,
      "learning_rate": 2.1048600356959133e-06,
      "loss": 1.1935,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1899
    },
    {
      "epoch": 1.191969887076537,
      "grad_norm": 1.8360897302627563,
      "learning_rate": 2.1021580923821876e-06,
      "loss": 1.2885,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1900
    },
    {
      "epoch": 1.1925972396486826,
      "grad_norm": 1.8737788200378418,
      "learning_rate": 2.099456625769872e-06,
      "loss": 1.1497,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1901
    },
    {
      "epoch": 1.1932245922208282,
      "grad_norm": 1.8200892210006714,
      "learning_rate": 2.0967556390959147e-06,
      "loss": 1.2012,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1902
    },
    {
      "epoch": 1.1938519447929736,
      "grad_norm": 1.9442538022994995,
      "learning_rate": 2.0940551355966856e-06,
      "loss": 1.5149,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1903
    },
    {
      "epoch": 1.1944792973651193,
      "grad_norm": 1.9775135517120361,
      "learning_rate": 2.0913551185079763e-06,
      "loss": 1.2177,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1904
    },
    {
      "epoch": 1.1951066499372647,
      "grad_norm": 1.7279037237167358,
      "learning_rate": 2.0886555910649987e-06,
      "loss": 1.2819,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1905
    },
    {
      "epoch": 1.1957340025094103,
      "grad_norm": 1.5370543003082275,
      "learning_rate": 2.0859565565023735e-06,
      "loss": 1.3726,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1906
    },
    {
      "epoch": 1.1963613550815557,
      "grad_norm": 1.768170952796936,
      "learning_rate": 2.0832580180541356e-06,
      "loss": 1.1752,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1907
    },
    {
      "epoch": 1.1969887076537014,
      "grad_norm": 1.8252557516098022,
      "learning_rate": 2.08055997895372e-06,
      "loss": 1.4465,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1908
    },
    {
      "epoch": 1.197616060225847,
      "grad_norm": 1.8621381521224976,
      "learning_rate": 2.0778624424339684e-06,
      "loss": 1.2775,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1909
    },
    {
      "epoch": 1.1982434127979924,
      "grad_norm": 2.329472780227661,
      "learning_rate": 2.0751654117271177e-06,
      "loss": 1.3053,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1910
    },
    {
      "epoch": 1.198870765370138,
      "grad_norm": 1.8072152137756348,
      "learning_rate": 2.0724688900647973e-06,
      "loss": 1.1612,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1911
    },
    {
      "epoch": 1.1994981179422837,
      "grad_norm": 1.9142813682556152,
      "learning_rate": 2.0697728806780308e-06,
      "loss": 1.2612,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1912
    },
    {
      "epoch": 1.200125470514429,
      "grad_norm": 1.831708550453186,
      "learning_rate": 2.0670773867972236e-06,
      "loss": 1.1592,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1913
    },
    {
      "epoch": 1.2007528230865747,
      "grad_norm": 1.7283164262771606,
      "learning_rate": 2.0643824116521683e-06,
      "loss": 1.2807,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1914
    },
    {
      "epoch": 1.2013801756587201,
      "grad_norm": 1.6546218395233154,
      "learning_rate": 2.0616879584720305e-06,
      "loss": 1.3963,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1915
    },
    {
      "epoch": 1.2020075282308658,
      "grad_norm": 1.7884093523025513,
      "learning_rate": 2.0589940304853552e-06,
      "loss": 1.1979,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1916
    },
    {
      "epoch": 1.2026348808030112,
      "grad_norm": 1.6815849542617798,
      "learning_rate": 2.0563006309200545e-06,
      "loss": 1.2687,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1917
    },
    {
      "epoch": 1.2032622333751568,
      "grad_norm": 1.7797585725784302,
      "learning_rate": 2.053607763003409e-06,
      "loss": 1.1443,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1918
    },
    {
      "epoch": 1.2038895859473024,
      "grad_norm": 1.181809663772583,
      "learning_rate": 2.0509154299620622e-06,
      "loss": 1.2201,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1919
    },
    {
      "epoch": 1.2045169385194479,
      "grad_norm": 1.8268485069274902,
      "learning_rate": 2.0482236350220152e-06,
      "loss": 1.2838,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1920
    },
    {
      "epoch": 1.2051442910915935,
      "grad_norm": 1.9953694343566895,
      "learning_rate": 2.0455323814086284e-06,
      "loss": 1.1425,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1921
    },
    {
      "epoch": 1.2057716436637391,
      "grad_norm": 1.6402983665466309,
      "learning_rate": 2.042841672346608e-06,
      "loss": 1.1408,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1922
    },
    {
      "epoch": 1.2063989962358845,
      "grad_norm": 1.7828822135925293,
      "learning_rate": 2.040151511060013e-06,
      "loss": 1.2727,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1923
    },
    {
      "epoch": 1.2070263488080302,
      "grad_norm": 1.570185899734497,
      "learning_rate": 2.0374619007722423e-06,
      "loss": 1.2971,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1924
    },
    {
      "epoch": 1.2076537013801756,
      "grad_norm": 1.9676752090454102,
      "learning_rate": 2.034772844706036e-06,
      "loss": 1.3004,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1925
    },
    {
      "epoch": 1.2082810539523212,
      "grad_norm": 1.4364757537841797,
      "learning_rate": 2.0320843460834698e-06,
      "loss": 1.1709,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1926
    },
    {
      "epoch": 1.2089084065244666,
      "grad_norm": 1.668623685836792,
      "learning_rate": 2.0293964081259514e-06,
      "loss": 1.2781,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1927
    },
    {
      "epoch": 1.2095357590966123,
      "grad_norm": 1.8750041723251343,
      "learning_rate": 2.026709034054218e-06,
      "loss": 1.2256,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1928
    },
    {
      "epoch": 1.210163111668758,
      "grad_norm": 2.012606620788574,
      "learning_rate": 2.024022227088329e-06,
      "loss": 1.228,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1929
    },
    {
      "epoch": 1.2107904642409033,
      "grad_norm": 1.8182164430618286,
      "learning_rate": 2.0213359904476666e-06,
      "loss": 1.2232,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1930
    },
    {
      "epoch": 1.211417816813049,
      "grad_norm": 1.8466030359268188,
      "learning_rate": 2.0186503273509276e-06,
      "loss": 1.2221,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1931
    },
    {
      "epoch": 1.2120451693851946,
      "grad_norm": 1.7908083200454712,
      "learning_rate": 2.015965241016122e-06,
      "loss": 1.1342,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1932
    },
    {
      "epoch": 1.21267252195734,
      "grad_norm": 1.8812059164047241,
      "learning_rate": 2.01328073466057e-06,
      "loss": 1.2676,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1933
    },
    {
      "epoch": 1.2132998745294856,
      "grad_norm": 1.8390413522720337,
      "learning_rate": 2.0105968115008957e-06,
      "loss": 1.2211,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1934
    },
    {
      "epoch": 1.213927227101631,
      "grad_norm": 1.9293330907821655,
      "learning_rate": 2.0079134747530244e-06,
      "loss": 1.2886,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1935
    },
    {
      "epoch": 1.2145545796737767,
      "grad_norm": 1.824219822883606,
      "learning_rate": 2.0052307276321793e-06,
      "loss": 1.305,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1936
    },
    {
      "epoch": 1.2151819322459223,
      "grad_norm": 1.8564656972885132,
      "learning_rate": 2.0025485733528754e-06,
      "loss": 1.2049,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1937
    },
    {
      "epoch": 1.2158092848180677,
      "grad_norm": 1.8456517457962036,
      "learning_rate": 1.9998670151289214e-06,
      "loss": 1.3585,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1938
    },
    {
      "epoch": 1.2164366373902133,
      "grad_norm": 1.81951904296875,
      "learning_rate": 1.9971860561734062e-06,
      "loss": 1.1953,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1939
    },
    {
      "epoch": 1.2170639899623588,
      "grad_norm": 1.8625571727752686,
      "learning_rate": 1.994505699698706e-06,
      "loss": 1.2339,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1940
    },
    {
      "epoch": 1.2176913425345044,
      "grad_norm": 1.8038089275360107,
      "learning_rate": 1.9918259489164717e-06,
      "loss": 1.2464,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1941
    },
    {
      "epoch": 1.21831869510665,
      "grad_norm": 1.580041766166687,
      "learning_rate": 1.9891468070376297e-06,
      "loss": 1.1855,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1942
    },
    {
      "epoch": 1.2189460476787954,
      "grad_norm": 1.5931955575942993,
      "learning_rate": 1.9864682772723757e-06,
      "loss": 1.2787,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1943
    },
    {
      "epoch": 1.219573400250941,
      "grad_norm": 1.7477593421936035,
      "learning_rate": 1.983790362830174e-06,
      "loss": 1.3233,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1944
    },
    {
      "epoch": 1.2202007528230865,
      "grad_norm": 1.6215864419937134,
      "learning_rate": 1.9811130669197496e-06,
      "loss": 1.2165,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1945
    },
    {
      "epoch": 1.2208281053952321,
      "grad_norm": 1.593785047531128,
      "learning_rate": 1.978436392749087e-06,
      "loss": 1.1529,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1946
    },
    {
      "epoch": 1.2214554579673778,
      "grad_norm": 1.6797493696212769,
      "learning_rate": 1.975760343525427e-06,
      "loss": 1.227,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1947
    },
    {
      "epoch": 1.2220828105395232,
      "grad_norm": 2.1317989826202393,
      "learning_rate": 1.9730849224552586e-06,
      "loss": 1.1288,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1948
    },
    {
      "epoch": 1.2227101631116688,
      "grad_norm": 1.771586537361145,
      "learning_rate": 1.970410132744322e-06,
      "loss": 1.2009,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1949
    },
    {
      "epoch": 1.2233375156838142,
      "grad_norm": 1.916454553604126,
      "learning_rate": 1.967735977597598e-06,
      "loss": 1.292,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1950
    },
    {
      "epoch": 1.2239648682559598,
      "grad_norm": 1.910947561264038,
      "learning_rate": 1.965062460219308e-06,
      "loss": 1.2213,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1951
    },
    {
      "epoch": 1.2245922208281055,
      "grad_norm": 1.6686664819717407,
      "learning_rate": 1.9623895838129094e-06,
      "loss": 1.2531,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1952
    },
    {
      "epoch": 1.225219573400251,
      "grad_norm": 2.0595884323120117,
      "learning_rate": 1.9597173515810896e-06,
      "loss": 1.0912,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1953
    },
    {
      "epoch": 1.2258469259723965,
      "grad_norm": 1.720428705215454,
      "learning_rate": 1.9570457667257686e-06,
      "loss": 1.137,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1954
    },
    {
      "epoch": 1.226474278544542,
      "grad_norm": 1.614362120628357,
      "learning_rate": 1.9543748324480856e-06,
      "loss": 1.207,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1955
    },
    {
      "epoch": 1.2271016311166876,
      "grad_norm": 1.9100805521011353,
      "learning_rate": 1.9517045519484046e-06,
      "loss": 1.2895,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1956
    },
    {
      "epoch": 1.2277289836888332,
      "grad_norm": 2.1220719814300537,
      "learning_rate": 1.9490349284263036e-06,
      "loss": 1.2398,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1957
    },
    {
      "epoch": 1.2283563362609786,
      "grad_norm": 1.9832006692886353,
      "learning_rate": 1.946365965080573e-06,
      "loss": 1.2474,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1958
    },
    {
      "epoch": 1.2289836888331243,
      "grad_norm": 1.8225581645965576,
      "learning_rate": 1.9436976651092143e-06,
      "loss": 1.3338,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1959
    },
    {
      "epoch": 1.2296110414052697,
      "grad_norm": 1.7378026247024536,
      "learning_rate": 1.9410300317094328e-06,
      "loss": 1.2641,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1960
    },
    {
      "epoch": 1.2302383939774153,
      "grad_norm": 1.4091269969940186,
      "learning_rate": 1.938363068077636e-06,
      "loss": 1.1238,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1961
    },
    {
      "epoch": 1.230865746549561,
      "grad_norm": 1.8257181644439697,
      "learning_rate": 1.9356967774094265e-06,
      "loss": 1.2355,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1962
    },
    {
      "epoch": 1.2314930991217063,
      "grad_norm": 1.5688260793685913,
      "learning_rate": 1.9330311628996046e-06,
      "loss": 1.2014,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1963
    },
    {
      "epoch": 1.232120451693852,
      "grad_norm": 2.238872528076172,
      "learning_rate": 1.930366227742157e-06,
      "loss": 1.2893,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1964
    },
    {
      "epoch": 1.2327478042659976,
      "grad_norm": 1.6153379678726196,
      "learning_rate": 1.9277019751302557e-06,
      "loss": 1.202,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1965
    },
    {
      "epoch": 1.233375156838143,
      "grad_norm": 1.7979259490966797,
      "learning_rate": 1.9250384082562603e-06,
      "loss": 1.2252,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1966
    },
    {
      "epoch": 1.2340025094102887,
      "grad_norm": 1.8436015844345093,
      "learning_rate": 1.922375530311702e-06,
      "loss": 1.1568,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1967
    },
    {
      "epoch": 1.234629861982434,
      "grad_norm": 1.7746145725250244,
      "learning_rate": 1.919713344487291e-06,
      "loss": 1.1825,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1968
    },
    {
      "epoch": 1.2352572145545797,
      "grad_norm": 1.67526113986969,
      "learning_rate": 1.9170518539729063e-06,
      "loss": 1.1145,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1969
    },
    {
      "epoch": 1.2358845671267251,
      "grad_norm": 1.6781444549560547,
      "learning_rate": 1.9143910619575952e-06,
      "loss": 1.1309,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1970
    },
    {
      "epoch": 1.2365119196988708,
      "grad_norm": 1.8381301164627075,
      "learning_rate": 1.9117309716295658e-06,
      "loss": 1.297,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1971
    },
    {
      "epoch": 1.2371392722710164,
      "grad_norm": 1.8952548503875732,
      "learning_rate": 1.9090715861761868e-06,
      "loss": 1.3334,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1972
    },
    {
      "epoch": 1.2377666248431618,
      "grad_norm": 1.9088690280914307,
      "learning_rate": 1.9064129087839833e-06,
      "loss": 1.2053,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1973
    },
    {
      "epoch": 1.2383939774153074,
      "grad_norm": 1.90237557888031,
      "learning_rate": 1.9037549426386304e-06,
      "loss": 1.2269,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1974
    },
    {
      "epoch": 1.239021329987453,
      "grad_norm": 1.8582751750946045,
      "learning_rate": 1.9010976909249506e-06,
      "loss": 1.3578,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1975
    },
    {
      "epoch": 1.2396486825595985,
      "grad_norm": 1.7843849658966064,
      "learning_rate": 1.8984411568269126e-06,
      "loss": 1.2032,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1976
    },
    {
      "epoch": 1.2402760351317441,
      "grad_norm": 1.9194496870040894,
      "learning_rate": 1.8957853435276238e-06,
      "loss": 1.211,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1977
    },
    {
      "epoch": 1.2409033877038895,
      "grad_norm": 1.8976136445999146,
      "learning_rate": 1.8931302542093274e-06,
      "loss": 1.1284,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1978
    },
    {
      "epoch": 1.2415307402760352,
      "grad_norm": 1.9729543924331665,
      "learning_rate": 1.8904758920533988e-06,
      "loss": 1.2495,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1979
    },
    {
      "epoch": 1.2421580928481806,
      "grad_norm": 1.7914001941680908,
      "learning_rate": 1.8878222602403452e-06,
      "loss": 1.3332,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1980
    },
    {
      "epoch": 1.2427854454203262,
      "grad_norm": 1.4101675748825073,
      "learning_rate": 1.885169361949794e-06,
      "loss": 1.2431,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1981
    },
    {
      "epoch": 1.2434127979924718,
      "grad_norm": 1.812052845954895,
      "learning_rate": 1.8825172003604988e-06,
      "loss": 1.2351,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1982
    },
    {
      "epoch": 1.2440401505646173,
      "grad_norm": 1.7212128639221191,
      "learning_rate": 1.879865778650326e-06,
      "loss": 1.4615,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1983
    },
    {
      "epoch": 1.2446675031367629,
      "grad_norm": 1.90878164768219,
      "learning_rate": 1.8772150999962588e-06,
      "loss": 1.3332,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1984
    },
    {
      "epoch": 1.2452948557089085,
      "grad_norm": 1.833099365234375,
      "learning_rate": 1.8745651675743876e-06,
      "loss": 1.268,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1985
    },
    {
      "epoch": 1.245922208281054,
      "grad_norm": 1.850371241569519,
      "learning_rate": 1.8719159845599087e-06,
      "loss": 1.1703,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1986
    },
    {
      "epoch": 1.2465495608531996,
      "grad_norm": 1.7974519729614258,
      "learning_rate": 1.8692675541271232e-06,
      "loss": 1.3099,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1987
    },
    {
      "epoch": 1.247176913425345,
      "grad_norm": 1.7612372636795044,
      "learning_rate": 1.866619879449427e-06,
      "loss": 1.2829,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1988
    },
    {
      "epoch": 1.2478042659974906,
      "grad_norm": 1.8083420991897583,
      "learning_rate": 1.863972963699314e-06,
      "loss": 1.1797,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1989
    },
    {
      "epoch": 1.248431618569636,
      "grad_norm": 1.6637266874313354,
      "learning_rate": 1.8613268100483655e-06,
      "loss": 1.2991,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1990
    },
    {
      "epoch": 1.2490589711417817,
      "grad_norm": 3.0416088104248047,
      "learning_rate": 1.8586814216672506e-06,
      "loss": 1.281,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1991
    },
    {
      "epoch": 1.2496863237139273,
      "grad_norm": 1.5195271968841553,
      "learning_rate": 1.8560368017257229e-06,
      "loss": 1.2255,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1992
    },
    {
      "epoch": 1.2503136762860727,
      "grad_norm": 1.4624004364013672,
      "learning_rate": 1.8533929533926132e-06,
      "loss": 1.1726,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1993
    },
    {
      "epoch": 1.2509410288582183,
      "grad_norm": 1.525188684463501,
      "learning_rate": 1.8507498798358298e-06,
      "loss": 1.2229,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1994
    },
    {
      "epoch": 1.251568381430364,
      "grad_norm": 1.8328171968460083,
      "learning_rate": 1.8481075842223495e-06,
      "loss": 1.1667,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1995
    },
    {
      "epoch": 1.2521957340025094,
      "grad_norm": 1.9061431884765625,
      "learning_rate": 1.845466069718221e-06,
      "loss": 1.1872,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1996
    },
    {
      "epoch": 1.252823086574655,
      "grad_norm": 1.416793704032898,
      "learning_rate": 1.8428253394885547e-06,
      "loss": 1.0642,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1997
    },
    {
      "epoch": 1.2534504391468004,
      "grad_norm": 1.6995455026626587,
      "learning_rate": 1.8401853966975197e-06,
      "loss": 1.3359,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1998
    },
    {
      "epoch": 1.254077791718946,
      "grad_norm": 1.8740665912628174,
      "learning_rate": 1.8375462445083464e-06,
      "loss": 1.1768,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 1999
    },
    {
      "epoch": 1.2547051442910915,
      "grad_norm": 1.7629276514053345,
      "learning_rate": 1.8349078860833125e-06,
      "loss": 1.3704,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2000
    },
    {
      "epoch": 1.2553324968632371,
      "grad_norm": 1.8292843103408813,
      "learning_rate": 1.8322703245837486e-06,
      "loss": 1.1974,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2001
    },
    {
      "epoch": 1.2559598494353827,
      "grad_norm": 1.968329668045044,
      "learning_rate": 1.8296335631700276e-06,
      "loss": 1.2561,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2002
    },
    {
      "epoch": 1.2565872020075282,
      "grad_norm": 1.3263936042785645,
      "learning_rate": 1.8269976050015665e-06,
      "loss": 1.1638,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2003
    },
    {
      "epoch": 1.2572145545796738,
      "grad_norm": 1.8674527406692505,
      "learning_rate": 1.8243624532368176e-06,
      "loss": 1.353,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2004
    },
    {
      "epoch": 1.2578419071518194,
      "grad_norm": 1.6902563571929932,
      "learning_rate": 1.8217281110332667e-06,
      "loss": 1.1342,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2005
    },
    {
      "epoch": 1.2584692597239648,
      "grad_norm": 1.9965534210205078,
      "learning_rate": 1.8190945815474323e-06,
      "loss": 1.3001,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2006
    },
    {
      "epoch": 1.2590966122961105,
      "grad_norm": 1.8367878198623657,
      "learning_rate": 1.8164618679348556e-06,
      "loss": 1.1216,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2007
    },
    {
      "epoch": 1.2597239648682559,
      "grad_norm": 2.2213966846466064,
      "learning_rate": 1.8138299733501036e-06,
      "loss": 1.3812,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2008
    },
    {
      "epoch": 1.2603513174404015,
      "grad_norm": 1.5876895189285278,
      "learning_rate": 1.811198900946759e-06,
      "loss": 1.2787,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2009
    },
    {
      "epoch": 1.260978670012547,
      "grad_norm": 1.8638761043548584,
      "learning_rate": 1.8085686538774216e-06,
      "loss": 1.3114,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2010
    },
    {
      "epoch": 1.2616060225846926,
      "grad_norm": 1.8115777969360352,
      "learning_rate": 1.8059392352937011e-06,
      "loss": 1.1951,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2011
    },
    {
      "epoch": 1.2622333751568382,
      "grad_norm": 1.8829171657562256,
      "learning_rate": 1.8033106483462131e-06,
      "loss": 1.4238,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2012
    },
    {
      "epoch": 1.2628607277289836,
      "grad_norm": 2.542158365249634,
      "learning_rate": 1.8006828961845807e-06,
      "loss": 1.2422,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2013
    },
    {
      "epoch": 1.2634880803011292,
      "grad_norm": 1.9012902975082397,
      "learning_rate": 1.7980559819574222e-06,
      "loss": 1.3113,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2014
    },
    {
      "epoch": 1.2641154328732749,
      "grad_norm": 1.7465624809265137,
      "learning_rate": 1.7954299088123561e-06,
      "loss": 1.1354,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2015
    },
    {
      "epoch": 1.2647427854454203,
      "grad_norm": 1.758912444114685,
      "learning_rate": 1.792804679895989e-06,
      "loss": 1.236,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2016
    },
    {
      "epoch": 1.265370138017566,
      "grad_norm": 1.8036067485809326,
      "learning_rate": 1.7901802983539201e-06,
      "loss": 1.1432,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2017
    },
    {
      "epoch": 1.2659974905897113,
      "grad_norm": 1.4807945489883423,
      "learning_rate": 1.787556767330729e-06,
      "loss": 1.3149,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2018
    },
    {
      "epoch": 1.266624843161857,
      "grad_norm": 1.527632713317871,
      "learning_rate": 1.784934089969979e-06,
      "loss": 1.1745,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2019
    },
    {
      "epoch": 1.2672521957340024,
      "grad_norm": 1.6842902898788452,
      "learning_rate": 1.782312269414211e-06,
      "loss": 1.3125,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2020
    },
    {
      "epoch": 1.267879548306148,
      "grad_norm": 1.6198992729187012,
      "learning_rate": 1.7796913088049355e-06,
      "loss": 1.1504,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2021
    },
    {
      "epoch": 1.2685069008782937,
      "grad_norm": 1.6432766914367676,
      "learning_rate": 1.7770712112826383e-06,
      "loss": 1.3909,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2022
    },
    {
      "epoch": 1.269134253450439,
      "grad_norm": 1.643363356590271,
      "learning_rate": 1.7744519799867649e-06,
      "loss": 1.278,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2023
    },
    {
      "epoch": 1.2697616060225847,
      "grad_norm": 1.7085602283477783,
      "learning_rate": 1.771833618055729e-06,
      "loss": 1.1347,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2024
    },
    {
      "epoch": 1.2703889585947303,
      "grad_norm": 1.7079628705978394,
      "learning_rate": 1.7692161286268971e-06,
      "loss": 1.3693,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2025
    },
    {
      "epoch": 1.2710163111668757,
      "grad_norm": 1.3713947534561157,
      "learning_rate": 1.7665995148365934e-06,
      "loss": 1.1312,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2026
    },
    {
      "epoch": 1.2716436637390214,
      "grad_norm": 1.4894335269927979,
      "learning_rate": 1.7639837798200923e-06,
      "loss": 1.1924,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2027
    },
    {
      "epoch": 1.272271016311167,
      "grad_norm": 1.779098629951477,
      "learning_rate": 1.7613689267116141e-06,
      "loss": 1.2831,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2028
    },
    {
      "epoch": 1.2728983688833124,
      "grad_norm": 1.7018905878067017,
      "learning_rate": 1.7587549586443253e-06,
      "loss": 1.0808,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2029
    },
    {
      "epoch": 1.2735257214554578,
      "grad_norm": 2.0568008422851562,
      "learning_rate": 1.7561418787503275e-06,
      "loss": 1.3071,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2030
    },
    {
      "epoch": 1.2741530740276035,
      "grad_norm": 1.819620966911316,
      "learning_rate": 1.7535296901606635e-06,
      "loss": 1.3893,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2031
    },
    {
      "epoch": 1.274780426599749,
      "grad_norm": 1.7510826587677002,
      "learning_rate": 1.7509183960053033e-06,
      "loss": 1.2589,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2032
    },
    {
      "epoch": 1.2754077791718945,
      "grad_norm": 1.9268629550933838,
      "learning_rate": 1.7483079994131468e-06,
      "loss": 1.4034,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2033
    },
    {
      "epoch": 1.2760351317440402,
      "grad_norm": 1.7096502780914307,
      "learning_rate": 1.7456985035120194e-06,
      "loss": 1.2468,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2034
    },
    {
      "epoch": 1.2766624843161858,
      "grad_norm": 2.8108818531036377,
      "learning_rate": 1.7430899114286665e-06,
      "loss": 1.4387,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2035
    },
    {
      "epoch": 1.2772898368883312,
      "grad_norm": 1.732362985610962,
      "learning_rate": 1.740482226288751e-06,
      "loss": 1.0624,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2036
    },
    {
      "epoch": 1.2779171894604768,
      "grad_norm": 1.6320720911026,
      "learning_rate": 1.737875451216848e-06,
      "loss": 1.3641,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2037
    },
    {
      "epoch": 1.2785445420326225,
      "grad_norm": 1.8338778018951416,
      "learning_rate": 1.735269589336442e-06,
      "loss": 1.2858,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2038
    },
    {
      "epoch": 1.2791718946047679,
      "grad_norm": 1.6709052324295044,
      "learning_rate": 1.7326646437699262e-06,
      "loss": 1.3238,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2039
    },
    {
      "epoch": 1.2797992471769133,
      "grad_norm": 1.803065299987793,
      "learning_rate": 1.7300606176385914e-06,
      "loss": 1.3778,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2040
    },
    {
      "epoch": 1.280426599749059,
      "grad_norm": 1.883104681968689,
      "learning_rate": 1.7274575140626318e-06,
      "loss": 1.2621,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2041
    },
    {
      "epoch": 1.2810539523212046,
      "grad_norm": 1.8484208583831787,
      "learning_rate": 1.7248553361611313e-06,
      "loss": 1.2509,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2042
    },
    {
      "epoch": 1.28168130489335,
      "grad_norm": 1.7838290929794312,
      "learning_rate": 1.722254087052068e-06,
      "loss": 1.1035,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2043
    },
    {
      "epoch": 1.2823086574654956,
      "grad_norm": 1.768359899520874,
      "learning_rate": 1.7196537698523052e-06,
      "loss": 1.2978,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2044
    },
    {
      "epoch": 1.2829360100376412,
      "grad_norm": 1.7369588613510132,
      "learning_rate": 1.7170543876775898e-06,
      "loss": 1.4121,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2045
    },
    {
      "epoch": 1.2835633626097867,
      "grad_norm": 1.9081089496612549,
      "learning_rate": 1.71445594364255e-06,
      "loss": 1.0285,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2046
    },
    {
      "epoch": 1.2841907151819323,
      "grad_norm": 1.3868578672409058,
      "learning_rate": 1.7118584408606876e-06,
      "loss": 1.177,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2047
    },
    {
      "epoch": 1.284818067754078,
      "grad_norm": 1.5141382217407227,
      "learning_rate": 1.709261882444379e-06,
      "loss": 1.1907,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2048
    },
    {
      "epoch": 1.2854454203262233,
      "grad_norm": 1.4479154348373413,
      "learning_rate": 1.7066662715048668e-06,
      "loss": 1.1626,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2049
    },
    {
      "epoch": 1.286072772898369,
      "grad_norm": 1.4316198825836182,
      "learning_rate": 1.7040716111522592e-06,
      "loss": 1.1732,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2050
    },
    {
      "epoch": 1.2867001254705144,
      "grad_norm": 1.5437365770339966,
      "learning_rate": 1.7014779044955262e-06,
      "loss": 1.207,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2051
    },
    {
      "epoch": 1.28732747804266,
      "grad_norm": 1.5371419191360474,
      "learning_rate": 1.6988851546424934e-06,
      "loss": 1.1898,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2052
    },
    {
      "epoch": 1.2879548306148054,
      "grad_norm": 1.7966153621673584,
      "learning_rate": 1.6962933646998413e-06,
      "loss": 1.3189,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2053
    },
    {
      "epoch": 1.288582183186951,
      "grad_norm": 2.556145191192627,
      "learning_rate": 1.6937025377730992e-06,
      "loss": 1.1015,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2054
    },
    {
      "epoch": 1.2892095357590967,
      "grad_norm": 1.7997864484786987,
      "learning_rate": 1.6911126769666442e-06,
      "loss": 1.2621,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2055
    },
    {
      "epoch": 1.289836888331242,
      "grad_norm": 2.656108856201172,
      "learning_rate": 1.6885237853836932e-06,
      "loss": 1.3861,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2056
    },
    {
      "epoch": 1.2904642409033877,
      "grad_norm": 1.6520987749099731,
      "learning_rate": 1.6859358661263048e-06,
      "loss": 1.2589,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2057
    },
    {
      "epoch": 1.2910915934755334,
      "grad_norm": 2.0792765617370605,
      "learning_rate": 1.68334892229537e-06,
      "loss": 1.1939,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2058
    },
    {
      "epoch": 1.2917189460476788,
      "grad_norm": 1.6274360418319702,
      "learning_rate": 1.6807629569906113e-06,
      "loss": 1.2164,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2059
    },
    {
      "epoch": 1.2923462986198244,
      "grad_norm": 2.0519096851348877,
      "learning_rate": 1.6781779733105813e-06,
      "loss": 1.2378,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2060
    },
    {
      "epoch": 1.2929736511919698,
      "grad_norm": 1.7964249849319458,
      "learning_rate": 1.6755939743526516e-06,
      "loss": 1.1686,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2061
    },
    {
      "epoch": 1.2936010037641155,
      "grad_norm": 1.4359616041183472,
      "learning_rate": 1.6730109632130199e-06,
      "loss": 1.2983,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2062
    },
    {
      "epoch": 1.2942283563362609,
      "grad_norm": 2.8610012531280518,
      "learning_rate": 1.6704289429866942e-06,
      "loss": 1.2046,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2063
    },
    {
      "epoch": 1.2948557089084065,
      "grad_norm": 1.8337393999099731,
      "learning_rate": 1.6678479167675005e-06,
      "loss": 1.1787,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2064
    },
    {
      "epoch": 1.2954830614805521,
      "grad_norm": 1.6933302879333496,
      "learning_rate": 1.665267887648071e-06,
      "loss": 1.2077,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2065
    },
    {
      "epoch": 1.2961104140526976,
      "grad_norm": 1.5457978248596191,
      "learning_rate": 1.6626888587198414e-06,
      "loss": 1.3236,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2066
    },
    {
      "epoch": 1.2967377666248432,
      "grad_norm": 1.985788106918335,
      "learning_rate": 1.6601108330730536e-06,
      "loss": 1.2518,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2067
    },
    {
      "epoch": 1.2973651191969888,
      "grad_norm": 2.003683090209961,
      "learning_rate": 1.657533813796744e-06,
      "loss": 1.1951,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2068
    },
    {
      "epoch": 1.2979924717691342,
      "grad_norm": 1.9997378587722778,
      "learning_rate": 1.6549578039787436e-06,
      "loss": 1.2089,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2069
    },
    {
      "epoch": 1.2986198243412799,
      "grad_norm": 1.4930788278579712,
      "learning_rate": 1.6523828067056739e-06,
      "loss": 1.1872,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2070
    },
    {
      "epoch": 1.2992471769134253,
      "grad_norm": 1.7217084169387817,
      "learning_rate": 1.6498088250629445e-06,
      "loss": 1.2626,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2071
    },
    {
      "epoch": 1.299874529485571,
      "grad_norm": 1.5882774591445923,
      "learning_rate": 1.6472358621347472e-06,
      "loss": 1.2439,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2072
    },
    {
      "epoch": 1.3005018820577163,
      "grad_norm": 1.8129900693893433,
      "learning_rate": 1.644663921004051e-06,
      "loss": 1.2066,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2073
    },
    {
      "epoch": 1.301129234629862,
      "grad_norm": 1.7803122997283936,
      "learning_rate": 1.6420930047526048e-06,
      "loss": 1.1882,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2074
    },
    {
      "epoch": 1.3017565872020076,
      "grad_norm": 1.6241403818130493,
      "learning_rate": 1.639523116460926e-06,
      "loss": 1.1229,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2075
    },
    {
      "epoch": 1.302383939774153,
      "grad_norm": 1.4273207187652588,
      "learning_rate": 1.636954259208302e-06,
      "loss": 1.1985,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2076
    },
    {
      "epoch": 1.3030112923462986,
      "grad_norm": 1.567501187324524,
      "learning_rate": 1.6343864360727835e-06,
      "loss": 1.1494,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2077
    },
    {
      "epoch": 1.3036386449184443,
      "grad_norm": 1.8070402145385742,
      "learning_rate": 1.6318196501311841e-06,
      "loss": 1.2294,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2078
    },
    {
      "epoch": 1.3042659974905897,
      "grad_norm": 1.795170545578003,
      "learning_rate": 1.629253904459073e-06,
      "loss": 1.2719,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2079
    },
    {
      "epoch": 1.3048933500627353,
      "grad_norm": 1.7698767185211182,
      "learning_rate": 1.6266892021307724e-06,
      "loss": 1.1697,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2080
    },
    {
      "epoch": 1.3055207026348807,
      "grad_norm": 2.0055277347564697,
      "learning_rate": 1.6241255462193565e-06,
      "loss": 1.1996,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2081
    },
    {
      "epoch": 1.3061480552070264,
      "grad_norm": 1.8001643419265747,
      "learning_rate": 1.6215629397966432e-06,
      "loss": 1.231,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2082
    },
    {
      "epoch": 1.3067754077791718,
      "grad_norm": 1.4703590869903564,
      "learning_rate": 1.6190013859331958e-06,
      "loss": 1.2711,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2083
    },
    {
      "epoch": 1.3074027603513174,
      "grad_norm": 1.8987573385238647,
      "learning_rate": 1.616440887698313e-06,
      "loss": 1.2926,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2084
    },
    {
      "epoch": 1.308030112923463,
      "grad_norm": 1.6749411821365356,
      "learning_rate": 1.6138814481600324e-06,
      "loss": 1.1826,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2085
    },
    {
      "epoch": 1.3086574654956085,
      "grad_norm": 1.7081942558288574,
      "learning_rate": 1.6113230703851194e-06,
      "loss": 1.2047,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2086
    },
    {
      "epoch": 1.309284818067754,
      "grad_norm": 1.7994329929351807,
      "learning_rate": 1.6087657574390681e-06,
      "loss": 1.209,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2087
    },
    {
      "epoch": 1.3099121706398997,
      "grad_norm": 1.7203141450881958,
      "learning_rate": 1.6062095123860996e-06,
      "loss": 1.2379,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2088
    },
    {
      "epoch": 1.3105395232120451,
      "grad_norm": 1.7341119050979614,
      "learning_rate": 1.6036543382891512e-06,
      "loss": 1.2684,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2089
    },
    {
      "epoch": 1.3111668757841908,
      "grad_norm": 1.6238679885864258,
      "learning_rate": 1.6011002382098806e-06,
      "loss": 1.1199,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2090
    },
    {
      "epoch": 1.3117942283563362,
      "grad_norm": 1.4414830207824707,
      "learning_rate": 1.5985472152086565e-06,
      "loss": 1.1536,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2091
    },
    {
      "epoch": 1.3124215809284818,
      "grad_norm": 1.963968276977539,
      "learning_rate": 1.5959952723445565e-06,
      "loss": 1.2495,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2092
    },
    {
      "epoch": 1.3130489335006272,
      "grad_norm": 2.341641902923584,
      "learning_rate": 1.5934444126753663e-06,
      "loss": 1.2921,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2093
    },
    {
      "epoch": 1.3136762860727729,
      "grad_norm": 1.4138715267181396,
      "learning_rate": 1.5908946392575713e-06,
      "loss": 1.1649,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2094
    },
    {
      "epoch": 1.3143036386449185,
      "grad_norm": 1.7310720682144165,
      "learning_rate": 1.5883459551463576e-06,
      "loss": 1.2476,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2095
    },
    {
      "epoch": 1.314930991217064,
      "grad_norm": 1.502193808555603,
      "learning_rate": 1.5857983633956027e-06,
      "loss": 1.1271,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2096
    },
    {
      "epoch": 1.3155583437892095,
      "grad_norm": 1.9654653072357178,
      "learning_rate": 1.5832518670578802e-06,
      "loss": 1.1828,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2097
    },
    {
      "epoch": 1.3161856963613552,
      "grad_norm": 2.250328779220581,
      "learning_rate": 1.580706469184447e-06,
      "loss": 1.1769,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2098
    },
    {
      "epoch": 1.3168130489335006,
      "grad_norm": 1.8096691370010376,
      "learning_rate": 1.5781621728252439e-06,
      "loss": 1.1981,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2099
    },
    {
      "epoch": 1.3174404015056462,
      "grad_norm": 1.3865689039230347,
      "learning_rate": 1.5756189810288952e-06,
      "loss": 1.1288,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2100
    },
    {
      "epoch": 1.3180677540777916,
      "grad_norm": 1.7595837116241455,
      "learning_rate": 1.5730768968426985e-06,
      "loss": 1.3046,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2101
    },
    {
      "epoch": 1.3186951066499373,
      "grad_norm": 2.089447498321533,
      "learning_rate": 1.5705359233126255e-06,
      "loss": 1.3244,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2102
    },
    {
      "epoch": 1.3193224592220827,
      "grad_norm": 1.8050563335418701,
      "learning_rate": 1.5679960634833164e-06,
      "loss": 1.2567,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2103
    },
    {
      "epoch": 1.3199498117942283,
      "grad_norm": 1.724447250366211,
      "learning_rate": 1.5654573203980782e-06,
      "loss": 1.1666,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2104
    },
    {
      "epoch": 1.320577164366374,
      "grad_norm": 1.9797168970108032,
      "learning_rate": 1.5629196970988791e-06,
      "loss": 1.1802,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2105
    },
    {
      "epoch": 1.3212045169385194,
      "grad_norm": 1.887502908706665,
      "learning_rate": 1.5603831966263444e-06,
      "loss": 1.3445,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2106
    },
    {
      "epoch": 1.321831869510665,
      "grad_norm": 1.614529013633728,
      "learning_rate": 1.5578478220197568e-06,
      "loss": 1.3333,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2107
    },
    {
      "epoch": 1.3224592220828106,
      "grad_norm": 1.4517871141433716,
      "learning_rate": 1.5553135763170466e-06,
      "loss": 1.3709,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2108
    },
    {
      "epoch": 1.323086574654956,
      "grad_norm": 1.386285662651062,
      "learning_rate": 1.5527804625547937e-06,
      "loss": 1.31,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2109
    },
    {
      "epoch": 1.3237139272271017,
      "grad_norm": 1.4335236549377441,
      "learning_rate": 1.5502484837682216e-06,
      "loss": 1.1913,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2110
    },
    {
      "epoch": 1.324341279799247,
      "grad_norm": 1.60616934299469,
      "learning_rate": 1.5477176429911934e-06,
      "loss": 1.3257,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2111
    },
    {
      "epoch": 1.3249686323713927,
      "grad_norm": 1.5817228555679321,
      "learning_rate": 1.5451879432562078e-06,
      "loss": 1.1697,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2112
    },
    {
      "epoch": 1.3255959849435381,
      "grad_norm": 1.670749306678772,
      "learning_rate": 1.5426593875943967e-06,
      "loss": 1.1722,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2113
    },
    {
      "epoch": 1.3262233375156838,
      "grad_norm": 2.0000078678131104,
      "learning_rate": 1.5401319790355232e-06,
      "loss": 1.2758,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2114
    },
    {
      "epoch": 1.3268506900878294,
      "grad_norm": 1.9140396118164062,
      "learning_rate": 1.537605720607972e-06,
      "loss": 1.252,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2115
    },
    {
      "epoch": 1.3274780426599748,
      "grad_norm": 1.599395990371704,
      "learning_rate": 1.5350806153387541e-06,
      "loss": 1.2188,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2116
    },
    {
      "epoch": 1.3281053952321205,
      "grad_norm": 1.6747828722000122,
      "learning_rate": 1.5325566662534948e-06,
      "loss": 1.3726,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2117
    },
    {
      "epoch": 1.328732747804266,
      "grad_norm": 1.457871675491333,
      "learning_rate": 1.5300338763764371e-06,
      "loss": 1.2079,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2118
    },
    {
      "epoch": 1.3293601003764115,
      "grad_norm": 1.839821219444275,
      "learning_rate": 1.5275122487304337e-06,
      "loss": 1.0779,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2119
    },
    {
      "epoch": 1.3299874529485571,
      "grad_norm": 1.8094121217727661,
      "learning_rate": 1.5249917863369426e-06,
      "loss": 1.3537,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2120
    },
    {
      "epoch": 1.3306148055207028,
      "grad_norm": 1.6778947114944458,
      "learning_rate": 1.5224724922160304e-06,
      "loss": 1.272,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2121
    },
    {
      "epoch": 1.3312421580928482,
      "grad_norm": 1.0478399991989136,
      "learning_rate": 1.519954369386359e-06,
      "loss": 1.094,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2122
    },
    {
      "epoch": 1.3318695106649936,
      "grad_norm": 1.8125706911087036,
      "learning_rate": 1.5174374208651913e-06,
      "loss": 1.3052,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2123
    },
    {
      "epoch": 1.3324968632371392,
      "grad_norm": 1.7263497114181519,
      "learning_rate": 1.5149216496683788e-06,
      "loss": 1.1937,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2124
    },
    {
      "epoch": 1.3331242158092849,
      "grad_norm": 1.8608651161193848,
      "learning_rate": 1.5124070588103648e-06,
      "loss": 1.2324,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2125
    },
    {
      "epoch": 1.3337515683814303,
      "grad_norm": 1.5468758344650269,
      "learning_rate": 1.5098936513041787e-06,
      "loss": 1.167,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2126
    },
    {
      "epoch": 1.334378920953576,
      "grad_norm": 1.8814101219177246,
      "learning_rate": 1.5073814301614298e-06,
      "loss": 1.1359,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2127
    },
    {
      "epoch": 1.3350062735257215,
      "grad_norm": 1.8458728790283203,
      "learning_rate": 1.5048703983923086e-06,
      "loss": 1.2031,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2128
    },
    {
      "epoch": 1.335633626097867,
      "grad_norm": 1.2960106134414673,
      "learning_rate": 1.5023605590055768e-06,
      "loss": 1.134,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2129
    },
    {
      "epoch": 1.3362609786700126,
      "grad_norm": 2.0509815216064453,
      "learning_rate": 1.4998519150085722e-06,
      "loss": 1.2158,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2130
    },
    {
      "epoch": 1.3368883312421582,
      "grad_norm": 1.7239158153533936,
      "learning_rate": 1.4973444694071957e-06,
      "loss": 1.199,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2131
    },
    {
      "epoch": 1.3375156838143036,
      "grad_norm": 1.9156348705291748,
      "learning_rate": 1.4948382252059158e-06,
      "loss": 1.2761,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2132
    },
    {
      "epoch": 1.338143036386449,
      "grad_norm": 1.6581004858016968,
      "learning_rate": 1.4923331854077592e-06,
      "loss": 1.3608,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2133
    },
    {
      "epoch": 1.3387703889585947,
      "grad_norm": 1.7229657173156738,
      "learning_rate": 1.4898293530143095e-06,
      "loss": 1.1478,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2134
    },
    {
      "epoch": 1.3393977415307403,
      "grad_norm": 1.7327300310134888,
      "learning_rate": 1.4873267310257056e-06,
      "loss": 1.1457,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2135
    },
    {
      "epoch": 1.3400250941028857,
      "grad_norm": 2.0124669075012207,
      "learning_rate": 1.4848253224406326e-06,
      "loss": 1.1386,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2136
    },
    {
      "epoch": 1.3406524466750314,
      "grad_norm": 1.6715619564056396,
      "learning_rate": 1.482325130256326e-06,
      "loss": 1.2899,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2137
    },
    {
      "epoch": 1.341279799247177,
      "grad_norm": 2.0009779930114746,
      "learning_rate": 1.4798261574685597e-06,
      "loss": 1.1755,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2138
    },
    {
      "epoch": 1.3419071518193224,
      "grad_norm": 1.9581701755523682,
      "learning_rate": 1.4773284070716504e-06,
      "loss": 1.2387,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2139
    },
    {
      "epoch": 1.342534504391468,
      "grad_norm": 1.7659634351730347,
      "learning_rate": 1.4748318820584468e-06,
      "loss": 1.228,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2140
    },
    {
      "epoch": 1.3431618569636137,
      "grad_norm": 2.4370534420013428,
      "learning_rate": 1.4723365854203298e-06,
      "loss": 1.2194,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2141
    },
    {
      "epoch": 1.343789209535759,
      "grad_norm": 2.0025267601013184,
      "learning_rate": 1.4698425201472111e-06,
      "loss": 1.1692,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2142
    },
    {
      "epoch": 1.3444165621079047,
      "grad_norm": 1.8108923435211182,
      "learning_rate": 1.4673496892275237e-06,
      "loss": 1.3146,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2143
    },
    {
      "epoch": 1.3450439146800501,
      "grad_norm": 1.5675235986709595,
      "learning_rate": 1.4648580956482238e-06,
      "loss": 1.0542,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2144
    },
    {
      "epoch": 1.3456712672521958,
      "grad_norm": 1.8576399087905884,
      "learning_rate": 1.4623677423947841e-06,
      "loss": 1.2311,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2145
    },
    {
      "epoch": 1.3462986198243412,
      "grad_norm": 1.8584723472595215,
      "learning_rate": 1.4598786324511892e-06,
      "loss": 1.0874,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2146
    },
    {
      "epoch": 1.3469259723964868,
      "grad_norm": 2.055356979370117,
      "learning_rate": 1.4573907687999383e-06,
      "loss": 1.2173,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2147
    },
    {
      "epoch": 1.3475533249686324,
      "grad_norm": 1.7574433088302612,
      "learning_rate": 1.4549041544220347e-06,
      "loss": 1.3364,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2148
    },
    {
      "epoch": 1.3481806775407779,
      "grad_norm": 1.7991924285888672,
      "learning_rate": 1.452418792296984e-06,
      "loss": 1.1384,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2149
    },
    {
      "epoch": 1.3488080301129235,
      "grad_norm": 1.873650074005127,
      "learning_rate": 1.4499346854027921e-06,
      "loss": 1.2564,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2150
    },
    {
      "epoch": 1.3494353826850691,
      "grad_norm": 1.9361118078231812,
      "learning_rate": 1.4474518367159623e-06,
      "loss": 1.3539,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2151
    },
    {
      "epoch": 1.3500627352572145,
      "grad_norm": 1.726911187171936,
      "learning_rate": 1.4449702492114886e-06,
      "loss": 1.1937,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2152
    },
    {
      "epoch": 1.3506900878293602,
      "grad_norm": 1.8655195236206055,
      "learning_rate": 1.4424899258628533e-06,
      "loss": 1.2017,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2153
    },
    {
      "epoch": 1.3513174404015056,
      "grad_norm": 1.840174674987793,
      "learning_rate": 1.4400108696420265e-06,
      "loss": 1.2439,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2154
    },
    {
      "epoch": 1.3519447929736512,
      "grad_norm": 1.859181523323059,
      "learning_rate": 1.4375330835194568e-06,
      "loss": 1.3436,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2155
    },
    {
      "epoch": 1.3525721455457966,
      "grad_norm": 1.7906042337417603,
      "learning_rate": 1.4350565704640746e-06,
      "loss": 1.1164,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2156
    },
    {
      "epoch": 1.3531994981179423,
      "grad_norm": 1.8030601739883423,
      "learning_rate": 1.4325813334432805e-06,
      "loss": 1.1413,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2157
    },
    {
      "epoch": 1.353826850690088,
      "grad_norm": 1.8038133382797241,
      "learning_rate": 1.4301073754229511e-06,
      "loss": 1.3058,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2158
    },
    {
      "epoch": 1.3544542032622333,
      "grad_norm": 6.459727764129639,
      "learning_rate": 1.4276346993674267e-06,
      "loss": 1.3058,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2159
    },
    {
      "epoch": 1.355081555834379,
      "grad_norm": 1.8177783489227295,
      "learning_rate": 1.4251633082395117e-06,
      "loss": 1.2954,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2160
    },
    {
      "epoch": 1.3557089084065246,
      "grad_norm": 1.9179670810699463,
      "learning_rate": 1.4226932050004735e-06,
      "loss": 1.2445,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2161
    },
    {
      "epoch": 1.35633626097867,
      "grad_norm": 1.9530935287475586,
      "learning_rate": 1.4202243926100345e-06,
      "loss": 1.2474,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2162
    },
    {
      "epoch": 1.3569636135508156,
      "grad_norm": 1.547910213470459,
      "learning_rate": 1.4177568740263692e-06,
      "loss": 1.3921,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2163
    },
    {
      "epoch": 1.357590966122961,
      "grad_norm": 1.682190179824829,
      "learning_rate": 1.415290652206105e-06,
      "loss": 1.1747,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2164
    },
    {
      "epoch": 1.3582183186951067,
      "grad_norm": 1.6370090246200562,
      "learning_rate": 1.412825730104313e-06,
      "loss": 1.148,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2165
    },
    {
      "epoch": 1.358845671267252,
      "grad_norm": 1.9257128238677979,
      "learning_rate": 1.410362110674508e-06,
      "loss": 1.2593,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2166
    },
    {
      "epoch": 1.3594730238393977,
      "grad_norm": 1.7159314155578613,
      "learning_rate": 1.4078997968686425e-06,
      "loss": 1.3904,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2167
    },
    {
      "epoch": 1.3601003764115434,
      "grad_norm": 1.8190382719039917,
      "learning_rate": 1.405438791637107e-06,
      "loss": 1.3654,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2168
    },
    {
      "epoch": 1.3607277289836888,
      "grad_norm": 1.6482884883880615,
      "learning_rate": 1.4029790979287217e-06,
      "loss": 1.1206,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2169
    },
    {
      "epoch": 1.3613550815558344,
      "grad_norm": 1.9584014415740967,
      "learning_rate": 1.400520718690738e-06,
      "loss": 1.1457,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2170
    },
    {
      "epoch": 1.36198243412798,
      "grad_norm": 1.7346889972686768,
      "learning_rate": 1.3980636568688283e-06,
      "loss": 1.2116,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2171
    },
    {
      "epoch": 1.3626097867001254,
      "grad_norm": 1.715471625328064,
      "learning_rate": 1.395607915407091e-06,
      "loss": 1.2171,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2172
    },
    {
      "epoch": 1.363237139272271,
      "grad_norm": 1.783653736114502,
      "learning_rate": 1.3931534972480392e-06,
      "loss": 1.2286,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2173
    },
    {
      "epoch": 1.3638644918444165,
      "grad_norm": 1.8066835403442383,
      "learning_rate": 1.3907004053326006e-06,
      "loss": 1.1425,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2174
    },
    {
      "epoch": 1.3644918444165621,
      "grad_norm": 1.775204062461853,
      "learning_rate": 1.3882486426001162e-06,
      "loss": 1.2009,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2175
    },
    {
      "epoch": 1.3651191969887075,
      "grad_norm": 1.8585458993911743,
      "learning_rate": 1.3857982119883313e-06,
      "loss": 1.2242,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2176
    },
    {
      "epoch": 1.3657465495608532,
      "grad_norm": 1.6320204734802246,
      "learning_rate": 1.3833491164333974e-06,
      "loss": 1.2934,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2177
    },
    {
      "epoch": 1.3663739021329988,
      "grad_norm": 1.6733769178390503,
      "learning_rate": 1.380901358869865e-06,
      "loss": 1.1086,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2178
    },
    {
      "epoch": 1.3670012547051442,
      "grad_norm": 1.6722503900527954,
      "learning_rate": 1.3784549422306808e-06,
      "loss": 1.0616,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2179
    },
    {
      "epoch": 1.3676286072772899,
      "grad_norm": 1.600703477859497,
      "learning_rate": 1.3760098694471876e-06,
      "loss": 1.1497,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2180
    },
    {
      "epoch": 1.3682559598494355,
      "grad_norm": 2.0433199405670166,
      "learning_rate": 1.373566143449115e-06,
      "loss": 1.3881,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2181
    },
    {
      "epoch": 1.368883312421581,
      "grad_norm": 1.7326481342315674,
      "learning_rate": 1.3711237671645802e-06,
      "loss": 1.2705,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2182
    },
    {
      "epoch": 1.3695106649937265,
      "grad_norm": 1.9093871116638184,
      "learning_rate": 1.3686827435200817e-06,
      "loss": 1.2108,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2183
    },
    {
      "epoch": 1.370138017565872,
      "grad_norm": 1.912520170211792,
      "learning_rate": 1.3662430754405004e-06,
      "loss": 1.1256,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2184
    },
    {
      "epoch": 1.3707653701380176,
      "grad_norm": 1.5855059623718262,
      "learning_rate": 1.363804765849089e-06,
      "loss": 1.2166,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2185
    },
    {
      "epoch": 1.371392722710163,
      "grad_norm": 1.4968267679214478,
      "learning_rate": 1.361367817667476e-06,
      "loss": 1.3215,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2186
    },
    {
      "epoch": 1.3720200752823086,
      "grad_norm": 1.8088436126708984,
      "learning_rate": 1.3589322338156571e-06,
      "loss": 1.2745,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2187
    },
    {
      "epoch": 1.3726474278544543,
      "grad_norm": 1.8022375106811523,
      "learning_rate": 1.3564980172119913e-06,
      "loss": 1.2414,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2188
    },
    {
      "epoch": 1.3732747804265997,
      "grad_norm": 1.7805216312408447,
      "learning_rate": 1.3540651707732036e-06,
      "loss": 1.2126,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2189
    },
    {
      "epoch": 1.3739021329987453,
      "grad_norm": 1.4937478303909302,
      "learning_rate": 1.3516336974143726e-06,
      "loss": 1.0723,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2190
    },
    {
      "epoch": 1.374529485570891,
      "grad_norm": 1.806618571281433,
      "learning_rate": 1.3492036000489361e-06,
      "loss": 1.2422,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2191
    },
    {
      "epoch": 1.3751568381430364,
      "grad_norm": 1.7576457262039185,
      "learning_rate": 1.34677488158868e-06,
      "loss": 1.2792,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2192
    },
    {
      "epoch": 1.375784190715182,
      "grad_norm": 2.2949795722961426,
      "learning_rate": 1.3443475449437377e-06,
      "loss": 1.14,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2193
    },
    {
      "epoch": 1.3764115432873274,
      "grad_norm": 1.6671180725097656,
      "learning_rate": 1.3419215930225898e-06,
      "loss": 1.2154,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2194
    },
    {
      "epoch": 1.377038895859473,
      "grad_norm": 1.8693598508834839,
      "learning_rate": 1.3394970287320553e-06,
      "loss": 1.1688,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2195
    },
    {
      "epoch": 1.3776662484316184,
      "grad_norm": 1.7913018465042114,
      "learning_rate": 1.3370738549772921e-06,
      "loss": 1.2461,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2196
    },
    {
      "epoch": 1.378293601003764,
      "grad_norm": 1.840532898902893,
      "learning_rate": 1.3346520746617903e-06,
      "loss": 1.2051,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2197
    },
    {
      "epoch": 1.3789209535759097,
      "grad_norm": 1.9473389387130737,
      "learning_rate": 1.3322316906873712e-06,
      "loss": 1.1106,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2198
    },
    {
      "epoch": 1.3795483061480551,
      "grad_norm": 1.4682437181472778,
      "learning_rate": 1.329812705954183e-06,
      "loss": 1.1503,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2199
    },
    {
      "epoch": 1.3801756587202008,
      "grad_norm": 1.787130355834961,
      "learning_rate": 1.327395123360696e-06,
      "loss": 1.2355,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2200
    },
    {
      "epoch": 1.3808030112923464,
      "grad_norm": 1.6980496644973755,
      "learning_rate": 1.324978945803704e-06,
      "loss": 1.1478,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2201
    },
    {
      "epoch": 1.3814303638644918,
      "grad_norm": 1.329012393951416,
      "learning_rate": 1.3225641761783126e-06,
      "loss": 1.2075,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2202
    },
    {
      "epoch": 1.3820577164366374,
      "grad_norm": 2.6499478816986084,
      "learning_rate": 1.320150817377945e-06,
      "loss": 1.2777,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2203
    },
    {
      "epoch": 1.382685069008783,
      "grad_norm": 2.159608840942383,
      "learning_rate": 1.317738872294329e-06,
      "loss": 1.1045,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2204
    },
    {
      "epoch": 1.3833124215809285,
      "grad_norm": 1.7414979934692383,
      "learning_rate": 1.3153283438175036e-06,
      "loss": 1.3571,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2205
    },
    {
      "epoch": 1.383939774153074,
      "grad_norm": 1.630829095840454,
      "learning_rate": 1.3129192348358066e-06,
      "loss": 1.1684,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2206
    },
    {
      "epoch": 1.3845671267252195,
      "grad_norm": 1.8356767892837524,
      "learning_rate": 1.3105115482358758e-06,
      "loss": 1.1264,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2207
    },
    {
      "epoch": 1.3851944792973652,
      "grad_norm": 1.557112455368042,
      "learning_rate": 1.308105286902646e-06,
      "loss": 1.0991,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2208
    },
    {
      "epoch": 1.3858218318695106,
      "grad_norm": 1.9413764476776123,
      "learning_rate": 1.3057004537193424e-06,
      "loss": 1.252,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2209
    },
    {
      "epoch": 1.3864491844416562,
      "grad_norm": 1.818488359451294,
      "learning_rate": 1.3032970515674814e-06,
      "loss": 1.3677,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2210
    },
    {
      "epoch": 1.3870765370138018,
      "grad_norm": 1.8553251028060913,
      "learning_rate": 1.3008950833268609e-06,
      "loss": 1.136,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2211
    },
    {
      "epoch": 1.3877038895859473,
      "grad_norm": 1.7978549003601074,
      "learning_rate": 1.2984945518755654e-06,
      "loss": 1.1121,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2212
    },
    {
      "epoch": 1.388331242158093,
      "grad_norm": 1.7926452159881592,
      "learning_rate": 1.2960954600899539e-06,
      "loss": 1.122,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2213
    },
    {
      "epoch": 1.3889585947302385,
      "grad_norm": 1.7817319631576538,
      "learning_rate": 1.2936978108446624e-06,
      "loss": 1.1541,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2214
    },
    {
      "epoch": 1.389585947302384,
      "grad_norm": 1.463486671447754,
      "learning_rate": 1.2913016070125971e-06,
      "loss": 1.0774,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2215
    },
    {
      "epoch": 1.3902132998745294,
      "grad_norm": 2.401296377182007,
      "learning_rate": 1.2889068514649328e-06,
      "loss": 1.2591,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2216
    },
    {
      "epoch": 1.390840652446675,
      "grad_norm": 1.7470189332962036,
      "learning_rate": 1.2865135470711107e-06,
      "loss": 1.2515,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2217
    },
    {
      "epoch": 1.3914680050188206,
      "grad_norm": 1.9201079607009888,
      "learning_rate": 1.2841216966988295e-06,
      "loss": 1.2297,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2218
    },
    {
      "epoch": 1.392095357590966,
      "grad_norm": 1.6405366659164429,
      "learning_rate": 1.2817313032140504e-06,
      "loss": 1.2744,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2219
    },
    {
      "epoch": 1.3927227101631117,
      "grad_norm": 2.254488468170166,
      "learning_rate": 1.2793423694809854e-06,
      "loss": 1.1555,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2220
    },
    {
      "epoch": 1.3933500627352573,
      "grad_norm": 1.7753210067749023,
      "learning_rate": 1.2769548983620963e-06,
      "loss": 1.1457,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2221
    },
    {
      "epoch": 1.3939774153074027,
      "grad_norm": 1.7737972736358643,
      "learning_rate": 1.2745688927180974e-06,
      "loss": 1.3371,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2222
    },
    {
      "epoch": 1.3946047678795483,
      "grad_norm": 1.9211421012878418,
      "learning_rate": 1.2721843554079418e-06,
      "loss": 1.2411,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2223
    },
    {
      "epoch": 1.395232120451694,
      "grad_norm": 2.360780954360962,
      "learning_rate": 1.2698012892888272e-06,
      "loss": 1.2071,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2224
    },
    {
      "epoch": 1.3958594730238394,
      "grad_norm": 1.9631240367889404,
      "learning_rate": 1.2674196972161856e-06,
      "loss": 1.2879,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2225
    },
    {
      "epoch": 1.396486825595985,
      "grad_norm": 1.9129877090454102,
      "learning_rate": 1.265039582043685e-06,
      "loss": 1.2065,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2226
    },
    {
      "epoch": 1.3971141781681304,
      "grad_norm": 1.8536701202392578,
      "learning_rate": 1.262660946623222e-06,
      "loss": 1.134,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2227
    },
    {
      "epoch": 1.397741530740276,
      "grad_norm": 2.0723443031311035,
      "learning_rate": 1.2602837938049204e-06,
      "loss": 1.4661,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2228
    },
    {
      "epoch": 1.3983688833124215,
      "grad_norm": 1.4898669719696045,
      "learning_rate": 1.257908126437129e-06,
      "loss": 1.0878,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2229
    },
    {
      "epoch": 1.3989962358845671,
      "grad_norm": 1.8184411525726318,
      "learning_rate": 1.2555339473664151e-06,
      "loss": 1.1638,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2230
    },
    {
      "epoch": 1.3996235884567128,
      "grad_norm": 1.7882477045059204,
      "learning_rate": 1.2531612594375636e-06,
      "loss": 1.2015,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2231
    },
    {
      "epoch": 1.4002509410288582,
      "grad_norm": 1.7373894453048706,
      "learning_rate": 1.2507900654935706e-06,
      "loss": 1.2848,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2232
    },
    {
      "epoch": 1.4008782936010038,
      "grad_norm": 1.6618754863739014,
      "learning_rate": 1.2484203683756466e-06,
      "loss": 1.1881,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2233
    },
    {
      "epoch": 1.4015056461731494,
      "grad_norm": 1.8542975187301636,
      "learning_rate": 1.2460521709232042e-06,
      "loss": 1.2826,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2234
    },
    {
      "epoch": 1.4021329987452948,
      "grad_norm": 1.7895851135253906,
      "learning_rate": 1.24368547597386e-06,
      "loss": 1.2008,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2235
    },
    {
      "epoch": 1.4027603513174405,
      "grad_norm": 1.808781623840332,
      "learning_rate": 1.2413202863634328e-06,
      "loss": 1.2159,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2236
    },
    {
      "epoch": 1.403387703889586,
      "grad_norm": 1.8326939344406128,
      "learning_rate": 1.238956604925934e-06,
      "loss": 1.3503,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2237
    },
    {
      "epoch": 1.4040150564617315,
      "grad_norm": 1.768405795097351,
      "learning_rate": 1.236594434493571e-06,
      "loss": 1.2804,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2238
    },
    {
      "epoch": 1.404642409033877,
      "grad_norm": 1.4438402652740479,
      "learning_rate": 1.2342337778967383e-06,
      "loss": 1.068,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2239
    },
    {
      "epoch": 1.4052697616060226,
      "grad_norm": 1.7598780393600464,
      "learning_rate": 1.2318746379640192e-06,
      "loss": 1.2313,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2240
    },
    {
      "epoch": 1.4058971141781682,
      "grad_norm": 1.9519436359405518,
      "learning_rate": 1.229517017522177e-06,
      "loss": 1.1764,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2241
    },
    {
      "epoch": 1.4065244667503136,
      "grad_norm": 1.7161890268325806,
      "learning_rate": 1.2271609193961544e-06,
      "loss": 1.2437,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2242
    },
    {
      "epoch": 1.4071518193224593,
      "grad_norm": 1.7799817323684692,
      "learning_rate": 1.2248063464090733e-06,
      "loss": 1.287,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2243
    },
    {
      "epoch": 1.4077791718946049,
      "grad_norm": 1.674679160118103,
      "learning_rate": 1.2224533013822237e-06,
      "loss": 1.4167,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2244
    },
    {
      "epoch": 1.4084065244667503,
      "grad_norm": 1.686480164527893,
      "learning_rate": 1.2201017871350687e-06,
      "loss": 1.1705,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2245
    },
    {
      "epoch": 1.409033877038896,
      "grad_norm": 1.886971354484558,
      "learning_rate": 1.217751806485235e-06,
      "loss": 1.2974,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2246
    },
    {
      "epoch": 1.4096612296110413,
      "grad_norm": 1.6448758840560913,
      "learning_rate": 1.2154033622485114e-06,
      "loss": 1.1193,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2247
    },
    {
      "epoch": 1.410288582183187,
      "grad_norm": 1.8690279722213745,
      "learning_rate": 1.2130564572388476e-06,
      "loss": 1.2275,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2248
    },
    {
      "epoch": 1.4109159347553324,
      "grad_norm": 2.0430450439453125,
      "learning_rate": 1.2107110942683459e-06,
      "loss": 1.3148,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2249
    },
    {
      "epoch": 1.411543287327478,
      "grad_norm": 1.8752175569534302,
      "learning_rate": 1.208367276147265e-06,
      "loss": 1.2259,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2250
    },
    {
      "epoch": 1.4121706398996237,
      "grad_norm": 1.7567509412765503,
      "learning_rate": 1.206025005684009e-06,
      "loss": 1.0586,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2251
    },
    {
      "epoch": 1.412797992471769,
      "grad_norm": 1.49163818359375,
      "learning_rate": 1.2036842856851303e-06,
      "loss": 1.1927,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2252
    },
    {
      "epoch": 1.4134253450439147,
      "grad_norm": 2.017422914505005,
      "learning_rate": 1.2013451189553217e-06,
      "loss": 1.3638,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2253
    },
    {
      "epoch": 1.4140526976160603,
      "grad_norm": 1.748540997505188,
      "learning_rate": 1.199007508297414e-06,
      "loss": 1.2116,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2254
    },
    {
      "epoch": 1.4146800501882058,
      "grad_norm": 2.0524532794952393,
      "learning_rate": 1.196671456512377e-06,
      "loss": 1.2057,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2255
    },
    {
      "epoch": 1.4153074027603514,
      "grad_norm": 2.4321794509887695,
      "learning_rate": 1.1943369663993083e-06,
      "loss": 1.1794,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2256
    },
    {
      "epoch": 1.4159347553324968,
      "grad_norm": 1.8823251724243164,
      "learning_rate": 1.1920040407554384e-06,
      "loss": 1.2584,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2257
    },
    {
      "epoch": 1.4165621079046424,
      "grad_norm": 1.7821834087371826,
      "learning_rate": 1.1896726823761195e-06,
      "loss": 1.3296,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2258
    },
    {
      "epoch": 1.4171894604767878,
      "grad_norm": 1.4383490085601807,
      "learning_rate": 1.1873428940548293e-06,
      "loss": 1.222,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2259
    },
    {
      "epoch": 1.4178168130489335,
      "grad_norm": 1.8384606838226318,
      "learning_rate": 1.1850146785831618e-06,
      "loss": 1.2738,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2260
    },
    {
      "epoch": 1.4184441656210791,
      "grad_norm": 1.805746078491211,
      "learning_rate": 1.1826880387508258e-06,
      "loss": 1.2846,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2261
    },
    {
      "epoch": 1.4190715181932245,
      "grad_norm": 1.5791975259780884,
      "learning_rate": 1.180362977345646e-06,
      "loss": 1.1872,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2262
    },
    {
      "epoch": 1.4196988707653702,
      "grad_norm": 1.7926671504974365,
      "learning_rate": 1.1780394971535516e-06,
      "loss": 1.3442,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2263
    },
    {
      "epoch": 1.4203262233375158,
      "grad_norm": 1.82709801197052,
      "learning_rate": 1.1757176009585795e-06,
      "loss": 1.0753,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2264
    },
    {
      "epoch": 1.4209535759096612,
      "grad_norm": 1.4206048250198364,
      "learning_rate": 1.1733972915428665e-06,
      "loss": 1.1006,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2265
    },
    {
      "epoch": 1.4215809284818068,
      "grad_norm": 1.908329963684082,
      "learning_rate": 1.171078571686652e-06,
      "loss": 1.2698,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2266
    },
    {
      "epoch": 1.4222082810539523,
      "grad_norm": 2.473623752593994,
      "learning_rate": 1.1687614441682667e-06,
      "loss": 1.4153,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2267
    },
    {
      "epoch": 1.4228356336260979,
      "grad_norm": 2.12664532661438,
      "learning_rate": 1.1664459117641344e-06,
      "loss": 1.2839,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2268
    },
    {
      "epoch": 1.4234629861982433,
      "grad_norm": 1.7684519290924072,
      "learning_rate": 1.16413197724877e-06,
      "loss": 1.1967,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2269
    },
    {
      "epoch": 1.424090338770389,
      "grad_norm": 1.3864423036575317,
      "learning_rate": 1.16181964339477e-06,
      "loss": 1.1686,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2270
    },
    {
      "epoch": 1.4247176913425346,
      "grad_norm": 1.5991979837417603,
      "learning_rate": 1.1595089129728165e-06,
      "loss": 1.3387,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2271
    },
    {
      "epoch": 1.42534504391468,
      "grad_norm": 1.8228894472122192,
      "learning_rate": 1.1571997887516672e-06,
      "loss": 1.1316,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2272
    },
    {
      "epoch": 1.4259723964868256,
      "grad_norm": 1.7148045301437378,
      "learning_rate": 1.154892273498159e-06,
      "loss": 1.2169,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2273
    },
    {
      "epoch": 1.4265997490589712,
      "grad_norm": 1.4883177280426025,
      "learning_rate": 1.1525863699771967e-06,
      "loss": 1.3343,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2274
    },
    {
      "epoch": 1.4272271016311167,
      "grad_norm": 1.6968177556991577,
      "learning_rate": 1.1502820809517557e-06,
      "loss": 1.4402,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2275
    },
    {
      "epoch": 1.4278544542032623,
      "grad_norm": 1.5625579357147217,
      "learning_rate": 1.1479794091828787e-06,
      "loss": 1.2919,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2276
    },
    {
      "epoch": 1.4284818067754077,
      "grad_norm": 1.4449665546417236,
      "learning_rate": 1.1456783574296668e-06,
      "loss": 1.2676,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2277
    },
    {
      "epoch": 1.4291091593475533,
      "grad_norm": 1.6299374103546143,
      "learning_rate": 1.1433789284492842e-06,
      "loss": 1.1585,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2278
    },
    {
      "epoch": 1.4297365119196987,
      "grad_norm": 1.7700270414352417,
      "learning_rate": 1.1410811249969475e-06,
      "loss": 1.1956,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2279
    },
    {
      "epoch": 1.4303638644918444,
      "grad_norm": 2.047250509262085,
      "learning_rate": 1.1387849498259265e-06,
      "loss": 1.2824,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2280
    },
    {
      "epoch": 1.43099121706399,
      "grad_norm": 1.733454704284668,
      "learning_rate": 1.1364904056875395e-06,
      "loss": 1.2353,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2281
    },
    {
      "epoch": 1.4316185696361354,
      "grad_norm": 2.0308773517608643,
      "learning_rate": 1.1341974953311524e-06,
      "loss": 1.2762,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2282
    },
    {
      "epoch": 1.432245922208281,
      "grad_norm": 1.8377238512039185,
      "learning_rate": 1.1319062215041715e-06,
      "loss": 1.2808,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2283
    },
    {
      "epoch": 1.4328732747804267,
      "grad_norm": 1.727290153503418,
      "learning_rate": 1.129616586952042e-06,
      "loss": 1.1955,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2284
    },
    {
      "epoch": 1.433500627352572,
      "grad_norm": 1.7196539640426636,
      "learning_rate": 1.1273285944182468e-06,
      "loss": 1.1532,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2285
    },
    {
      "epoch": 1.4341279799247177,
      "grad_norm": 2.2919390201568604,
      "learning_rate": 1.1250422466442992e-06,
      "loss": 1.2986,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2286
    },
    {
      "epoch": 1.4347553324968632,
      "grad_norm": 1.6838818788528442,
      "learning_rate": 1.122757546369744e-06,
      "loss": 1.3198,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2287
    },
    {
      "epoch": 1.4353826850690088,
      "grad_norm": 1.4195574522018433,
      "learning_rate": 1.12047449633215e-06,
      "loss": 1.1265,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2288
    },
    {
      "epoch": 1.4360100376411542,
      "grad_norm": 1.9881501197814941,
      "learning_rate": 1.1181930992671078e-06,
      "loss": 1.2263,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2289
    },
    {
      "epoch": 1.4366373902132998,
      "grad_norm": 1.729966640472412,
      "learning_rate": 1.1159133579082315e-06,
      "loss": 1.1496,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2290
    },
    {
      "epoch": 1.4372647427854455,
      "grad_norm": 1.7988170385360718,
      "learning_rate": 1.1136352749871462e-06,
      "loss": 1.2376,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2291
    },
    {
      "epoch": 1.4378920953575909,
      "grad_norm": 1.5249148607254028,
      "learning_rate": 1.1113588532334946e-06,
      "loss": 1.1522,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2292
    },
    {
      "epoch": 1.4385194479297365,
      "grad_norm": 1.36632239818573,
      "learning_rate": 1.1090840953749253e-06,
      "loss": 1.3422,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2293
    },
    {
      "epoch": 1.4391468005018822,
      "grad_norm": 1.8093537092208862,
      "learning_rate": 1.1068110041370938e-06,
      "loss": 1.2057,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2294
    },
    {
      "epoch": 1.4397741530740276,
      "grad_norm": 2.2041056156158447,
      "learning_rate": 1.1045395822436617e-06,
      "loss": 1.3304,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2295
    },
    {
      "epoch": 1.4404015056461732,
      "grad_norm": 1.9485527276992798,
      "learning_rate": 1.102269832416286e-06,
      "loss": 1.2654,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2296
    },
    {
      "epoch": 1.4410288582183188,
      "grad_norm": 1.7615538835525513,
      "learning_rate": 1.1000017573746225e-06,
      "loss": 1.173,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2297
    },
    {
      "epoch": 1.4416562107904642,
      "grad_norm": 1.8701543807983398,
      "learning_rate": 1.097735359836321e-06,
      "loss": 1.221,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2298
    },
    {
      "epoch": 1.4422835633626097,
      "grad_norm": 1.8322738409042358,
      "learning_rate": 1.0954706425170198e-06,
      "loss": 1.3195,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2299
    },
    {
      "epoch": 1.4429109159347553,
      "grad_norm": 1.7369928359985352,
      "learning_rate": 1.0932076081303442e-06,
      "loss": 1.3012,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2300
    },
    {
      "epoch": 1.443538268506901,
      "grad_norm": 1.7098748683929443,
      "learning_rate": 1.0909462593879021e-06,
      "loss": 1.1589,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2301
    },
    {
      "epoch": 1.4441656210790463,
      "grad_norm": 1.913375735282898,
      "learning_rate": 1.0886865989992847e-06,
      "loss": 1.3016,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2302
    },
    {
      "epoch": 1.444792973651192,
      "grad_norm": 2.0486834049224854,
      "learning_rate": 1.0864286296720566e-06,
      "loss": 1.1933,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2303
    },
    {
      "epoch": 1.4454203262233376,
      "grad_norm": 2.107297420501709,
      "learning_rate": 1.0841723541117594e-06,
      "loss": 1.1922,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2304
    },
    {
      "epoch": 1.446047678795483,
      "grad_norm": 1.906996726989746,
      "learning_rate": 1.0819177750219013e-06,
      "loss": 1.2011,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2305
    },
    {
      "epoch": 1.4466750313676286,
      "grad_norm": 1.8115808963775635,
      "learning_rate": 1.0796648951039621e-06,
      "loss": 1.2571,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2306
    },
    {
      "epoch": 1.4473023839397743,
      "grad_norm": 1.6775972843170166,
      "learning_rate": 1.0774137170573826e-06,
      "loss": 1.0947,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2307
    },
    {
      "epoch": 1.4479297365119197,
      "grad_norm": 1.692162036895752,
      "learning_rate": 1.0751642435795642e-06,
      "loss": 1.2644,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2308
    },
    {
      "epoch": 1.448557089084065,
      "grad_norm": 1.9846559762954712,
      "learning_rate": 1.0729164773658692e-06,
      "loss": 1.2951,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2309
    },
    {
      "epoch": 1.4491844416562107,
      "grad_norm": 1.7943240404129028,
      "learning_rate": 1.0706704211096098e-06,
      "loss": 1.2268,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2310
    },
    {
      "epoch": 1.4498117942283564,
      "grad_norm": 1.8496971130371094,
      "learning_rate": 1.0684260775020532e-06,
      "loss": 1.2008,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2311
    },
    {
      "epoch": 1.4504391468005018,
      "grad_norm": 1.930126667022705,
      "learning_rate": 1.0661834492324121e-06,
      "loss": 1.2997,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2312
    },
    {
      "epoch": 1.4510664993726474,
      "grad_norm": 1.8001741170883179,
      "learning_rate": 1.0639425389878435e-06,
      "loss": 1.2232,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2313
    },
    {
      "epoch": 1.451693851944793,
      "grad_norm": 1.777064561843872,
      "learning_rate": 1.0617033494534486e-06,
      "loss": 1.1836,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2314
    },
    {
      "epoch": 1.4523212045169385,
      "grad_norm": 1.6934930086135864,
      "learning_rate": 1.0594658833122642e-06,
      "loss": 1.2539,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2315
    },
    {
      "epoch": 1.452948557089084,
      "grad_norm": 2.0415735244750977,
      "learning_rate": 1.0572301432452626e-06,
      "loss": 1.2492,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2316
    },
    {
      "epoch": 1.4535759096612297,
      "grad_norm": 1.4712687730789185,
      "learning_rate": 1.0549961319313476e-06,
      "loss": 1.145,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2317
    },
    {
      "epoch": 1.4542032622333751,
      "grad_norm": 1.9153761863708496,
      "learning_rate": 1.052763852047354e-06,
      "loss": 1.1663,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2318
    },
    {
      "epoch": 1.4548306148055208,
      "grad_norm": 1.9745244979858398,
      "learning_rate": 1.0505333062680383e-06,
      "loss": 1.2157,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2319
    },
    {
      "epoch": 1.4554579673776662,
      "grad_norm": 1.7918163537979126,
      "learning_rate": 1.048304497266083e-06,
      "loss": 1.151,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2320
    },
    {
      "epoch": 1.4560853199498118,
      "grad_norm": 1.8916409015655518,
      "learning_rate": 1.0460774277120866e-06,
      "loss": 1.1889,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2321
    },
    {
      "epoch": 1.4567126725219572,
      "grad_norm": 1.5514240264892578,
      "learning_rate": 1.0438521002745634e-06,
      "loss": 1.1767,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2322
    },
    {
      "epoch": 1.4573400250941029,
      "grad_norm": 1.4099035263061523,
      "learning_rate": 1.0416285176199425e-06,
      "loss": 1.2045,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2323
    },
    {
      "epoch": 1.4579673776662485,
      "grad_norm": 1.954008936882019,
      "learning_rate": 1.0394066824125604e-06,
      "loss": 1.2397,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2324
    },
    {
      "epoch": 1.458594730238394,
      "grad_norm": 1.4865626096725464,
      "learning_rate": 1.0371865973146613e-06,
      "loss": 1.1996,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2325
    },
    {
      "epoch": 1.4592220828105396,
      "grad_norm": 1.4902006387710571,
      "learning_rate": 1.03496826498639e-06,
      "loss": 1.1556,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2326
    },
    {
      "epoch": 1.4598494353826852,
      "grad_norm": 2.036508083343506,
      "learning_rate": 1.0327516880857949e-06,
      "loss": 1.3385,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2327
    },
    {
      "epoch": 1.4604767879548306,
      "grad_norm": 1.7831565141677856,
      "learning_rate": 1.0305368692688175e-06,
      "loss": 1.1753,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2328
    },
    {
      "epoch": 1.4611041405269762,
      "grad_norm": 1.7710812091827393,
      "learning_rate": 1.0283238111892929e-06,
      "loss": 1.202,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2329
    },
    {
      "epoch": 1.4617314930991216,
      "grad_norm": 1.7439368963241577,
      "learning_rate": 1.0261125164989497e-06,
      "loss": 1.3092,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2330
    },
    {
      "epoch": 1.4623588456712673,
      "grad_norm": 2.3144192695617676,
      "learning_rate": 1.0239029878474001e-06,
      "loss": 1.1785,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2331
    },
    {
      "epoch": 1.4629861982434127,
      "grad_norm": 1.8720273971557617,
      "learning_rate": 1.0216952278821423e-06,
      "loss": 1.2678,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2332
    },
    {
      "epoch": 1.4636135508155583,
      "grad_norm": 1.4683690071105957,
      "learning_rate": 1.019489239248553e-06,
      "loss": 1.0986,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2333
    },
    {
      "epoch": 1.464240903387704,
      "grad_norm": 1.8687940835952759,
      "learning_rate": 1.0172850245898893e-06,
      "loss": 1.213,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2334
    },
    {
      "epoch": 1.4648682559598494,
      "grad_norm": 1.6996122598648071,
      "learning_rate": 1.0150825865472813e-06,
      "loss": 1.1504,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2335
    },
    {
      "epoch": 1.465495608531995,
      "grad_norm": 1.8337829113006592,
      "learning_rate": 1.0128819277597287e-06,
      "loss": 1.2712,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2336
    },
    {
      "epoch": 1.4661229611041406,
      "grad_norm": 1.7464158535003662,
      "learning_rate": 1.0106830508641032e-06,
      "loss": 1.0795,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2337
    },
    {
      "epoch": 1.466750313676286,
      "grad_norm": 1.7541254758834839,
      "learning_rate": 1.008485958495137e-06,
      "loss": 1.3494,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2338
    },
    {
      "epoch": 1.4673776662484317,
      "grad_norm": 1.6948381662368774,
      "learning_rate": 1.0062906532854284e-06,
      "loss": 1.2058,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2339
    },
    {
      "epoch": 1.468005018820577,
      "grad_norm": 1.8352493047714233,
      "learning_rate": 1.0040971378654294e-06,
      "loss": 1.3595,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2340
    },
    {
      "epoch": 1.4686323713927227,
      "grad_norm": 1.7481765747070312,
      "learning_rate": 1.0019054148634525e-06,
      "loss": 1.3388,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2341
    },
    {
      "epoch": 1.4692597239648681,
      "grad_norm": 2.0751793384552,
      "learning_rate": 9.997154869056588e-07,
      "loss": 1.205,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2342
    },
    {
      "epoch": 1.4698870765370138,
      "grad_norm": 6.425692558288574,
      "learning_rate": 9.975273566160592e-07,
      "loss": 1.2013,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2343
    },
    {
      "epoch": 1.4705144291091594,
      "grad_norm": 1.7943389415740967,
      "learning_rate": 9.953410266165131e-07,
      "loss": 1.1807,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2344
    },
    {
      "epoch": 1.4711417816813048,
      "grad_norm": 1.886940836906433,
      "learning_rate": 9.93156499526719e-07,
      "loss": 1.0951,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2345
    },
    {
      "epoch": 1.4717691342534505,
      "grad_norm": 1.7454280853271484,
      "learning_rate": 9.909737779642187e-07,
      "loss": 1.0865,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2346
    },
    {
      "epoch": 1.472396486825596,
      "grad_norm": 1.8696519136428833,
      "learning_rate": 9.887928645443878e-07,
      "loss": 1.2695,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2347
    },
    {
      "epoch": 1.4730238393977415,
      "grad_norm": 1.8949179649353027,
      "learning_rate": 9.866137618804366e-07,
      "loss": 1.2531,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2348
    },
    {
      "epoch": 1.4736511919698871,
      "grad_norm": 1.9233238697052002,
      "learning_rate": 9.844364725834058e-07,
      "loss": 1.3932,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2349
    },
    {
      "epoch": 1.4742785445420326,
      "grad_norm": 1.780385136604309,
      "learning_rate": 9.822609992621618e-07,
      "loss": 1.2219,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2350
    },
    {
      "epoch": 1.4749058971141782,
      "grad_norm": 2.673675775527954,
      "learning_rate": 9.800873445233984e-07,
      "loss": 1.1437,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2351
    },
    {
      "epoch": 1.4755332496863236,
      "grad_norm": 1.3347368240356445,
      "learning_rate": 9.77915510971626e-07,
      "loss": 1.1426,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2352
    },
    {
      "epoch": 1.4761606022584692,
      "grad_norm": 1.907494306564331,
      "learning_rate": 9.757455012091774e-07,
      "loss": 1.2547,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2353
    },
    {
      "epoch": 1.4767879548306149,
      "grad_norm": 1.7126989364624023,
      "learning_rate": 9.735773178361965e-07,
      "loss": 1.3163,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2354
    },
    {
      "epoch": 1.4774153074027603,
      "grad_norm": 2.050947904586792,
      "learning_rate": 9.714109634506393e-07,
      "loss": 1.0594,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2355
    },
    {
      "epoch": 1.478042659974906,
      "grad_norm": 1.7997150421142578,
      "learning_rate": 9.692464406482727e-07,
      "loss": 1.2088,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2356
    },
    {
      "epoch": 1.4786700125470515,
      "grad_norm": 1.9468494653701782,
      "learning_rate": 9.67083752022666e-07,
      "loss": 1.2096,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2357
    },
    {
      "epoch": 1.479297365119197,
      "grad_norm": 1.8462549448013306,
      "learning_rate": 9.64922900165193e-07,
      "loss": 1.143,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2358
    },
    {
      "epoch": 1.4799247176913426,
      "grad_norm": 1.697049617767334,
      "learning_rate": 9.627638876650245e-07,
      "loss": 1.3182,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2359
    },
    {
      "epoch": 1.480552070263488,
      "grad_norm": 1.7648268938064575,
      "learning_rate": 9.606067171091301e-07,
      "loss": 1.0697,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2360
    },
    {
      "epoch": 1.4811794228356336,
      "grad_norm": 1.8276077508926392,
      "learning_rate": 9.584513910822704e-07,
      "loss": 1.3043,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2361
    },
    {
      "epoch": 1.481806775407779,
      "grad_norm": 1.9977463483810425,
      "learning_rate": 9.562979121669946e-07,
      "loss": 1.3174,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2362
    },
    {
      "epoch": 1.4824341279799247,
      "grad_norm": 1.7347925901412964,
      "learning_rate": 9.541462829436426e-07,
      "loss": 1.2394,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2363
    },
    {
      "epoch": 1.4830614805520703,
      "grad_norm": 1.7249211072921753,
      "learning_rate": 9.519965059903349e-07,
      "loss": 1.1495,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2364
    },
    {
      "epoch": 1.4836888331242157,
      "grad_norm": 1.6533867120742798,
      "learning_rate": 9.498485838829727e-07,
      "loss": 1.3614,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2365
    },
    {
      "epoch": 1.4843161856963614,
      "grad_norm": 1.8468972444534302,
      "learning_rate": 9.477025191952352e-07,
      "loss": 1.36,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2366
    },
    {
      "epoch": 1.484943538268507,
      "grad_norm": 2.0067813396453857,
      "learning_rate": 9.455583144985772e-07,
      "loss": 1.2394,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2367
    },
    {
      "epoch": 1.4855708908406524,
      "grad_norm": 1.959493637084961,
      "learning_rate": 9.434159723622235e-07,
      "loss": 1.2166,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2368
    },
    {
      "epoch": 1.486198243412798,
      "grad_norm": 1.8856251239776611,
      "learning_rate": 9.412754953531664e-07,
      "loss": 1.2236,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2369
    },
    {
      "epoch": 1.4868255959849435,
      "grad_norm": 1.546231985092163,
      "learning_rate": 9.39136886036166e-07,
      "loss": 1.1945,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2370
    },
    {
      "epoch": 1.487452948557089,
      "grad_norm": 1.9689160585403442,
      "learning_rate": 9.370001469737411e-07,
      "loss": 1.2117,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2371
    },
    {
      "epoch": 1.4880803011292345,
      "grad_norm": 1.8257304430007935,
      "learning_rate": 9.348652807261732e-07,
      "loss": 1.2046,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2372
    },
    {
      "epoch": 1.4887076537013801,
      "grad_norm": 1.7765759229660034,
      "learning_rate": 9.32732289851496e-07,
      "loss": 1.1348,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2373
    },
    {
      "epoch": 1.4893350062735258,
      "grad_norm": 1.97781240940094,
      "learning_rate": 9.306011769054999e-07,
      "loss": 1.1807,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2374
    },
    {
      "epoch": 1.4899623588456712,
      "grad_norm": 1.512803316116333,
      "learning_rate": 9.28471944441722e-07,
      "loss": 1.1873,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2375
    },
    {
      "epoch": 1.4905897114178168,
      "grad_norm": 1.8602982759475708,
      "learning_rate": 9.263445950114469e-07,
      "loss": 1.2918,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2376
    },
    {
      "epoch": 1.4912170639899625,
      "grad_norm": 1.9396471977233887,
      "learning_rate": 9.24219131163705e-07,
      "loss": 1.3401,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2377
    },
    {
      "epoch": 1.4918444165621079,
      "grad_norm": 1.779157280921936,
      "learning_rate": 9.220955554452637e-07,
      "loss": 1.1911,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2378
    },
    {
      "epoch": 1.4924717691342535,
      "grad_norm": 1.6677671670913696,
      "learning_rate": 9.199738704006322e-07,
      "loss": 1.2692,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2379
    },
    {
      "epoch": 1.4930991217063991,
      "grad_norm": 1.3902426958084106,
      "learning_rate": 9.17854078572051e-07,
      "loss": 1.2505,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2380
    },
    {
      "epoch": 1.4937264742785445,
      "grad_norm": 1.772935152053833,
      "learning_rate": 9.157361824994937e-07,
      "loss": 1.2273,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2381
    },
    {
      "epoch": 1.49435382685069,
      "grad_norm": 1.8907947540283203,
      "learning_rate": 9.13620184720661e-07,
      "loss": 1.2794,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2382
    },
    {
      "epoch": 1.4949811794228356,
      "grad_norm": 1.882448434829712,
      "learning_rate": 9.1150608777098e-07,
      "loss": 1.1436,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2383
    },
    {
      "epoch": 1.4956085319949812,
      "grad_norm": 2.5027365684509277,
      "learning_rate": 9.093938941836012e-07,
      "loss": 1.1959,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2384
    },
    {
      "epoch": 1.4962358845671266,
      "grad_norm": 1.830198049545288,
      "learning_rate": 9.072836064893917e-07,
      "loss": 1.327,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2385
    },
    {
      "epoch": 1.4968632371392723,
      "grad_norm": 1.606887698173523,
      "learning_rate": 9.051752272169386e-07,
      "loss": 1.2297,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2386
    },
    {
      "epoch": 1.497490589711418,
      "grad_norm": 1.3970919847488403,
      "learning_rate": 9.030687588925388e-07,
      "loss": 1.1981,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2387
    },
    {
      "epoch": 1.4981179422835633,
      "grad_norm": 2.031083822250366,
      "learning_rate": 9.009642040402022e-07,
      "loss": 1.2114,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2388
    },
    {
      "epoch": 1.498745294855709,
      "grad_norm": 1.7160199880599976,
      "learning_rate": 8.98861565181644e-07,
      "loss": 1.3597,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2389
    },
    {
      "epoch": 1.4993726474278546,
      "grad_norm": 1.7950799465179443,
      "learning_rate": 8.967608448362836e-07,
      "loss": 1.1441,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2390
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.7221378087997437,
      "learning_rate": 8.946620455212438e-07,
      "loss": 1.2303,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2391
    },
    {
      "epoch": 1.5006273525721454,
      "grad_norm": 1.899416208267212,
      "learning_rate": 8.925651697513424e-07,
      "loss": 1.3782,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2392
    },
    {
      "epoch": 1.501254705144291,
      "grad_norm": 1.5485550165176392,
      "learning_rate": 8.90470220039096e-07,
      "loss": 1.1964,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2393
    },
    {
      "epoch": 1.5018820577164367,
      "grad_norm": 1.6856316328048706,
      "learning_rate": 8.883771988947099e-07,
      "loss": 1.3119,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2394
    },
    {
      "epoch": 1.502509410288582,
      "grad_norm": 1.5902036428451538,
      "learning_rate": 8.862861088260816e-07,
      "loss": 1.1776,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2395
    },
    {
      "epoch": 1.5031367628607277,
      "grad_norm": 2.432138442993164,
      "learning_rate": 8.841969523387922e-07,
      "loss": 1.4684,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2396
    },
    {
      "epoch": 1.5037641154328734,
      "grad_norm": 1.3001375198364258,
      "learning_rate": 8.821097319361074e-07,
      "loss": 1.0717,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2397
    },
    {
      "epoch": 1.5043914680050188,
      "grad_norm": 1.747187614440918,
      "learning_rate": 8.800244501189722e-07,
      "loss": 1.3016,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2398
    },
    {
      "epoch": 1.5050188205771644,
      "grad_norm": 1.7088016271591187,
      "learning_rate": 8.77941109386009e-07,
      "loss": 1.2023,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2399
    },
    {
      "epoch": 1.50564617314931,
      "grad_norm": 1.6368725299835205,
      "learning_rate": 8.758597122335158e-07,
      "loss": 1.1448,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2400
    },
    {
      "epoch": 1.5062735257214555,
      "grad_norm": 1.5822457075119019,
      "learning_rate": 8.737802611554596e-07,
      "loss": 1.1753,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2401
    },
    {
      "epoch": 1.5069008782936009,
      "grad_norm": 1.209599256515503,
      "learning_rate": 8.717027586434759e-07,
      "loss": 1.1381,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2402
    },
    {
      "epoch": 1.5075282308657465,
      "grad_norm": 1.73714017868042,
      "learning_rate": 8.696272071868678e-07,
      "loss": 1.1301,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2403
    },
    {
      "epoch": 1.5081555834378921,
      "grad_norm": 2.023517370223999,
      "learning_rate": 8.675536092725967e-07,
      "loss": 1.2922,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2404
    },
    {
      "epoch": 1.5087829360100375,
      "grad_norm": 1.518594741821289,
      "learning_rate": 8.654819673852874e-07,
      "loss": 1.2086,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2405
    },
    {
      "epoch": 1.5094102885821832,
      "grad_norm": 2.004223108291626,
      "learning_rate": 8.634122840072171e-07,
      "loss": 1.3216,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2406
    },
    {
      "epoch": 1.5100376411543288,
      "grad_norm": 1.823880672454834,
      "learning_rate": 8.613445616183197e-07,
      "loss": 1.3131,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2407
    },
    {
      "epoch": 1.5106649937264742,
      "grad_norm": 1.5020653009414673,
      "learning_rate": 8.592788026961769e-07,
      "loss": 1.2801,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2408
    },
    {
      "epoch": 1.5112923462986199,
      "grad_norm": 1.8484547138214111,
      "learning_rate": 8.572150097160179e-07,
      "loss": 1.1977,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2409
    },
    {
      "epoch": 1.5119196988707655,
      "grad_norm": 1.4522210359573364,
      "learning_rate": 8.551531851507186e-07,
      "loss": 1.1126,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2410
    },
    {
      "epoch": 1.512547051442911,
      "grad_norm": 1.6940292119979858,
      "learning_rate": 8.530933314707932e-07,
      "loss": 1.1495,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2411
    },
    {
      "epoch": 1.5131744040150563,
      "grad_norm": 2.1096079349517822,
      "learning_rate": 8.510354511443975e-07,
      "loss": 1.3074,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2412
    },
    {
      "epoch": 1.513801756587202,
      "grad_norm": 1.2838313579559326,
      "learning_rate": 8.489795466373199e-07,
      "loss": 1.2711,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2413
    },
    {
      "epoch": 1.5144291091593476,
      "grad_norm": 1.7509784698486328,
      "learning_rate": 8.469256204129827e-07,
      "loss": 1.3424,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2414
    },
    {
      "epoch": 1.515056461731493,
      "grad_norm": 1.9112739562988281,
      "learning_rate": 8.448736749324382e-07,
      "loss": 1.2579,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2415
    },
    {
      "epoch": 1.5156838143036386,
      "grad_norm": 1.7965364456176758,
      "learning_rate": 8.428237126543634e-07,
      "loss": 1.2314,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2416
    },
    {
      "epoch": 1.5163111668757843,
      "grad_norm": 1.9476574659347534,
      "learning_rate": 8.407757360350624e-07,
      "loss": 1.32,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2417
    },
    {
      "epoch": 1.5169385194479297,
      "grad_norm": 1.7615290880203247,
      "learning_rate": 8.387297475284562e-07,
      "loss": 1.2656,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2418
    },
    {
      "epoch": 1.5175658720200753,
      "grad_norm": 2.400646686553955,
      "learning_rate": 8.366857495860869e-07,
      "loss": 1.2182,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2419
    },
    {
      "epoch": 1.518193224592221,
      "grad_norm": 2.004615306854248,
      "learning_rate": 8.34643744657109e-07,
      "loss": 1.1982,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2420
    },
    {
      "epoch": 1.5188205771643664,
      "grad_norm": 1.8327479362487793,
      "learning_rate": 8.32603735188291e-07,
      "loss": 1.2677,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2421
    },
    {
      "epoch": 1.5194479297365118,
      "grad_norm": 1.8293399810791016,
      "learning_rate": 8.305657236240086e-07,
      "loss": 1.2551,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2422
    },
    {
      "epoch": 1.5200752823086576,
      "grad_norm": 1.6446075439453125,
      "learning_rate": 8.285297124062439e-07,
      "loss": 1.314,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2423
    },
    {
      "epoch": 1.520702634880803,
      "grad_norm": 1.8601378202438354,
      "learning_rate": 8.264957039745835e-07,
      "loss": 1.2594,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2424
    },
    {
      "epoch": 1.5213299874529485,
      "grad_norm": 1.7208938598632812,
      "learning_rate": 8.244637007662126e-07,
      "loss": 1.2329,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2425
    },
    {
      "epoch": 1.521957340025094,
      "grad_norm": 1.8246501684188843,
      "learning_rate": 8.224337052159154e-07,
      "loss": 1.1892,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2426
    },
    {
      "epoch": 1.5225846925972397,
      "grad_norm": 1.7943869829177856,
      "learning_rate": 8.204057197560681e-07,
      "loss": 1.1979,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2427
    },
    {
      "epoch": 1.5232120451693851,
      "grad_norm": 1.817122220993042,
      "learning_rate": 8.18379746816641e-07,
      "loss": 1.3934,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2428
    },
    {
      "epoch": 1.5238393977415308,
      "grad_norm": 1.78376305103302,
      "learning_rate": 8.163557888251916e-07,
      "loss": 1.3509,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2429
    },
    {
      "epoch": 1.5244667503136764,
      "grad_norm": 2.2729251384735107,
      "learning_rate": 8.143338482068628e-07,
      "loss": 1.1322,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2430
    },
    {
      "epoch": 1.5250941028858218,
      "grad_norm": 2.0838749408721924,
      "learning_rate": 8.123139273843802e-07,
      "loss": 1.1556,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2431
    },
    {
      "epoch": 1.5257214554579672,
      "grad_norm": 1.7555264234542847,
      "learning_rate": 8.102960287780498e-07,
      "loss": 1.216,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2432
    },
    {
      "epoch": 1.526348808030113,
      "grad_norm": 3.866431474685669,
      "learning_rate": 8.082801548057553e-07,
      "loss": 1.2201,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2433
    },
    {
      "epoch": 1.5269761606022585,
      "grad_norm": 1.5737709999084473,
      "learning_rate": 8.062663078829524e-07,
      "loss": 1.2154,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2434
    },
    {
      "epoch": 1.527603513174404,
      "grad_norm": 1.448691964149475,
      "learning_rate": 8.042544904226704e-07,
      "loss": 1.2301,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2435
    },
    {
      "epoch": 1.5282308657465495,
      "grad_norm": 2.6556127071380615,
      "learning_rate": 8.02244704835505e-07,
      "loss": 1.2506,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2436
    },
    {
      "epoch": 1.5288582183186952,
      "grad_norm": 1.8680992126464844,
      "learning_rate": 8.002369535296168e-07,
      "loss": 1.2128,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2437
    },
    {
      "epoch": 1.5294855708908406,
      "grad_norm": 1.5463523864746094,
      "learning_rate": 7.982312389107319e-07,
      "loss": 1.1466,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2438
    },
    {
      "epoch": 1.5301129234629862,
      "grad_norm": 1.889672040939331,
      "learning_rate": 7.962275633821321e-07,
      "loss": 1.3728,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2439
    },
    {
      "epoch": 1.5307402760351319,
      "grad_norm": 1.8735464811325073,
      "learning_rate": 7.942259293446594e-07,
      "loss": 1.2968,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2440
    },
    {
      "epoch": 1.5313676286072773,
      "grad_norm": 1.7368452548980713,
      "learning_rate": 7.922263391967069e-07,
      "loss": 1.2953,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2441
    },
    {
      "epoch": 1.5319949811794227,
      "grad_norm": 1.8116888999938965,
      "learning_rate": 7.902287953342213e-07,
      "loss": 1.2372,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2442
    },
    {
      "epoch": 1.5326223337515685,
      "grad_norm": 1.717750906944275,
      "learning_rate": 7.882333001506953e-07,
      "loss": 1.1085,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2443
    },
    {
      "epoch": 1.533249686323714,
      "grad_norm": 2.0107264518737793,
      "learning_rate": 7.862398560371665e-07,
      "loss": 1.1305,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2444
    },
    {
      "epoch": 1.5338770388958594,
      "grad_norm": 1.9042125940322876,
      "learning_rate": 7.842484653822177e-07,
      "loss": 1.2038,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2445
    },
    {
      "epoch": 1.534504391468005,
      "grad_norm": 1.7293494939804077,
      "learning_rate": 7.822591305719684e-07,
      "loss": 1.1182,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2446
    },
    {
      "epoch": 1.5351317440401506,
      "grad_norm": 1.9139995574951172,
      "learning_rate": 7.802718539900761e-07,
      "loss": 1.2507,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2447
    },
    {
      "epoch": 1.535759096612296,
      "grad_norm": 1.5856516361236572,
      "learning_rate": 7.782866380177307e-07,
      "loss": 1.285,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2448
    },
    {
      "epoch": 1.5363864491844417,
      "grad_norm": 1.739535927772522,
      "learning_rate": 7.763034850336554e-07,
      "loss": 1.2812,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2449
    },
    {
      "epoch": 1.5370138017565873,
      "grad_norm": 1.9271031618118286,
      "learning_rate": 7.743223974141001e-07,
      "loss": 1.1226,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2450
    },
    {
      "epoch": 1.5376411543287327,
      "grad_norm": 1.4703642129898071,
      "learning_rate": 7.723433775328385e-07,
      "loss": 1.1203,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2451
    },
    {
      "epoch": 1.5382685069008781,
      "grad_norm": 1.5959513187408447,
      "learning_rate": 7.703664277611702e-07,
      "loss": 1.381,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2452
    },
    {
      "epoch": 1.538895859473024,
      "grad_norm": 1.8221240043640137,
      "learning_rate": 7.683915504679109e-07,
      "loss": 1.2892,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2453
    },
    {
      "epoch": 1.5395232120451694,
      "grad_norm": 1.7736937999725342,
      "learning_rate": 7.66418748019396e-07,
      "loss": 1.0808,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2454
    },
    {
      "epoch": 1.5401505646173148,
      "grad_norm": 1.6039260625839233,
      "learning_rate": 7.644480227794723e-07,
      "loss": 1.2808,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2455
    },
    {
      "epoch": 1.5407779171894604,
      "grad_norm": 1.7460005283355713,
      "learning_rate": 7.624793771094979e-07,
      "loss": 1.1979,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2456
    },
    {
      "epoch": 1.541405269761606,
      "grad_norm": 1.8570126295089722,
      "learning_rate": 7.605128133683414e-07,
      "loss": 1.104,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2457
    },
    {
      "epoch": 1.5420326223337515,
      "grad_norm": 2.0892014503479004,
      "learning_rate": 7.585483339123739e-07,
      "loss": 1.3555,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2458
    },
    {
      "epoch": 1.5426599749058971,
      "grad_norm": 1.9000020027160645,
      "learning_rate": 7.565859410954718e-07,
      "loss": 1.401,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2459
    },
    {
      "epoch": 1.5432873274780428,
      "grad_norm": 2.0705411434173584,
      "learning_rate": 7.546256372690086e-07,
      "loss": 1.2129,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2460
    },
    {
      "epoch": 1.5439146800501882,
      "grad_norm": 1.8334147930145264,
      "learning_rate": 7.526674247818569e-07,
      "loss": 1.3082,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2461
    },
    {
      "epoch": 1.5445420326223338,
      "grad_norm": 1.836845874786377,
      "learning_rate": 7.507113059803825e-07,
      "loss": 1.2692,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2462
    },
    {
      "epoch": 1.5451693851944794,
      "grad_norm": 1.8347852230072021,
      "learning_rate": 7.48757283208442e-07,
      "loss": 1.1857,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2463
    },
    {
      "epoch": 1.5457967377666249,
      "grad_norm": 1.5334773063659668,
      "learning_rate": 7.468053588073803e-07,
      "loss": 1.1636,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2464
    },
    {
      "epoch": 1.5464240903387703,
      "grad_norm": 1.8923527002334595,
      "learning_rate": 7.448555351160308e-07,
      "loss": 1.3495,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2465
    },
    {
      "epoch": 1.547051442910916,
      "grad_norm": 2.0651772022247314,
      "learning_rate": 7.429078144707064e-07,
      "loss": 1.145,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2466
    },
    {
      "epoch": 1.5476787954830615,
      "grad_norm": 1.8705108165740967,
      "learning_rate": 7.409621992052007e-07,
      "loss": 1.207,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2467
    },
    {
      "epoch": 1.548306148055207,
      "grad_norm": 1.8827093839645386,
      "learning_rate": 7.390186916507869e-07,
      "loss": 1.2828,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2468
    },
    {
      "epoch": 1.5489335006273526,
      "grad_norm": 1.8180711269378662,
      "learning_rate": 7.370772941362106e-07,
      "loss": 1.3038,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2469
    },
    {
      "epoch": 1.5495608531994982,
      "grad_norm": 1.8148471117019653,
      "learning_rate": 7.351380089876886e-07,
      "loss": 1.1912,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2470
    },
    {
      "epoch": 1.5501882057716436,
      "grad_norm": 1.8098262548446655,
      "learning_rate": 7.332008385289091e-07,
      "loss": 1.1659,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2471
    },
    {
      "epoch": 1.5508155583437893,
      "grad_norm": 1.2533210515975952,
      "learning_rate": 7.312657850810234e-07,
      "loss": 1.0858,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2472
    },
    {
      "epoch": 1.551442910915935,
      "grad_norm": 1.871936559677124,
      "learning_rate": 7.293328509626497e-07,
      "loss": 1.2199,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2473
    },
    {
      "epoch": 1.5520702634880803,
      "grad_norm": 1.6073999404907227,
      "learning_rate": 7.274020384898628e-07,
      "loss": 1.2168,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2474
    },
    {
      "epoch": 1.5526976160602257,
      "grad_norm": 1.662189245223999,
      "learning_rate": 7.254733499761993e-07,
      "loss": 1.239,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2475
    },
    {
      "epoch": 1.5533249686323714,
      "grad_norm": 1.7758508920669556,
      "learning_rate": 7.235467877326476e-07,
      "loss": 1.3394,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2476
    },
    {
      "epoch": 1.553952321204517,
      "grad_norm": 1.8011035919189453,
      "learning_rate": 7.216223540676495e-07,
      "loss": 1.4568,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2477
    },
    {
      "epoch": 1.5545796737766624,
      "grad_norm": 5.772345542907715,
      "learning_rate": 7.197000512870974e-07,
      "loss": 1.3203,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2478
    },
    {
      "epoch": 1.555207026348808,
      "grad_norm": 1.6594387292861938,
      "learning_rate": 7.177798816943288e-07,
      "loss": 1.1403,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2479
    },
    {
      "epoch": 1.5558343789209537,
      "grad_norm": 2.6319494247436523,
      "learning_rate": 7.158618475901252e-07,
      "loss": 1.2109,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2480
    },
    {
      "epoch": 1.556461731493099,
      "grad_norm": 1.813653588294983,
      "learning_rate": 7.139459512727115e-07,
      "loss": 1.1186,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2481
    },
    {
      "epoch": 1.5570890840652447,
      "grad_norm": 1.7633485794067383,
      "learning_rate": 7.120321950377487e-07,
      "loss": 1.148,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2482
    },
    {
      "epoch": 1.5577164366373903,
      "grad_norm": 1.852625846862793,
      "learning_rate": 7.101205811783343e-07,
      "loss": 1.1101,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2483
    },
    {
      "epoch": 1.5583437892095358,
      "grad_norm": 1.7904287576675415,
      "learning_rate": 7.08211111984998e-07,
      "loss": 1.2955,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2484
    },
    {
      "epoch": 1.5589711417816812,
      "grad_norm": 1.7286176681518555,
      "learning_rate": 7.063037897457023e-07,
      "loss": 1.2696,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2485
    },
    {
      "epoch": 1.5595984943538268,
      "grad_norm": 1.7110837697982788,
      "learning_rate": 7.043986167458336e-07,
      "loss": 1.2207,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2486
    },
    {
      "epoch": 1.5602258469259724,
      "grad_norm": 1.8859424591064453,
      "learning_rate": 7.024955952682064e-07,
      "loss": 1.1963,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2487
    },
    {
      "epoch": 1.5608531994981178,
      "grad_norm": 1.711631417274475,
      "learning_rate": 7.005947275930541e-07,
      "loss": 1.1367,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2488
    },
    {
      "epoch": 1.5614805520702635,
      "grad_norm": 1.6996835470199585,
      "learning_rate": 6.986960159980327e-07,
      "loss": 1.1912,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2489
    },
    {
      "epoch": 1.5621079046424091,
      "grad_norm": 1.7976754903793335,
      "learning_rate": 6.96799462758212e-07,
      "loss": 1.222,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2490
    },
    {
      "epoch": 1.5627352572145545,
      "grad_norm": 1.5648021697998047,
      "learning_rate": 6.949050701460758e-07,
      "loss": 1.09,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2491
    },
    {
      "epoch": 1.5633626097867002,
      "grad_norm": 1.8570146560668945,
      "learning_rate": 6.930128404315214e-07,
      "loss": 1.2234,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2492
    },
    {
      "epoch": 1.5639899623588458,
      "grad_norm": 1.8520839214324951,
      "learning_rate": 6.911227758818515e-07,
      "loss": 1.2482,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2493
    },
    {
      "epoch": 1.5646173149309912,
      "grad_norm": 2.30804181098938,
      "learning_rate": 6.89234878761777e-07,
      "loss": 1.2595,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2494
    },
    {
      "epoch": 1.5652446675031366,
      "grad_norm": 1.8586586713790894,
      "learning_rate": 6.873491513334099e-07,
      "loss": 1.304,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2495
    },
    {
      "epoch": 1.5658720200752823,
      "grad_norm": 1.7122609615325928,
      "learning_rate": 6.854655958562625e-07,
      "loss": 1.3498,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2496
    },
    {
      "epoch": 1.566499372647428,
      "grad_norm": 1.9733279943466187,
      "learning_rate": 6.835842145872468e-07,
      "loss": 1.2857,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2497
    },
    {
      "epoch": 1.5671267252195733,
      "grad_norm": 1.871863842010498,
      "learning_rate": 6.817050097806674e-07,
      "loss": 1.1074,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2498
    },
    {
      "epoch": 1.567754077791719,
      "grad_norm": 1.6728519201278687,
      "learning_rate": 6.798279836882211e-07,
      "loss": 1.2012,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2499
    },
    {
      "epoch": 1.5683814303638646,
      "grad_norm": 1.7090595960617065,
      "learning_rate": 6.779531385589944e-07,
      "loss": 1.1679,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2500
    },
    {
      "epoch": 1.56900878293601,
      "grad_norm": 1.8407670259475708,
      "learning_rate": 6.760804766394624e-07,
      "loss": 1.1853,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2501
    },
    {
      "epoch": 1.5696361355081556,
      "grad_norm": 1.6820379495620728,
      "learning_rate": 6.742100001734817e-07,
      "loss": 1.1855,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2502
    },
    {
      "epoch": 1.5702634880803013,
      "grad_norm": 1.8091737031936646,
      "learning_rate": 6.723417114022907e-07,
      "loss": 1.1984,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2503
    },
    {
      "epoch": 1.5708908406524467,
      "grad_norm": 1.671722173690796,
      "learning_rate": 6.704756125645082e-07,
      "loss": 1.3158,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2504
    },
    {
      "epoch": 1.571518193224592,
      "grad_norm": 1.5467654466629028,
      "learning_rate": 6.686117058961267e-07,
      "loss": 1.1931,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2505
    },
    {
      "epoch": 1.572145545796738,
      "grad_norm": 1.8012003898620605,
      "learning_rate": 6.667499936305142e-07,
      "loss": 1.202,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2506
    },
    {
      "epoch": 1.5727728983688833,
      "grad_norm": 1.6461151838302612,
      "learning_rate": 6.648904779984067e-07,
      "loss": 1.0794,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2507
    },
    {
      "epoch": 1.5734002509410288,
      "grad_norm": 1.5227487087249756,
      "learning_rate": 6.630331612279112e-07,
      "loss": 1.2715,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2508
    },
    {
      "epoch": 1.5740276035131744,
      "grad_norm": 1.8711963891983032,
      "learning_rate": 6.61178045544498e-07,
      "loss": 1.2422,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2509
    },
    {
      "epoch": 1.57465495608532,
      "grad_norm": 1.4674845933914185,
      "learning_rate": 6.593251331709993e-07,
      "loss": 1.181,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2510
    },
    {
      "epoch": 1.5752823086574654,
      "grad_norm": 1.8265960216522217,
      "learning_rate": 6.5747442632761e-07,
      "loss": 1.2212,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2511
    },
    {
      "epoch": 1.575909661229611,
      "grad_norm": 1.6680943965911865,
      "learning_rate": 6.556259272318797e-07,
      "loss": 1.1853,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2512
    },
    {
      "epoch": 1.5765370138017567,
      "grad_norm": 1.9921751022338867,
      "learning_rate": 6.537796380987144e-07,
      "loss": 1.1803,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2513
    },
    {
      "epoch": 1.5771643663739021,
      "grad_norm": 1.7228193283081055,
      "learning_rate": 6.51935561140371e-07,
      "loss": 1.3567,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2514
    },
    {
      "epoch": 1.5777917189460475,
      "grad_norm": 1.6541950702667236,
      "learning_rate": 6.500936985664561e-07,
      "loss": 1.2213,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2515
    },
    {
      "epoch": 1.5784190715181934,
      "grad_norm": 1.579052209854126,
      "learning_rate": 6.482540525839226e-07,
      "loss": 1.2772,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2516
    },
    {
      "epoch": 1.5790464240903388,
      "grad_norm": 1.8425257205963135,
      "learning_rate": 6.464166253970672e-07,
      "loss": 1.2691,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2517
    },
    {
      "epoch": 1.5796737766624842,
      "grad_norm": 1.6395012140274048,
      "learning_rate": 6.445814192075304e-07,
      "loss": 1.1531,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2518
    },
    {
      "epoch": 1.5803011292346298,
      "grad_norm": 1.2978256940841675,
      "learning_rate": 6.427484362142877e-07,
      "loss": 1.1268,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2519
    },
    {
      "epoch": 1.5809284818067755,
      "grad_norm": 1.783844232559204,
      "learning_rate": 6.409176786136542e-07,
      "loss": 1.1135,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2520
    },
    {
      "epoch": 1.5815558343789209,
      "grad_norm": 1.8337434530258179,
      "learning_rate": 6.390891485992757e-07,
      "loss": 1.2873,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2521
    },
    {
      "epoch": 1.5821831869510665,
      "grad_norm": 1.6792529821395874,
      "learning_rate": 6.372628483621315e-07,
      "loss": 1.1688,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2522
    },
    {
      "epoch": 1.5828105395232122,
      "grad_norm": 1.58247971534729,
      "learning_rate": 6.35438780090527e-07,
      "loss": 1.1782,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2523
    },
    {
      "epoch": 1.5834378920953576,
      "grad_norm": 1.7473936080932617,
      "learning_rate": 6.336169459700933e-07,
      "loss": 1.1864,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2524
    },
    {
      "epoch": 1.584065244667503,
      "grad_norm": 1.6084119081497192,
      "learning_rate": 6.317973481837866e-07,
      "loss": 1.2161,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2525
    },
    {
      "epoch": 1.5846925972396488,
      "grad_norm": 1.8634337186813354,
      "learning_rate": 6.299799889118808e-07,
      "loss": 1.1492,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2526
    },
    {
      "epoch": 1.5853199498117942,
      "grad_norm": 1.4772895574569702,
      "learning_rate": 6.281648703319698e-07,
      "loss": 1.1645,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2527
    },
    {
      "epoch": 1.5859473023839397,
      "grad_norm": 1.8228148221969604,
      "learning_rate": 6.263519946189608e-07,
      "loss": 1.2619,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2528
    },
    {
      "epoch": 1.5865746549560853,
      "grad_norm": 1.7683981657028198,
      "learning_rate": 6.245413639450757e-07,
      "loss": 1.1456,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2529
    },
    {
      "epoch": 1.587202007528231,
      "grad_norm": 1.4007647037506104,
      "learning_rate": 6.227329804798444e-07,
      "loss": 1.1303,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2530
    },
    {
      "epoch": 1.5878293601003763,
      "grad_norm": 1.8269098997116089,
      "learning_rate": 6.209268463901047e-07,
      "loss": 1.2663,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2531
    },
    {
      "epoch": 1.588456712672522,
      "grad_norm": 2.1203908920288086,
      "learning_rate": 6.191229638399995e-07,
      "loss": 1.1958,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2532
    },
    {
      "epoch": 1.5890840652446676,
      "grad_norm": 1.8962397575378418,
      "learning_rate": 6.17321334990973e-07,
      "loss": 1.3296,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2533
    },
    {
      "epoch": 1.589711417816813,
      "grad_norm": 1.7916409969329834,
      "learning_rate": 6.155219620017707e-07,
      "loss": 1.2352,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2534
    },
    {
      "epoch": 1.5903387703889584,
      "grad_norm": 1.7805792093276978,
      "learning_rate": 6.137248470284332e-07,
      "loss": 1.4115,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2535
    },
    {
      "epoch": 1.5909661229611043,
      "grad_norm": 1.6655528545379639,
      "learning_rate": 6.119299922242975e-07,
      "loss": 1.1256,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2536
    },
    {
      "epoch": 1.5915934755332497,
      "grad_norm": 2.02083420753479,
      "learning_rate": 6.101373997399907e-07,
      "loss": 1.3301,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2537
    },
    {
      "epoch": 1.5922208281053951,
      "grad_norm": 1.6873060464859009,
      "learning_rate": 6.083470717234285e-07,
      "loss": 1.2483,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2538
    },
    {
      "epoch": 1.5928481806775407,
      "grad_norm": 1.4721137285232544,
      "learning_rate": 6.065590103198166e-07,
      "loss": 1.2483,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2539
    },
    {
      "epoch": 1.5934755332496864,
      "grad_norm": 1.7473325729370117,
      "learning_rate": 6.047732176716409e-07,
      "loss": 1.2648,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2540
    },
    {
      "epoch": 1.5941028858218318,
      "grad_norm": 1.768115520477295,
      "learning_rate": 6.029896959186721e-07,
      "loss": 1.3562,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2541
    },
    {
      "epoch": 1.5947302383939774,
      "grad_norm": 1.7181570529937744,
      "learning_rate": 6.012084471979571e-07,
      "loss": 1.1847,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2542
    },
    {
      "epoch": 1.595357590966123,
      "grad_norm": 1.8423340320587158,
      "learning_rate": 5.994294736438219e-07,
      "loss": 1.2472,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2543
    },
    {
      "epoch": 1.5959849435382685,
      "grad_norm": 1.721067190170288,
      "learning_rate": 5.97652777387864e-07,
      "loss": 1.3293,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2544
    },
    {
      "epoch": 1.5966122961104139,
      "grad_norm": 1.8208516836166382,
      "learning_rate": 5.95878360558953e-07,
      "loss": 1.2584,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2545
    },
    {
      "epoch": 1.5972396486825597,
      "grad_norm": 1.639552354812622,
      "learning_rate": 5.941062252832283e-07,
      "loss": 1.1548,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2546
    },
    {
      "epoch": 1.5978670012547052,
      "grad_norm": 1.8412017822265625,
      "learning_rate": 5.923363736840946e-07,
      "loss": 1.2064,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2547
    },
    {
      "epoch": 1.5984943538268506,
      "grad_norm": 1.747814416885376,
      "learning_rate": 5.905688078822192e-07,
      "loss": 1.3132,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2548
    },
    {
      "epoch": 1.5991217063989962,
      "grad_norm": 1.5099161863327026,
      "learning_rate": 5.888035299955325e-07,
      "loss": 1.0845,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2549
    },
    {
      "epoch": 1.5997490589711418,
      "grad_norm": 1.4273351430892944,
      "learning_rate": 5.870405421392217e-07,
      "loss": 1.2287,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2550
    },
    {
      "epoch": 1.6003764115432872,
      "grad_norm": 1.7674329280853271,
      "learning_rate": 5.852798464257323e-07,
      "loss": 1.3752,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2551
    },
    {
      "epoch": 1.6010037641154329,
      "grad_norm": 1.6482888460159302,
      "learning_rate": 5.835214449647602e-07,
      "loss": 1.405,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2552
    },
    {
      "epoch": 1.6016311166875785,
      "grad_norm": 1.7106801271438599,
      "learning_rate": 5.817653398632559e-07,
      "loss": 1.1694,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2553
    },
    {
      "epoch": 1.602258469259724,
      "grad_norm": 1.8246630430221558,
      "learning_rate": 5.800115332254144e-07,
      "loss": 1.1745,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2554
    },
    {
      "epoch": 1.6028858218318696,
      "grad_norm": 1.8861138820648193,
      "learning_rate": 5.782600271526812e-07,
      "loss": 1.2363,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2555
    },
    {
      "epoch": 1.6035131744040152,
      "grad_norm": 1.6869688034057617,
      "learning_rate": 5.765108237437409e-07,
      "loss": 1.0971,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2556
    },
    {
      "epoch": 1.6041405269761606,
      "grad_norm": 1.5465879440307617,
      "learning_rate": 5.747639250945208e-07,
      "loss": 1.1556,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2557
    },
    {
      "epoch": 1.604767879548306,
      "grad_norm": 1.6890803575515747,
      "learning_rate": 5.730193332981879e-07,
      "loss": 1.1229,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2558
    },
    {
      "epoch": 1.6053952321204517,
      "grad_norm": 1.8298060894012451,
      "learning_rate": 5.712770504451426e-07,
      "loss": 1.2463,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2559
    },
    {
      "epoch": 1.6060225846925973,
      "grad_norm": 2.0945799350738525,
      "learning_rate": 5.695370786230208e-07,
      "loss": 1.458,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2560
    },
    {
      "epoch": 1.6066499372647427,
      "grad_norm": 1.846012830734253,
      "learning_rate": 5.677994199166875e-07,
      "loss": 1.3112,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2561
    },
    {
      "epoch": 1.6072772898368883,
      "grad_norm": 1.8249188661575317,
      "learning_rate": 5.660640764082381e-07,
      "loss": 1.1594,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2562
    },
    {
      "epoch": 1.607904642409034,
      "grad_norm": 1.952448844909668,
      "learning_rate": 5.643310501769924e-07,
      "loss": 1.3515,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2563
    },
    {
      "epoch": 1.6085319949811794,
      "grad_norm": 2.193516254425049,
      "learning_rate": 5.626003432994934e-07,
      "loss": 1.2284,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2564
    },
    {
      "epoch": 1.609159347553325,
      "grad_norm": 1.8074815273284912,
      "learning_rate": 5.608719578495059e-07,
      "loss": 1.1216,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2565
    },
    {
      "epoch": 1.6097867001254706,
      "grad_norm": 1.4073917865753174,
      "learning_rate": 5.591458958980123e-07,
      "loss": 1.1088,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2566
    },
    {
      "epoch": 1.610414052697616,
      "grad_norm": 1.7987900972366333,
      "learning_rate": 5.574221595132124e-07,
      "loss": 1.2133,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2567
    },
    {
      "epoch": 1.6110414052697615,
      "grad_norm": 1.6239173412322998,
      "learning_rate": 5.557007507605172e-07,
      "loss": 1.0926,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2568
    },
    {
      "epoch": 1.611668757841907,
      "grad_norm": 1.8376796245574951,
      "learning_rate": 5.539816717025515e-07,
      "loss": 1.2056,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2569
    },
    {
      "epoch": 1.6122961104140527,
      "grad_norm": 1.8480087518692017,
      "learning_rate": 5.522649243991463e-07,
      "loss": 1.251,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2570
    },
    {
      "epoch": 1.6129234629861982,
      "grad_norm": 1.8004862070083618,
      "learning_rate": 5.50550510907339e-07,
      "loss": 1.1975,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2571
    },
    {
      "epoch": 1.6135508155583438,
      "grad_norm": 1.6661462783813477,
      "learning_rate": 5.488384332813718e-07,
      "loss": 1.4225,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2572
    },
    {
      "epoch": 1.6141781681304894,
      "grad_norm": 1.7336554527282715,
      "learning_rate": 5.471286935726866e-07,
      "loss": 1.1692,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2573
    },
    {
      "epoch": 1.6148055207026348,
      "grad_norm": 1.893118977546692,
      "learning_rate": 5.454212938299256e-07,
      "loss": 1.1888,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2574
    },
    {
      "epoch": 1.6154328732747805,
      "grad_norm": 1.7622027397155762,
      "learning_rate": 5.437162360989251e-07,
      "loss": 1.298,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2575
    },
    {
      "epoch": 1.616060225846926,
      "grad_norm": 1.1501035690307617,
      "learning_rate": 5.420135224227174e-07,
      "loss": 1.2508,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2576
    },
    {
      "epoch": 1.6166875784190715,
      "grad_norm": 1.8274246454238892,
      "learning_rate": 5.403131548415244e-07,
      "loss": 1.2622,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2577
    },
    {
      "epoch": 1.617314930991217,
      "grad_norm": 1.746263027191162,
      "learning_rate": 5.386151353927569e-07,
      "loss": 1.3823,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2578
    },
    {
      "epoch": 1.6179422835633626,
      "grad_norm": 1.7620964050292969,
      "learning_rate": 5.369194661110138e-07,
      "loss": 1.2652,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2579
    },
    {
      "epoch": 1.6185696361355082,
      "grad_norm": 1.7409865856170654,
      "learning_rate": 5.352261490280767e-07,
      "loss": 1.1592,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2580
    },
    {
      "epoch": 1.6191969887076536,
      "grad_norm": 1.9257817268371582,
      "learning_rate": 5.335351861729082e-07,
      "loss": 1.2101,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2581
    },
    {
      "epoch": 1.6198243412797992,
      "grad_norm": 2.0328640937805176,
      "learning_rate": 5.318465795716507e-07,
      "loss": 1.2135,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2582
    },
    {
      "epoch": 1.6204516938519449,
      "grad_norm": 1.7458486557006836,
      "learning_rate": 5.301603312476242e-07,
      "loss": 1.1808,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2583
    },
    {
      "epoch": 1.6210790464240903,
      "grad_norm": 1.7722233533859253,
      "learning_rate": 5.284764432213222e-07,
      "loss": 1.2224,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2584
    },
    {
      "epoch": 1.621706398996236,
      "grad_norm": 1.5155515670776367,
      "learning_rate": 5.267949175104084e-07,
      "loss": 1.2597,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2585
    },
    {
      "epoch": 1.6223337515683816,
      "grad_norm": 1.7921104431152344,
      "learning_rate": 5.251157561297193e-07,
      "loss": 1.1613,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2586
    },
    {
      "epoch": 1.622961104140527,
      "grad_norm": 1.710188388824463,
      "learning_rate": 5.234389610912552e-07,
      "loss": 1.2128,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2587
    },
    {
      "epoch": 1.6235884567126724,
      "grad_norm": 1.8385692834854126,
      "learning_rate": 5.217645344041838e-07,
      "loss": 1.3291,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2588
    },
    {
      "epoch": 1.624215809284818,
      "grad_norm": 2.3026092052459717,
      "learning_rate": 5.200924780748323e-07,
      "loss": 1.1366,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2589
    },
    {
      "epoch": 1.6248431618569636,
      "grad_norm": 1.6695224046707153,
      "learning_rate": 5.184227941066902e-07,
      "loss": 1.198,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2590
    },
    {
      "epoch": 1.625470514429109,
      "grad_norm": 1.778435230255127,
      "learning_rate": 5.167554845004028e-07,
      "loss": 1.2428,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2591
    },
    {
      "epoch": 1.6260978670012547,
      "grad_norm": 1.6701916456222534,
      "learning_rate": 5.150905512537699e-07,
      "loss": 1.1933,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2592
    },
    {
      "epoch": 1.6267252195734003,
      "grad_norm": 1.9535346031188965,
      "learning_rate": 5.134279963617461e-07,
      "loss": 1.2006,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2593
    },
    {
      "epoch": 1.6273525721455457,
      "grad_norm": 1.6672227382659912,
      "learning_rate": 5.117678218164337e-07,
      "loss": 1.2237,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2594
    },
    {
      "epoch": 1.6279799247176914,
      "grad_norm": 1.7902626991271973,
      "learning_rate": 5.101100296070852e-07,
      "loss": 1.3211,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2595
    },
    {
      "epoch": 1.628607277289837,
      "grad_norm": 1.9515424966812134,
      "learning_rate": 5.084546217200972e-07,
      "loss": 1.2774,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2596
    },
    {
      "epoch": 1.6292346298619824,
      "grad_norm": 1.8537187576293945,
      "learning_rate": 5.068016001390088e-07,
      "loss": 1.2868,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2597
    },
    {
      "epoch": 1.6298619824341278,
      "grad_norm": 1.8684711456298828,
      "learning_rate": 5.051509668445012e-07,
      "loss": 1.1947,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2598
    },
    {
      "epoch": 1.6304893350062737,
      "grad_norm": 1.3285757303237915,
      "learning_rate": 5.035027238143925e-07,
      "loss": 1.2076,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2599
    },
    {
      "epoch": 1.631116687578419,
      "grad_norm": 1.7497682571411133,
      "learning_rate": 5.018568730236384e-07,
      "loss": 1.1428,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2600
    },
    {
      "epoch": 1.6317440401505645,
      "grad_norm": 2.2294180393218994,
      "learning_rate": 5.002134164443262e-07,
      "loss": 1.3661,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2601
    },
    {
      "epoch": 1.6323713927227101,
      "grad_norm": 1.6695797443389893,
      "learning_rate": 4.985723560456765e-07,
      "loss": 1.3749,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2602
    },
    {
      "epoch": 1.6329987452948558,
      "grad_norm": 1.7357219457626343,
      "learning_rate": 4.969336937940378e-07,
      "loss": 1.3069,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2603
    },
    {
      "epoch": 1.6336260978670012,
      "grad_norm": 4.611532211303711,
      "learning_rate": 4.952974316528833e-07,
      "loss": 1.3523,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2604
    },
    {
      "epoch": 1.6342534504391468,
      "grad_norm": 1.919092059135437,
      "learning_rate": 4.936635715828145e-07,
      "loss": 1.1854,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2605
    },
    {
      "epoch": 1.6348808030112925,
      "grad_norm": 1.867214560508728,
      "learning_rate": 4.9203211554155e-07,
      "loss": 1.262,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2606
    },
    {
      "epoch": 1.6355081555834379,
      "grad_norm": 1.6036155223846436,
      "learning_rate": 4.904030654839317e-07,
      "loss": 1.2831,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2607
    },
    {
      "epoch": 1.6361355081555833,
      "grad_norm": 1.7611279487609863,
      "learning_rate": 4.887764233619163e-07,
      "loss": 1.2201,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2608
    },
    {
      "epoch": 1.6367628607277291,
      "grad_norm": 1.7776987552642822,
      "learning_rate": 4.871521911245764e-07,
      "loss": 1.3487,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2609
    },
    {
      "epoch": 1.6373902132998746,
      "grad_norm": 1.8058147430419922,
      "learning_rate": 4.855303707180964e-07,
      "loss": 1.2534,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2610
    },
    {
      "epoch": 1.63801756587202,
      "grad_norm": 1.6661456823349,
      "learning_rate": 4.839109640857706e-07,
      "loss": 1.1765,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2611
    },
    {
      "epoch": 1.6386449184441656,
      "grad_norm": 1.8750739097595215,
      "learning_rate": 4.822939731680021e-07,
      "loss": 1.2644,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2612
    },
    {
      "epoch": 1.6392722710163112,
      "grad_norm": 1.8761341571807861,
      "learning_rate": 4.806793999022988e-07,
      "loss": 1.2247,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2613
    },
    {
      "epoch": 1.6398996235884566,
      "grad_norm": 1.7048670053482056,
      "learning_rate": 4.790672462232715e-07,
      "loss": 1.1726,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2614
    },
    {
      "epoch": 1.6405269761606023,
      "grad_norm": 1.754637360572815,
      "learning_rate": 4.774575140626317e-07,
      "loss": 1.1735,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2615
    },
    {
      "epoch": 1.641154328732748,
      "grad_norm": 1.8188670873641968,
      "learning_rate": 4.758502053491906e-07,
      "loss": 1.2341,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2616
    },
    {
      "epoch": 1.6417816813048933,
      "grad_norm": 1.8739845752716064,
      "learning_rate": 4.7424532200885417e-07,
      "loss": 1.0668,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2617
    },
    {
      "epoch": 1.6424090338770387,
      "grad_norm": 1.9493358135223389,
      "learning_rate": 4.726428659646226e-07,
      "loss": 1.3107,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2618
    },
    {
      "epoch": 1.6430363864491846,
      "grad_norm": 1.6212058067321777,
      "learning_rate": 4.710428391365887e-07,
      "loss": 1.2477,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2619
    },
    {
      "epoch": 1.64366373902133,
      "grad_norm": 1.7489856481552124,
      "learning_rate": 4.6944524344193247e-07,
      "loss": 1.1277,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2620
    },
    {
      "epoch": 1.6442910915934754,
      "grad_norm": 1.7265174388885498,
      "learning_rate": 4.6785008079492377e-07,
      "loss": 1.3101,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2621
    },
    {
      "epoch": 1.644918444165621,
      "grad_norm": 1.923108696937561,
      "learning_rate": 4.6625735310691396e-07,
      "loss": 1.2839,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2622
    },
    {
      "epoch": 1.6455457967377667,
      "grad_norm": 1.7697632312774658,
      "learning_rate": 4.6466706228633973e-07,
      "loss": 1.2111,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2623
    },
    {
      "epoch": 1.646173149309912,
      "grad_norm": 1.936873197555542,
      "learning_rate": 4.630792102387155e-07,
      "loss": 1.2291,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2624
    },
    {
      "epoch": 1.6468005018820577,
      "grad_norm": 1.563767433166504,
      "learning_rate": 4.614937988666343e-07,
      "loss": 1.22,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2625
    },
    {
      "epoch": 1.6474278544542034,
      "grad_norm": 1.9267933368682861,
      "learning_rate": 4.5991083006976607e-07,
      "loss": 1.296,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2626
    },
    {
      "epoch": 1.6480552070263488,
      "grad_norm": 1.8290936946868896,
      "learning_rate": 4.5833030574485097e-07,
      "loss": 1.1191,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2627
    },
    {
      "epoch": 1.6486825595984942,
      "grad_norm": 1.7071551084518433,
      "learning_rate": 4.567522277857042e-07,
      "loss": 1.2689,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2628
    },
    {
      "epoch": 1.64930991217064,
      "grad_norm": 1.4760491847991943,
      "learning_rate": 4.55176598083206e-07,
      "loss": 1.1789,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2629
    },
    {
      "epoch": 1.6499372647427855,
      "grad_norm": 1.6776413917541504,
      "learning_rate": 4.536034185253052e-07,
      "loss": 1.2469,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2630
    },
    {
      "epoch": 1.6505646173149309,
      "grad_norm": 1.4032368659973145,
      "learning_rate": 4.5203269099701296e-07,
      "loss": 1.1217,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2631
    },
    {
      "epoch": 1.6511919698870765,
      "grad_norm": 1.7307270765304565,
      "learning_rate": 4.504644173804054e-07,
      "loss": 1.2958,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2632
    },
    {
      "epoch": 1.6518193224592221,
      "grad_norm": 1.848423719406128,
      "learning_rate": 4.488985995546152e-07,
      "loss": 1.2684,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2633
    },
    {
      "epoch": 1.6524466750313676,
      "grad_norm": 1.5907129049301147,
      "learning_rate": 4.473352393958338e-07,
      "loss": 1.1161,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2634
    },
    {
      "epoch": 1.6530740276035132,
      "grad_norm": 1.4280104637145996,
      "learning_rate": 4.4577433877730865e-07,
      "loss": 1.0568,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2635
    },
    {
      "epoch": 1.6537013801756588,
      "grad_norm": 1.9540001153945923,
      "learning_rate": 4.4421589956933827e-07,
      "loss": 1.3103,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2636
    },
    {
      "epoch": 1.6543287327478042,
      "grad_norm": 1.7712022066116333,
      "learning_rate": 4.4265992363927367e-07,
      "loss": 1.261,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2637
    },
    {
      "epoch": 1.6549560853199499,
      "grad_norm": 2.0369656085968018,
      "learning_rate": 4.411064128515133e-07,
      "loss": 1.3026,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2638
    },
    {
      "epoch": 1.6555834378920955,
      "grad_norm": 1.9376763105392456,
      "learning_rate": 4.395553690675014e-07,
      "loss": 1.4054,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2639
    },
    {
      "epoch": 1.656210790464241,
      "grad_norm": 1.9523491859436035,
      "learning_rate": 4.380067941457278e-07,
      "loss": 1.155,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2640
    },
    {
      "epoch": 1.6568381430363863,
      "grad_norm": 1.5371203422546387,
      "learning_rate": 4.3646068994172233e-07,
      "loss": 1.1868,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2641
    },
    {
      "epoch": 1.657465495608532,
      "grad_norm": 1.8343383073806763,
      "learning_rate": 4.349170583080559e-07,
      "loss": 1.2096,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2642
    },
    {
      "epoch": 1.6580928481806776,
      "grad_norm": 1.9491466283798218,
      "learning_rate": 4.3337590109433505e-07,
      "loss": 1.1406,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2643
    },
    {
      "epoch": 1.658720200752823,
      "grad_norm": 1.6798421144485474,
      "learning_rate": 4.3183722014720375e-07,
      "loss": 1.254,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2644
    },
    {
      "epoch": 1.6593475533249686,
      "grad_norm": 1.7423664331436157,
      "learning_rate": 4.303010173103367e-07,
      "loss": 1.0916,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2645
    },
    {
      "epoch": 1.6599749058971143,
      "grad_norm": 1.8117595911026,
      "learning_rate": 4.287672944244403e-07,
      "loss": 1.2391,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2646
    },
    {
      "epoch": 1.6606022584692597,
      "grad_norm": 1.9419714212417603,
      "learning_rate": 4.272360533272482e-07,
      "loss": 1.2536,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2647
    },
    {
      "epoch": 1.6612296110414053,
      "grad_norm": 1.3489598035812378,
      "learning_rate": 4.25707295853523e-07,
      "loss": 1.2114,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2648
    },
    {
      "epoch": 1.661856963613551,
      "grad_norm": 1.895057201385498,
      "learning_rate": 4.2418102383504884e-07,
      "loss": 1.3033,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2649
    },
    {
      "epoch": 1.6624843161856964,
      "grad_norm": 1.6380215883255005,
      "learning_rate": 4.22657239100632e-07,
      "loss": 1.0711,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2650
    },
    {
      "epoch": 1.6631116687578418,
      "grad_norm": 1.456254243850708,
      "learning_rate": 4.211359434761003e-07,
      "loss": 1.1179,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2651
    },
    {
      "epoch": 1.6637390213299874,
      "grad_norm": 1.8337856531143188,
      "learning_rate": 4.196171387842976e-07,
      "loss": 1.2768,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2652
    },
    {
      "epoch": 1.664366373902133,
      "grad_norm": 1.382093906402588,
      "learning_rate": 4.18100826845082e-07,
      "loss": 1.1812,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2653
    },
    {
      "epoch": 1.6649937264742785,
      "grad_norm": 1.6331734657287598,
      "learning_rate": 4.16587009475328e-07,
      "loss": 1.1856,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2654
    },
    {
      "epoch": 1.665621079046424,
      "grad_norm": 2.566094160079956,
      "learning_rate": 4.1507568848891706e-07,
      "loss": 1.1711,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2655
    },
    {
      "epoch": 1.6662484316185697,
      "grad_norm": 1.8740427494049072,
      "learning_rate": 4.1356686569674344e-07,
      "loss": 1.2985,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2656
    },
    {
      "epoch": 1.6668757841907151,
      "grad_norm": 1.7602894306182861,
      "learning_rate": 4.1206054290670537e-07,
      "loss": 1.2358,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2657
    },
    {
      "epoch": 1.6675031367628608,
      "grad_norm": 1.8847901821136475,
      "learning_rate": 4.10556721923705e-07,
      "loss": 1.3158,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2658
    },
    {
      "epoch": 1.6681304893350064,
      "grad_norm": 1.524640440940857,
      "learning_rate": 4.0905540454965004e-07,
      "loss": 1.1987,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2659
    },
    {
      "epoch": 1.6687578419071518,
      "grad_norm": 1.7790837287902832,
      "learning_rate": 4.0755659258344434e-07,
      "loss": 1.1568,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2660
    },
    {
      "epoch": 1.6693851944792972,
      "grad_norm": 1.7558435201644897,
      "learning_rate": 4.060602878209932e-07,
      "loss": 1.1112,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2661
    },
    {
      "epoch": 1.6700125470514429,
      "grad_norm": 1.9152045249938965,
      "learning_rate": 4.045664920551959e-07,
      "loss": 1.2686,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2662
    },
    {
      "epoch": 1.6706398996235885,
      "grad_norm": 1.7652149200439453,
      "learning_rate": 4.0307520707594446e-07,
      "loss": 1.3869,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2663
    },
    {
      "epoch": 1.671267252195734,
      "grad_norm": 1.9139306545257568,
      "learning_rate": 4.015864346701251e-07,
      "loss": 1.3363,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2664
    },
    {
      "epoch": 1.6718946047678795,
      "grad_norm": 2.0330417156219482,
      "learning_rate": 4.001001766216117e-07,
      "loss": 1.2961,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2665
    },
    {
      "epoch": 1.6725219573400252,
      "grad_norm": 2.176459789276123,
      "learning_rate": 3.986164347112656e-07,
      "loss": 1.3539,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2666
    },
    {
      "epoch": 1.6731493099121706,
      "grad_norm": 1.9700795412063599,
      "learning_rate": 3.971352107169324e-07,
      "loss": 1.2131,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2667
    },
    {
      "epoch": 1.6737766624843162,
      "grad_norm": 1.8541291952133179,
      "learning_rate": 3.956565064134432e-07,
      "loss": 1.2052,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2668
    },
    {
      "epoch": 1.6744040150564619,
      "grad_norm": 1.6921569108963013,
      "learning_rate": 3.941803235726069e-07,
      "loss": 1.0926,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2669
    },
    {
      "epoch": 1.6750313676286073,
      "grad_norm": 2.4293434619903564,
      "learning_rate": 3.9270666396321404e-07,
      "loss": 1.2696,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2670
    },
    {
      "epoch": 1.6756587202007527,
      "grad_norm": 1.8745760917663574,
      "learning_rate": 3.9123552935102976e-07,
      "loss": 1.2597,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2671
    },
    {
      "epoch": 1.6762860727728983,
      "grad_norm": 1.8291773796081543,
      "learning_rate": 3.897669214987937e-07,
      "loss": 1.4168,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2672
    },
    {
      "epoch": 1.676913425345044,
      "grad_norm": 1.7694315910339355,
      "learning_rate": 3.883008421662196e-07,
      "loss": 1.2302,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2673
    },
    {
      "epoch": 1.6775407779171894,
      "grad_norm": 1.8569852113723755,
      "learning_rate": 3.8683729310998926e-07,
      "loss": 1.1905,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2674
    },
    {
      "epoch": 1.678168130489335,
      "grad_norm": 1.8917303085327148,
      "learning_rate": 3.8537627608375485e-07,
      "loss": 1.2325,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2675
    },
    {
      "epoch": 1.6787954830614806,
      "grad_norm": 1.6169365644454956,
      "learning_rate": 3.839177928381327e-07,
      "loss": 1.4392,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2676
    },
    {
      "epoch": 1.679422835633626,
      "grad_norm": 1.8867363929748535,
      "learning_rate": 3.8246184512070486e-07,
      "loss": 1.1633,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2677
    },
    {
      "epoch": 1.6800501882057717,
      "grad_norm": 2.2046451568603516,
      "learning_rate": 3.81008434676014e-07,
      "loss": 1.1018,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2678
    },
    {
      "epoch": 1.6806775407779173,
      "grad_norm": 1.8099101781845093,
      "learning_rate": 3.79557563245562e-07,
      "loss": 1.3516,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2679
    },
    {
      "epoch": 1.6813048933500627,
      "grad_norm": 1.7646654844284058,
      "learning_rate": 3.78109232567811e-07,
      "loss": 1.2145,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2680
    },
    {
      "epoch": 1.6819322459222081,
      "grad_norm": 1.616289734840393,
      "learning_rate": 3.766634443781761e-07,
      "loss": 1.2265,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2681
    },
    {
      "epoch": 1.682559598494354,
      "grad_norm": 1.3728324174880981,
      "learning_rate": 3.752202004090275e-07,
      "loss": 1.2257,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2682
    },
    {
      "epoch": 1.6831869510664994,
      "grad_norm": 1.857109785079956,
      "learning_rate": 3.7377950238968566e-07,
      "loss": 1.2864,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2683
    },
    {
      "epoch": 1.6838143036386448,
      "grad_norm": 1.9126770496368408,
      "learning_rate": 3.7234135204642195e-07,
      "loss": 1.2166,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2684
    },
    {
      "epoch": 1.6844416562107905,
      "grad_norm": 2.712088108062744,
      "learning_rate": 3.709057511024541e-07,
      "loss": 1.2432,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2685
    },
    {
      "epoch": 1.685069008782936,
      "grad_norm": 1.7454075813293457,
      "learning_rate": 3.6947270127794475e-07,
      "loss": 1.2589,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2686
    },
    {
      "epoch": 1.6856963613550815,
      "grad_norm": 1.8178859949111938,
      "learning_rate": 3.6804220429000116e-07,
      "loss": 1.2416,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2687
    },
    {
      "epoch": 1.6863237139272271,
      "grad_norm": 1.7671403884887695,
      "learning_rate": 3.666142618526697e-07,
      "loss": 1.4289,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2688
    },
    {
      "epoch": 1.6869510664993728,
      "grad_norm": 1.6030900478363037,
      "learning_rate": 3.651888756769381e-07,
      "loss": 1.3057,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2689
    },
    {
      "epoch": 1.6875784190715182,
      "grad_norm": 1.710805058479309,
      "learning_rate": 3.637660474707291e-07,
      "loss": 1.2315,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2690
    },
    {
      "epoch": 1.6882057716436636,
      "grad_norm": 1.9641602039337158,
      "learning_rate": 3.6234577893890235e-07,
      "loss": 1.1884,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2691
    },
    {
      "epoch": 1.6888331242158094,
      "grad_norm": 1.7865450382232666,
      "learning_rate": 3.609280717832489e-07,
      "loss": 1.2465,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2692
    },
    {
      "epoch": 1.6894604767879549,
      "grad_norm": 1.943463683128357,
      "learning_rate": 3.595129277024906e-07,
      "loss": 1.2143,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2693
    },
    {
      "epoch": 1.6900878293601003,
      "grad_norm": 1.588843822479248,
      "learning_rate": 3.581003483922801e-07,
      "loss": 1.2295,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2694
    },
    {
      "epoch": 1.690715181932246,
      "grad_norm": 1.748705506324768,
      "learning_rate": 3.5669033554519424e-07,
      "loss": 1.2691,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2695
    },
    {
      "epoch": 1.6913425345043915,
      "grad_norm": 1.7085033655166626,
      "learning_rate": 3.552828908507375e-07,
      "loss": 1.2281,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2696
    },
    {
      "epoch": 1.691969887076537,
      "grad_norm": 1.6979553699493408,
      "learning_rate": 3.538780159953348e-07,
      "loss": 1.2615,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2697
    },
    {
      "epoch": 1.6925972396486826,
      "grad_norm": 1.8602803945541382,
      "learning_rate": 3.5247571266233323e-07,
      "loss": 1.3764,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2698
    },
    {
      "epoch": 1.6932245922208282,
      "grad_norm": 2.2552378177642822,
      "learning_rate": 3.510759825319976e-07,
      "loss": 1.2772,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2699
    },
    {
      "epoch": 1.6938519447929736,
      "grad_norm": 1.6664947271347046,
      "learning_rate": 3.4967882728150994e-07,
      "loss": 1.1499,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2700
    },
    {
      "epoch": 1.694479297365119,
      "grad_norm": 1.85127592086792,
      "learning_rate": 3.48284248584968e-07,
      "loss": 1.2625,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2701
    },
    {
      "epoch": 1.695106649937265,
      "grad_norm": 1.7261724472045898,
      "learning_rate": 3.468922481133802e-07,
      "loss": 1.2575,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2702
    },
    {
      "epoch": 1.6957340025094103,
      "grad_norm": 1.5232841968536377,
      "learning_rate": 3.455028275346678e-07,
      "loss": 1.2934,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2703
    },
    {
      "epoch": 1.6963613550815557,
      "grad_norm": 1.9840103387832642,
      "learning_rate": 3.4411598851365967e-07,
      "loss": 1.3476,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2704
    },
    {
      "epoch": 1.6969887076537014,
      "grad_norm": 1.9216537475585938,
      "learning_rate": 3.4273173271209047e-07,
      "loss": 1.1801,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2705
    },
    {
      "epoch": 1.697616060225847,
      "grad_norm": 1.321739912033081,
      "learning_rate": 3.413500617886023e-07,
      "loss": 1.2259,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2706
    },
    {
      "epoch": 1.6982434127979924,
      "grad_norm": 1.701654314994812,
      "learning_rate": 3.39970977398737e-07,
      "loss": 1.2265,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2707
    },
    {
      "epoch": 1.698870765370138,
      "grad_norm": 1.864776611328125,
      "learning_rate": 3.3859448119493957e-07,
      "loss": 1.4043,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2708
    },
    {
      "epoch": 1.6994981179422837,
      "grad_norm": 1.7704063653945923,
      "learning_rate": 3.3722057482655224e-07,
      "loss": 1.2881,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2709
    },
    {
      "epoch": 1.700125470514429,
      "grad_norm": 1.7865586280822754,
      "learning_rate": 3.3584925993981497e-07,
      "loss": 1.2118,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2710
    },
    {
      "epoch": 1.7007528230865745,
      "grad_norm": 2.1016578674316406,
      "learning_rate": 3.344805381778621e-07,
      "loss": 1.1508,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2711
    },
    {
      "epoch": 1.7013801756587204,
      "grad_norm": 1.7841296195983887,
      "learning_rate": 3.331144111807205e-07,
      "loss": 1.1504,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2712
    },
    {
      "epoch": 1.7020075282308658,
      "grad_norm": 1.661086916923523,
      "learning_rate": 3.3175088058530925e-07,
      "loss": 1.1409,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2713
    },
    {
      "epoch": 1.7026348808030112,
      "grad_norm": 1.675002098083496,
      "learning_rate": 3.3038994802543466e-07,
      "loss": 1.1987,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2714
    },
    {
      "epoch": 1.7032622333751568,
      "grad_norm": 1.4157485961914062,
      "learning_rate": 3.2903161513179156e-07,
      "loss": 1.177,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2715
    },
    {
      "epoch": 1.7038895859473024,
      "grad_norm": 1.4867208003997803,
      "learning_rate": 3.276758835319582e-07,
      "loss": 1.2039,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2716
    },
    {
      "epoch": 1.7045169385194479,
      "grad_norm": 1.752272367477417,
      "learning_rate": 3.2632275485039806e-07,
      "loss": 1.326,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2717
    },
    {
      "epoch": 1.7051442910915935,
      "grad_norm": 1.7673453092575073,
      "learning_rate": 3.2497223070845377e-07,
      "loss": 1.2458,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2718
    },
    {
      "epoch": 1.7057716436637391,
      "grad_norm": 1.7563871145248413,
      "learning_rate": 3.236243127243477e-07,
      "loss": 1.2054,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2719
    },
    {
      "epoch": 1.7063989962358845,
      "grad_norm": 1.822948932647705,
      "learning_rate": 3.2227900251318055e-07,
      "loss": 1.1937,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2720
    },
    {
      "epoch": 1.70702634880803,
      "grad_norm": 1.6502693891525269,
      "learning_rate": 3.2093630168692656e-07,
      "loss": 1.2332,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2721
    },
    {
      "epoch": 1.7076537013801758,
      "grad_norm": 1.5615298748016357,
      "learning_rate": 3.1959621185443516e-07,
      "loss": 1.189,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2722
    },
    {
      "epoch": 1.7082810539523212,
      "grad_norm": 1.8436845541000366,
      "learning_rate": 3.182587346214255e-07,
      "loss": 1.2043,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2723
    },
    {
      "epoch": 1.7089084065244666,
      "grad_norm": 2.5546927452087402,
      "learning_rate": 3.169238715904882e-07,
      "loss": 1.1197,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2724
    },
    {
      "epoch": 1.7095357590966123,
      "grad_norm": 1.7388148307800293,
      "learning_rate": 3.1559162436107987e-07,
      "loss": 1.2678,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2725
    },
    {
      "epoch": 1.710163111668758,
      "grad_norm": 1.6913169622421265,
      "learning_rate": 3.142619945295225e-07,
      "loss": 1.1965,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2726
    },
    {
      "epoch": 1.7107904642409033,
      "grad_norm": 1.6821208000183105,
      "learning_rate": 3.1293498368900414e-07,
      "loss": 1.2556,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2727
    },
    {
      "epoch": 1.711417816813049,
      "grad_norm": 2.5397050380706787,
      "learning_rate": 3.116105934295724e-07,
      "loss": 1.098,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2728
    },
    {
      "epoch": 1.7120451693851946,
      "grad_norm": 1.8930282592773438,
      "learning_rate": 3.1028882533813643e-07,
      "loss": 1.1857,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2729
    },
    {
      "epoch": 1.71267252195734,
      "grad_norm": 1.8789072036743164,
      "learning_rate": 3.089696809984624e-07,
      "loss": 1.311,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2730
    },
    {
      "epoch": 1.7132998745294856,
      "grad_norm": 2.0739526748657227,
      "learning_rate": 3.076531619911735e-07,
      "loss": 1.1883,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2731
    },
    {
      "epoch": 1.7139272271016313,
      "grad_norm": 1.430984616279602,
      "learning_rate": 3.063392698937462e-07,
      "loss": 1.2454,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2732
    },
    {
      "epoch": 1.7145545796737767,
      "grad_norm": 1.6760432720184326,
      "learning_rate": 3.0502800628050946e-07,
      "loss": 1.2448,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2733
    },
    {
      "epoch": 1.715181932245922,
      "grad_norm": 1.83548104763031,
      "learning_rate": 3.0371937272264454e-07,
      "loss": 1.2657,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2734
    },
    {
      "epoch": 1.7158092848180677,
      "grad_norm": 2.0283336639404297,
      "learning_rate": 3.024133707881785e-07,
      "loss": 1.272,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2735
    },
    {
      "epoch": 1.7164366373902133,
      "grad_norm": 1.5249745845794678,
      "learning_rate": 3.0111000204198825e-07,
      "loss": 1.205,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2736
    },
    {
      "epoch": 1.7170639899623588,
      "grad_norm": 1.977151870727539,
      "learning_rate": 2.998092680457923e-07,
      "loss": 1.286,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2737
    },
    {
      "epoch": 1.7176913425345044,
      "grad_norm": 1.2337921857833862,
      "learning_rate": 2.98511170358155e-07,
      "loss": 1.1496,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2738
    },
    {
      "epoch": 1.71831869510665,
      "grad_norm": 2.6582040786743164,
      "learning_rate": 2.9721571053448056e-07,
      "loss": 1.1556,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2739
    },
    {
      "epoch": 1.7189460476787954,
      "grad_norm": 1.7614023685455322,
      "learning_rate": 2.9592289012701113e-07,
      "loss": 1.1573,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2740
    },
    {
      "epoch": 1.719573400250941,
      "grad_norm": 1.8160935640335083,
      "learning_rate": 2.9463271068482955e-07,
      "loss": 1.1242,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2741
    },
    {
      "epoch": 1.7202007528230867,
      "grad_norm": 1.9184051752090454,
      "learning_rate": 2.933451737538506e-07,
      "loss": 1.2449,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2742
    },
    {
      "epoch": 1.7208281053952321,
      "grad_norm": 1.8110864162445068,
      "learning_rate": 2.920602808768258e-07,
      "loss": 1.2209,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2743
    },
    {
      "epoch": 1.7214554579673775,
      "grad_norm": 1.450143575668335,
      "learning_rate": 2.9077803359333607e-07,
      "loss": 1.1317,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2744
    },
    {
      "epoch": 1.7220828105395232,
      "grad_norm": 1.7388874292373657,
      "learning_rate": 2.8949843343979435e-07,
      "loss": 1.1769,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2745
    },
    {
      "epoch": 1.7227101631116688,
      "grad_norm": 1.4511213302612305,
      "learning_rate": 2.8822148194944055e-07,
      "loss": 1.1455,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2746
    },
    {
      "epoch": 1.7233375156838142,
      "grad_norm": 1.4643691778182983,
      "learning_rate": 2.8694718065234105e-07,
      "loss": 1.2187,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2747
    },
    {
      "epoch": 1.7239648682559598,
      "grad_norm": 1.8728381395339966,
      "learning_rate": 2.856755310753867e-07,
      "loss": 1.2591,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2748
    },
    {
      "epoch": 1.7245922208281055,
      "grad_norm": 1.712773084640503,
      "learning_rate": 2.8440653474229083e-07,
      "loss": 1.315,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2749
    },
    {
      "epoch": 1.725219573400251,
      "grad_norm": 1.9405332803726196,
      "learning_rate": 2.831401931735892e-07,
      "loss": 1.1987,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2750
    },
    {
      "epoch": 1.7258469259723965,
      "grad_norm": 1.7674938440322876,
      "learning_rate": 2.8187650788663366e-07,
      "loss": 1.3141,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2751
    },
    {
      "epoch": 1.7264742785445422,
      "grad_norm": 1.6644740104675293,
      "learning_rate": 2.8061548039559625e-07,
      "loss": 1.1981,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2752
    },
    {
      "epoch": 1.7271016311166876,
      "grad_norm": 2.2521865367889404,
      "learning_rate": 2.793571122114627e-07,
      "loss": 1.2635,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2753
    },
    {
      "epoch": 1.727728983688833,
      "grad_norm": 1.648939609527588,
      "learning_rate": 2.781014048420319e-07,
      "loss": 1.1418,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2754
    },
    {
      "epoch": 1.7283563362609786,
      "grad_norm": 1.3256694078445435,
      "learning_rate": 2.7684835979191664e-07,
      "loss": 1.0624,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2755
    },
    {
      "epoch": 1.7289836888331243,
      "grad_norm": 1.8163894414901733,
      "learning_rate": 2.755979785625368e-07,
      "loss": 1.1717,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2756
    },
    {
      "epoch": 1.7296110414052697,
      "grad_norm": 1.7268973588943481,
      "learning_rate": 2.7435026265212327e-07,
      "loss": 1.2758,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2757
    },
    {
      "epoch": 1.7302383939774153,
      "grad_norm": 1.8375306129455566,
      "learning_rate": 2.731052135557111e-07,
      "loss": 1.2875,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2758
    },
    {
      "epoch": 1.730865746549561,
      "grad_norm": 1.4706637859344482,
      "learning_rate": 2.718628327651407e-07,
      "loss": 1.315,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2759
    },
    {
      "epoch": 1.7314930991217063,
      "grad_norm": 1.4504117965698242,
      "learning_rate": 2.7062312176905606e-07,
      "loss": 1.1671,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2760
    },
    {
      "epoch": 1.732120451693852,
      "grad_norm": 2.374910593032837,
      "learning_rate": 2.693860820529004e-07,
      "loss": 1.167,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2761
    },
    {
      "epoch": 1.7327478042659976,
      "grad_norm": 1.5370956659317017,
      "learning_rate": 2.681517150989185e-07,
      "loss": 1.3897,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2762
    },
    {
      "epoch": 1.733375156838143,
      "grad_norm": 1.623807668685913,
      "learning_rate": 2.669200223861504e-07,
      "loss": 1.0385,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2763
    },
    {
      "epoch": 1.7340025094102884,
      "grad_norm": 1.8441228866577148,
      "learning_rate": 2.6569100539043326e-07,
      "loss": 1.268,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2764
    },
    {
      "epoch": 1.734629861982434,
      "grad_norm": 1.8143815994262695,
      "learning_rate": 2.6446466558439657e-07,
      "loss": 1.2536,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2765
    },
    {
      "epoch": 1.7352572145545797,
      "grad_norm": 1.8156932592391968,
      "learning_rate": 2.632410044374642e-07,
      "loss": 1.2525,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2766
    },
    {
      "epoch": 1.7358845671267251,
      "grad_norm": 1.8268598318099976,
      "learning_rate": 2.6202002341584875e-07,
      "loss": 1.3156,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2767
    },
    {
      "epoch": 1.7365119196988708,
      "grad_norm": 1.4789742231369019,
      "learning_rate": 2.608017239825511e-07,
      "loss": 1.1006,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2768
    },
    {
      "epoch": 1.7371392722710164,
      "grad_norm": 1.5572457313537598,
      "learning_rate": 2.5958610759736133e-07,
      "loss": 1.2384,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2769
    },
    {
      "epoch": 1.7377666248431618,
      "grad_norm": 1.9117459058761597,
      "learning_rate": 2.5837317571685175e-07,
      "loss": 1.2199,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2770
    },
    {
      "epoch": 1.7383939774153074,
      "grad_norm": 1.9667168855667114,
      "learning_rate": 2.5716292979438037e-07,
      "loss": 1.2503,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2771
    },
    {
      "epoch": 1.739021329987453,
      "grad_norm": 2.239107370376587,
      "learning_rate": 2.559553712800852e-07,
      "loss": 1.203,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2772
    },
    {
      "epoch": 1.7396486825595985,
      "grad_norm": 2.7901501655578613,
      "learning_rate": 2.547505016208851e-07,
      "loss": 1.2129,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2773
    },
    {
      "epoch": 1.740276035131744,
      "grad_norm": 1.6809381246566772,
      "learning_rate": 2.53548322260477e-07,
      "loss": 1.218,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2774
    },
    {
      "epoch": 1.7409033877038897,
      "grad_norm": 1.3651955127716064,
      "learning_rate": 2.5234883463933323e-07,
      "loss": 1.1659,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2775
    },
    {
      "epoch": 1.7415307402760352,
      "grad_norm": 1.5471340417861938,
      "learning_rate": 2.511520401947032e-07,
      "loss": 1.2559,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2776
    },
    {
      "epoch": 1.7421580928481806,
      "grad_norm": 1.870573878288269,
      "learning_rate": 2.4995794036060616e-07,
      "loss": 1.2515,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2777
    },
    {
      "epoch": 1.7427854454203262,
      "grad_norm": 1.6339471340179443,
      "learning_rate": 2.4876653656783543e-07,
      "loss": 1.2609,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2778
    },
    {
      "epoch": 1.7434127979924718,
      "grad_norm": 1.815114974975586,
      "learning_rate": 2.4757783024395244e-07,
      "loss": 1.3496,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2779
    },
    {
      "epoch": 1.7440401505646173,
      "grad_norm": 1.5389330387115479,
      "learning_rate": 2.463918228132864e-07,
      "loss": 1.1197,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2780
    },
    {
      "epoch": 1.7446675031367629,
      "grad_norm": 1.1387971639633179,
      "learning_rate": 2.4520851569693275e-07,
      "loss": 1.1247,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2781
    },
    {
      "epoch": 1.7452948557089085,
      "grad_norm": 2.002338171005249,
      "learning_rate": 2.440279103127524e-07,
      "loss": 1.2492,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2782
    },
    {
      "epoch": 1.745922208281054,
      "grad_norm": 1.8052780628204346,
      "learning_rate": 2.428500080753676e-07,
      "loss": 1.2869,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2783
    },
    {
      "epoch": 1.7465495608531993,
      "grad_norm": 1.9134352207183838,
      "learning_rate": 2.416748103961625e-07,
      "loss": 1.13,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2784
    },
    {
      "epoch": 1.7471769134253452,
      "grad_norm": 1.935655951499939,
      "learning_rate": 2.4050231868328026e-07,
      "loss": 1.3445,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2785
    },
    {
      "epoch": 1.7478042659974906,
      "grad_norm": 1.5137767791748047,
      "learning_rate": 2.3933253434162193e-07,
      "loss": 1.0256,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2786
    },
    {
      "epoch": 1.748431618569636,
      "grad_norm": 1.7359564304351807,
      "learning_rate": 2.381654587728438e-07,
      "loss": 1.268,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2787
    },
    {
      "epoch": 1.7490589711417817,
      "grad_norm": 1.6612578630447388,
      "learning_rate": 2.3700109337535803e-07,
      "loss": 1.4344,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2788
    },
    {
      "epoch": 1.7496863237139273,
      "grad_norm": 1.733493685722351,
      "learning_rate": 2.3583943954432725e-07,
      "loss": 1.3073,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2789
    },
    {
      "epoch": 1.7503136762860727,
      "grad_norm": 1.6784566640853882,
      "learning_rate": 2.3468049867166747e-07,
      "loss": 1.1749,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2790
    },
    {
      "epoch": 1.7509410288582183,
      "grad_norm": 1.6308798789978027,
      "learning_rate": 2.335242721460415e-07,
      "loss": 1.2556,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2791
    },
    {
      "epoch": 1.751568381430364,
      "grad_norm": 1.7757388353347778,
      "learning_rate": 2.3237076135286224e-07,
      "loss": 1.3496,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2792
    },
    {
      "epoch": 1.7521957340025094,
      "grad_norm": 2.260532855987549,
      "learning_rate": 2.3121996767428655e-07,
      "loss": 1.1303,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2793
    },
    {
      "epoch": 1.7528230865746548,
      "grad_norm": 1.8579128980636597,
      "learning_rate": 2.300718924892159e-07,
      "loss": 1.3093,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2794
    },
    {
      "epoch": 1.7534504391468007,
      "grad_norm": 1.677127718925476,
      "learning_rate": 2.28926537173296e-07,
      "loss": 1.2947,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2795
    },
    {
      "epoch": 1.754077791718946,
      "grad_norm": 1.9110498428344727,
      "learning_rate": 2.2778390309891178e-07,
      "loss": 1.1955,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2796
    },
    {
      "epoch": 1.7547051442910915,
      "grad_norm": 1.6374671459197998,
      "learning_rate": 2.2664399163518786e-07,
      "loss": 1.2818,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2797
    },
    {
      "epoch": 1.7553324968632371,
      "grad_norm": 1.7491819858551025,
      "learning_rate": 2.255068041479877e-07,
      "loss": 1.2249,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2798
    },
    {
      "epoch": 1.7559598494353827,
      "grad_norm": 1.700737476348877,
      "learning_rate": 2.243723419999097e-07,
      "loss": 1.1558,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2799
    },
    {
      "epoch": 1.7565872020075282,
      "grad_norm": 1.699294090270996,
      "learning_rate": 2.232406065502868e-07,
      "loss": 1.2829,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2800
    },
    {
      "epoch": 1.7572145545796738,
      "grad_norm": 1.671245813369751,
      "learning_rate": 2.2211159915518477e-07,
      "loss": 1.1871,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2801
    },
    {
      "epoch": 1.7578419071518194,
      "grad_norm": 1.7521545886993408,
      "learning_rate": 2.2098532116740152e-07,
      "loss": 1.3888,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2802
    },
    {
      "epoch": 1.7584692597239648,
      "grad_norm": 1.4503552913665771,
      "learning_rate": 2.1986177393646307e-07,
      "loss": 1.1318,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2803
    },
    {
      "epoch": 1.7590966122961103,
      "grad_norm": 2.4087252616882324,
      "learning_rate": 2.1874095880862505e-07,
      "loss": 1.3339,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2804
    },
    {
      "epoch": 1.759723964868256,
      "grad_norm": 1.7991889715194702,
      "learning_rate": 2.176228771268682e-07,
      "loss": 1.1475,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2805
    },
    {
      "epoch": 1.7603513174404015,
      "grad_norm": 1.8192427158355713,
      "learning_rate": 2.165075302308975e-07,
      "loss": 1.1967,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2806
    },
    {
      "epoch": 1.760978670012547,
      "grad_norm": 1.9129115343093872,
      "learning_rate": 2.1539491945714337e-07,
      "loss": 1.264,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2807
    },
    {
      "epoch": 1.7616060225846926,
      "grad_norm": 1.306735634803772,
      "learning_rate": 2.1428504613875466e-07,
      "loss": 1.0545,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2808
    },
    {
      "epoch": 1.7622333751568382,
      "grad_norm": 1.6830686330795288,
      "learning_rate": 2.131779116056032e-07,
      "loss": 1.3171,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2809
    },
    {
      "epoch": 1.7628607277289836,
      "grad_norm": 1.8796741962432861,
      "learning_rate": 2.1207351718427667e-07,
      "loss": 1.1889,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2810
    },
    {
      "epoch": 1.7634880803011292,
      "grad_norm": 1.8853201866149902,
      "learning_rate": 2.1097186419808151e-07,
      "loss": 1.1654,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2811
    },
    {
      "epoch": 1.7641154328732749,
      "grad_norm": 1.8712329864501953,
      "learning_rate": 2.0987295396703772e-07,
      "loss": 1.2522,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2812
    },
    {
      "epoch": 1.7647427854454203,
      "grad_norm": 1.628964900970459,
      "learning_rate": 2.0877678780787918e-07,
      "loss": 1.1069,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2813
    },
    {
      "epoch": 1.765370138017566,
      "grad_norm": 1.801734447479248,
      "learning_rate": 2.0768336703405335e-07,
      "loss": 1.1379,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2814
    },
    {
      "epoch": 1.7659974905897116,
      "grad_norm": 1.8092901706695557,
      "learning_rate": 2.0659269295571603e-07,
      "loss": 1.1901,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2815
    },
    {
      "epoch": 1.766624843161857,
      "grad_norm": 3.3052401542663574,
      "learning_rate": 2.0550476687973274e-07,
      "loss": 1.1334,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2816
    },
    {
      "epoch": 1.7672521957340024,
      "grad_norm": 1.5364779233932495,
      "learning_rate": 2.04419590109676e-07,
      "loss": 1.2076,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2817
    },
    {
      "epoch": 1.767879548306148,
      "grad_norm": 1.5080270767211914,
      "learning_rate": 2.0333716394582536e-07,
      "loss": 1.1618,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2818
    },
    {
      "epoch": 1.7685069008782937,
      "grad_norm": 1.584987759590149,
      "learning_rate": 2.0225748968516284e-07,
      "loss": 1.1514,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2819
    },
    {
      "epoch": 1.769134253450439,
      "grad_norm": 1.841313362121582,
      "learning_rate": 2.0118056862137358e-07,
      "loss": 1.1889,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2820
    },
    {
      "epoch": 1.7697616060225847,
      "grad_norm": 1.4839106798171997,
      "learning_rate": 2.0010640204484455e-07,
      "loss": 1.124,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2821
    },
    {
      "epoch": 1.7703889585947303,
      "grad_norm": 2.0453951358795166,
      "learning_rate": 1.9903499124266118e-07,
      "loss": 1.2448,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2822
    },
    {
      "epoch": 1.7710163111668757,
      "grad_norm": 1.7970421314239502,
      "learning_rate": 1.9796633749860795e-07,
      "loss": 1.4263,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2823
    },
    {
      "epoch": 1.7716436637390214,
      "grad_norm": 1.8288103342056274,
      "learning_rate": 1.9690044209316445e-07,
      "loss": 1.1157,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2824
    },
    {
      "epoch": 1.772271016311167,
      "grad_norm": 1.8434520959854126,
      "learning_rate": 1.958373063035071e-07,
      "loss": 1.3835,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2825
    },
    {
      "epoch": 1.7728983688833124,
      "grad_norm": 1.594417691230774,
      "learning_rate": 1.947769314035036e-07,
      "loss": 1.1717,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2826
    },
    {
      "epoch": 1.7735257214554578,
      "grad_norm": 1.8379368782043457,
      "learning_rate": 1.937193186637143e-07,
      "loss": 1.2795,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2827
    },
    {
      "epoch": 1.7741530740276035,
      "grad_norm": 1.6096771955490112,
      "learning_rate": 1.926644693513907e-07,
      "loss": 1.2423,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2828
    },
    {
      "epoch": 1.774780426599749,
      "grad_norm": 1.4958261251449585,
      "learning_rate": 1.916123847304721e-07,
      "loss": 1.2482,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2829
    },
    {
      "epoch": 1.7754077791718945,
      "grad_norm": 2.062771797180176,
      "learning_rate": 1.9056306606158563e-07,
      "loss": 1.2613,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2830
    },
    {
      "epoch": 1.7760351317440402,
      "grad_norm": 1.9094319343566895,
      "learning_rate": 1.895165146020439e-07,
      "loss": 1.2073,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2831
    },
    {
      "epoch": 1.7766624843161858,
      "grad_norm": 1.3328020572662354,
      "learning_rate": 1.8847273160584378e-07,
      "loss": 1.1798,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2832
    },
    {
      "epoch": 1.7772898368883312,
      "grad_norm": 1.8016719818115234,
      "learning_rate": 1.8743171832366512e-07,
      "loss": 1.106,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2833
    },
    {
      "epoch": 1.7779171894604768,
      "grad_norm": 1.3392291069030762,
      "learning_rate": 1.8639347600286877e-07,
      "loss": 1.2876,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2834
    },
    {
      "epoch": 1.7785445420326225,
      "grad_norm": 1.6724977493286133,
      "learning_rate": 1.8535800588749598e-07,
      "loss": 1.1286,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2835
    },
    {
      "epoch": 1.7791718946047679,
      "grad_norm": 1.430114984512329,
      "learning_rate": 1.8432530921826537e-07,
      "loss": 1.1618,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2836
    },
    {
      "epoch": 1.7797992471769133,
      "grad_norm": 1.1978938579559326,
      "learning_rate": 1.8329538723257352e-07,
      "loss": 1.209,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2837
    },
    {
      "epoch": 1.780426599749059,
      "grad_norm": 1.695947527885437,
      "learning_rate": 1.8226824116449076e-07,
      "loss": 1.3051,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2838
    },
    {
      "epoch": 1.7810539523212046,
      "grad_norm": 1.7131558656692505,
      "learning_rate": 1.8124387224476347e-07,
      "loss": 1.2666,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2839
    },
    {
      "epoch": 1.78168130489335,
      "grad_norm": 1.789499044418335,
      "learning_rate": 1.802222817008084e-07,
      "loss": 1.1785,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2840
    },
    {
      "epoch": 1.7823086574654956,
      "grad_norm": 1.9134174585342407,
      "learning_rate": 1.7920347075671335e-07,
      "loss": 1.3004,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2841
    },
    {
      "epoch": 1.7829360100376412,
      "grad_norm": 1.8754029273986816,
      "learning_rate": 1.7818744063323735e-07,
      "loss": 1.2351,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2842
    },
    {
      "epoch": 1.7835633626097867,
      "grad_norm": 1.7078591585159302,
      "learning_rate": 1.7717419254780488e-07,
      "loss": 1.0923,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2843
    },
    {
      "epoch": 1.7841907151819323,
      "grad_norm": 1.804856538772583,
      "learning_rate": 1.761637277145095e-07,
      "loss": 1.0874,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2844
    },
    {
      "epoch": 1.784818067754078,
      "grad_norm": 1.6386200189590454,
      "learning_rate": 1.7515604734410745e-07,
      "loss": 1.195,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2845
    },
    {
      "epoch": 1.7854454203262233,
      "grad_norm": 1.5458471775054932,
      "learning_rate": 1.7415115264402065e-07,
      "loss": 1.2134,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2846
    },
    {
      "epoch": 1.7860727728983687,
      "grad_norm": 1.5821287631988525,
      "learning_rate": 1.7314904481833178e-07,
      "loss": 1.2658,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2847
    },
    {
      "epoch": 1.7867001254705144,
      "grad_norm": 1.736161231994629,
      "learning_rate": 1.7214972506778476e-07,
      "loss": 1.1598,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2848
    },
    {
      "epoch": 1.78732747804266,
      "grad_norm": 1.7591357231140137,
      "learning_rate": 1.7115319458978236e-07,
      "loss": 1.2577,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2849
    },
    {
      "epoch": 1.7879548306148054,
      "grad_norm": 1.8174993991851807,
      "learning_rate": 1.7015945457838524e-07,
      "loss": 1.1333,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2850
    },
    {
      "epoch": 1.788582183186951,
      "grad_norm": 1.7169201374053955,
      "learning_rate": 1.6916850622431175e-07,
      "loss": 1.3502,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2851
    },
    {
      "epoch": 1.7892095357590967,
      "grad_norm": 2.1507155895233154,
      "learning_rate": 1.681803507149335e-07,
      "loss": 1.2511,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2852
    },
    {
      "epoch": 1.789836888331242,
      "grad_norm": 1.5810800790786743,
      "learning_rate": 1.6719498923427697e-07,
      "loss": 1.1941,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2853
    },
    {
      "epoch": 1.7904642409033877,
      "grad_norm": 1.6978322267532349,
      "learning_rate": 1.6621242296301966e-07,
      "loss": 1.2615,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2854
    },
    {
      "epoch": 1.7910915934755334,
      "grad_norm": 1.8557056188583374,
      "learning_rate": 1.6523265307849035e-07,
      "loss": 1.1719,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2855
    },
    {
      "epoch": 1.7917189460476788,
      "grad_norm": 1.9752377271652222,
      "learning_rate": 1.6425568075466775e-07,
      "loss": 1.2727,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2856
    },
    {
      "epoch": 1.7923462986198242,
      "grad_norm": 1.8385305404663086,
      "learning_rate": 1.6328150716217682e-07,
      "loss": 1.1641,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2857
    },
    {
      "epoch": 1.79297365119197,
      "grad_norm": 4.611957550048828,
      "learning_rate": 1.6231013346829138e-07,
      "loss": 1.258,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2858
    },
    {
      "epoch": 1.7936010037641155,
      "grad_norm": 1.7183728218078613,
      "learning_rate": 1.6134156083692843e-07,
      "loss": 1.2715,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2859
    },
    {
      "epoch": 1.7942283563362609,
      "grad_norm": 1.6445415019989014,
      "learning_rate": 1.6037579042864876e-07,
      "loss": 1.1439,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2860
    },
    {
      "epoch": 1.7948557089084065,
      "grad_norm": 1.642172932624817,
      "learning_rate": 1.59412823400657e-07,
      "loss": 1.1275,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2861
    },
    {
      "epoch": 1.7954830614805521,
      "grad_norm": 1.7930870056152344,
      "learning_rate": 1.584526609067971e-07,
      "loss": 1.1603,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2862
    },
    {
      "epoch": 1.7961104140526976,
      "grad_norm": 1.725569248199463,
      "learning_rate": 1.574953040975538e-07,
      "loss": 1.1794,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2863
    },
    {
      "epoch": 1.7967377666248432,
      "grad_norm": 1.5991744995117188,
      "learning_rate": 1.5654075412004893e-07,
      "loss": 1.3509,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2864
    },
    {
      "epoch": 1.7973651191969888,
      "grad_norm": 1.9339492321014404,
      "learning_rate": 1.55589012118042e-07,
      "loss": 1.18,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2865
    },
    {
      "epoch": 1.7979924717691342,
      "grad_norm": 1.4420740604400635,
      "learning_rate": 1.5464007923192719e-07,
      "loss": 1.2373,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2866
    },
    {
      "epoch": 1.7986198243412796,
      "grad_norm": 1.912462592124939,
      "learning_rate": 1.5369395659873305e-07,
      "loss": 1.3134,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2867
    },
    {
      "epoch": 1.7992471769134255,
      "grad_norm": 1.7321563959121704,
      "learning_rate": 1.5275064535212186e-07,
      "loss": 1.1722,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2868
    },
    {
      "epoch": 1.799874529485571,
      "grad_norm": 1.7911001443862915,
      "learning_rate": 1.5181014662238507e-07,
      "loss": 1.1812,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2869
    },
    {
      "epoch": 1.8005018820577163,
      "grad_norm": 1.928369164466858,
      "learning_rate": 1.5087246153644651e-07,
      "loss": 1.3139,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2870
    },
    {
      "epoch": 1.801129234629862,
      "grad_norm": 1.7555097341537476,
      "learning_rate": 1.4993759121785635e-07,
      "loss": 1.1508,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2871
    },
    {
      "epoch": 1.8017565872020076,
      "grad_norm": 1.8406260013580322,
      "learning_rate": 1.4900553678679457e-07,
      "loss": 1.1167,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2872
    },
    {
      "epoch": 1.802383939774153,
      "grad_norm": 1.290350317955017,
      "learning_rate": 1.4807629936006485e-07,
      "loss": 1.13,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2873
    },
    {
      "epoch": 1.8030112923462986,
      "grad_norm": 1.650330662727356,
      "learning_rate": 1.471498800510962e-07,
      "loss": 1.2608,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2874
    },
    {
      "epoch": 1.8036386449184443,
      "grad_norm": 1.4686672687530518,
      "learning_rate": 1.4622627996994154e-07,
      "loss": 1.1727,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2875
    },
    {
      "epoch": 1.8042659974905897,
      "grad_norm": 2.0812888145446777,
      "learning_rate": 1.4530550022327532e-07,
      "loss": 1.1651,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2876
    },
    {
      "epoch": 1.804893350062735,
      "grad_norm": 1.7088572978973389,
      "learning_rate": 1.4438754191439254e-07,
      "loss": 1.2074,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2877
    },
    {
      "epoch": 1.805520702634881,
      "grad_norm": 1.4053688049316406,
      "learning_rate": 1.4347240614320717e-07,
      "loss": 1.2422,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2878
    },
    {
      "epoch": 1.8061480552070264,
      "grad_norm": 1.803846836090088,
      "learning_rate": 1.4256009400625216e-07,
      "loss": 1.3032,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2879
    },
    {
      "epoch": 1.8067754077791718,
      "grad_norm": 2.0060648918151855,
      "learning_rate": 1.4165060659667606e-07,
      "loss": 1.231,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2880
    },
    {
      "epoch": 1.8074027603513174,
      "grad_norm": 1.4810571670532227,
      "learning_rate": 1.407439450042433e-07,
      "loss": 1.1486,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2881
    },
    {
      "epoch": 1.808030112923463,
      "grad_norm": 1.8344682455062866,
      "learning_rate": 1.398401103153324e-07,
      "loss": 1.2441,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2882
    },
    {
      "epoch": 1.8086574654956085,
      "grad_norm": 1.9145503044128418,
      "learning_rate": 1.3893910361293422e-07,
      "loss": 1.2898,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2883
    },
    {
      "epoch": 1.809284818067754,
      "grad_norm": 1.8603219985961914,
      "learning_rate": 1.3804092597665186e-07,
      "loss": 1.1516,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2884
    },
    {
      "epoch": 1.8099121706398997,
      "grad_norm": 1.771806240081787,
      "learning_rate": 1.3714557848269777e-07,
      "loss": 1.3643,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2885
    },
    {
      "epoch": 1.8105395232120451,
      "grad_norm": 2.0314760208129883,
      "learning_rate": 1.3625306220389378e-07,
      "loss": 1.2896,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2886
    },
    {
      "epoch": 1.8111668757841906,
      "grad_norm": 1.787047028541565,
      "learning_rate": 1.3536337820966915e-07,
      "loss": 1.2796,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2887
    },
    {
      "epoch": 1.8117942283563364,
      "grad_norm": 1.9423733949661255,
      "learning_rate": 1.3447652756605894e-07,
      "loss": 1.122,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2888
    },
    {
      "epoch": 1.8124215809284818,
      "grad_norm": 1.8853334188461304,
      "learning_rate": 1.3359251133570422e-07,
      "loss": 1.3846,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2889
    },
    {
      "epoch": 1.8130489335006272,
      "grad_norm": 1.9998983144760132,
      "learning_rate": 1.3271133057784906e-07,
      "loss": 1.3012,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2890
    },
    {
      "epoch": 1.8136762860727729,
      "grad_norm": 1.855903148651123,
      "learning_rate": 1.3183298634834086e-07,
      "loss": 1.2702,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2891
    },
    {
      "epoch": 1.8143036386449185,
      "grad_norm": 2.0532174110412598,
      "learning_rate": 1.3095747969962686e-07,
      "loss": 1.1433,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2892
    },
    {
      "epoch": 1.814930991217064,
      "grad_norm": 1.7956229448318481,
      "learning_rate": 1.3008481168075571e-07,
      "loss": 1.1543,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2893
    },
    {
      "epoch": 1.8155583437892095,
      "grad_norm": 1.9425783157348633,
      "learning_rate": 1.2921498333737375e-07,
      "loss": 1.3422,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2894
    },
    {
      "epoch": 1.8161856963613552,
      "grad_norm": 1.777909278869629,
      "learning_rate": 1.283479957117248e-07,
      "loss": 1.317,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2895
    },
    {
      "epoch": 1.8168130489335006,
      "grad_norm": 1.7789652347564697,
      "learning_rate": 1.2748384984265004e-07,
      "loss": 1.3224,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2896
    },
    {
      "epoch": 1.817440401505646,
      "grad_norm": 1.7923486232757568,
      "learning_rate": 1.2662254676558406e-07,
      "loss": 1.2084,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2897
    },
    {
      "epoch": 1.8180677540777919,
      "grad_norm": 1.7949541807174683,
      "learning_rate": 1.2576408751255624e-07,
      "loss": 1.294,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2898
    },
    {
      "epoch": 1.8186951066499373,
      "grad_norm": 1.7665704488754272,
      "learning_rate": 1.2490847311218773e-07,
      "loss": 1.3085,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2899
    },
    {
      "epoch": 1.8193224592220827,
      "grad_norm": 1.8380930423736572,
      "learning_rate": 1.2405570458969173e-07,
      "loss": 1.3413,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2900
    },
    {
      "epoch": 1.8199498117942283,
      "grad_norm": 1.823759913444519,
      "learning_rate": 1.232057829668709e-07,
      "loss": 1.3157,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2901
    },
    {
      "epoch": 1.820577164366374,
      "grad_norm": 1.807655930519104,
      "learning_rate": 1.223587092621162e-07,
      "loss": 1.3335,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2902
    },
    {
      "epoch": 1.8212045169385194,
      "grad_norm": 1.7860008478164673,
      "learning_rate": 1.2151448449040759e-07,
      "loss": 1.3931,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2903
    },
    {
      "epoch": 1.821831869510665,
      "grad_norm": 0.9793686270713806,
      "learning_rate": 1.2067310966330958e-07,
      "loss": 1.1069,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2904
    },
    {
      "epoch": 1.8224592220828106,
      "grad_norm": 1.4274057149887085,
      "learning_rate": 1.1983458578897404e-07,
      "loss": 1.2951,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2905
    },
    {
      "epoch": 1.823086574654956,
      "grad_norm": 1.787569284439087,
      "learning_rate": 1.1899891387213425e-07,
      "loss": 1.1813,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2906
    },
    {
      "epoch": 1.8237139272271017,
      "grad_norm": 2.3724148273468018,
      "learning_rate": 1.1816609491410864e-07,
      "loss": 1.2176,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2907
    },
    {
      "epoch": 1.8243412797992473,
      "grad_norm": 1.8867779970169067,
      "learning_rate": 1.1733612991279536e-07,
      "loss": 1.2545,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2908
    },
    {
      "epoch": 1.8249686323713927,
      "grad_norm": 1.7058035135269165,
      "learning_rate": 1.1650901986267365e-07,
      "loss": 1.1629,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2909
    },
    {
      "epoch": 1.8255959849435381,
      "grad_norm": 1.5623539686203003,
      "learning_rate": 1.1568476575480186e-07,
      "loss": 1.1691,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2910
    },
    {
      "epoch": 1.8262233375156838,
      "grad_norm": 1.742311716079712,
      "learning_rate": 1.1486336857681635e-07,
      "loss": 1.2359,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2911
    },
    {
      "epoch": 1.8268506900878294,
      "grad_norm": 1.8960965871810913,
      "learning_rate": 1.1404482931293009e-07,
      "loss": 1.2743,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2912
    },
    {
      "epoch": 1.8274780426599748,
      "grad_norm": 1.8217061758041382,
      "learning_rate": 1.1322914894393188e-07,
      "loss": 1.3185,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2913
    },
    {
      "epoch": 1.8281053952321205,
      "grad_norm": 1.695849895477295,
      "learning_rate": 1.1241632844718464e-07,
      "loss": 1.2637,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2914
    },
    {
      "epoch": 1.828732747804266,
      "grad_norm": 1.9500839710235596,
      "learning_rate": 1.1160636879662456e-07,
      "loss": 1.2737,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2915
    },
    {
      "epoch": 1.8293601003764115,
      "grad_norm": 1.6682372093200684,
      "learning_rate": 1.1079927096275978e-07,
      "loss": 1.1241,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2916
    },
    {
      "epoch": 1.8299874529485571,
      "grad_norm": 1.7813045978546143,
      "learning_rate": 1.099950359126703e-07,
      "loss": 1.3649,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2917
    },
    {
      "epoch": 1.8306148055207028,
      "grad_norm": 1.7439370155334473,
      "learning_rate": 1.0919366461000447e-07,
      "loss": 1.1933,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2918
    },
    {
      "epoch": 1.8312421580928482,
      "grad_norm": 2.0448148250579834,
      "learning_rate": 1.0839515801498085e-07,
      "loss": 1.1745,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2919
    },
    {
      "epoch": 1.8318695106649936,
      "grad_norm": 1.9162275791168213,
      "learning_rate": 1.0759951708438382e-07,
      "loss": 1.2988,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2920
    },
    {
      "epoch": 1.8324968632371392,
      "grad_norm": 1.8003709316253662,
      "learning_rate": 1.0680674277156467e-07,
      "loss": 1.195,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2921
    },
    {
      "epoch": 1.8331242158092849,
      "grad_norm": 1.7913392782211304,
      "learning_rate": 1.0601683602644102e-07,
      "loss": 1.3184,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2922
    },
    {
      "epoch": 1.8337515683814303,
      "grad_norm": 1.3972041606903076,
      "learning_rate": 1.052297977954922e-07,
      "loss": 1.1933,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2923
    },
    {
      "epoch": 1.834378920953576,
      "grad_norm": 2.2087419033050537,
      "learning_rate": 1.0444562902176297e-07,
      "loss": 1.1391,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2924
    },
    {
      "epoch": 1.8350062735257215,
      "grad_norm": 1.5424485206604004,
      "learning_rate": 1.0366433064485759e-07,
      "loss": 1.1697,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2925
    },
    {
      "epoch": 1.835633626097867,
      "grad_norm": 1.494396686553955,
      "learning_rate": 1.0288590360094275e-07,
      "loss": 1.1096,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2926
    },
    {
      "epoch": 1.8362609786700126,
      "grad_norm": 2.128662586212158,
      "learning_rate": 1.0211034882274345e-07,
      "loss": 1.1278,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2927
    },
    {
      "epoch": 1.8368883312421582,
      "grad_norm": 1.8385462760925293,
      "learning_rate": 1.0133766723954358e-07,
      "loss": 1.2563,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2928
    },
    {
      "epoch": 1.8375156838143036,
      "grad_norm": 1.823840856552124,
      "learning_rate": 1.0056785977718448e-07,
      "loss": 1.3391,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2929
    },
    {
      "epoch": 1.838143036386449,
      "grad_norm": 1.6410809755325317,
      "learning_rate": 9.98009273580633e-08,
      "loss": 1.116,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2930
    },
    {
      "epoch": 1.8387703889585947,
      "grad_norm": 1.597440242767334,
      "learning_rate": 9.903687090113246e-08,
      "loss": 1.1979,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2931
    },
    {
      "epoch": 1.8393977415307403,
      "grad_norm": 1.8229918479919434,
      "learning_rate": 9.827569132189824e-08,
      "loss": 1.2552,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2932
    },
    {
      "epoch": 1.8400250941028857,
      "grad_norm": 1.5379549264907837,
      "learning_rate": 9.751738953241996e-08,
      "loss": 1.2338,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2933
    },
    {
      "epoch": 1.8406524466750314,
      "grad_norm": 2.0025312900543213,
      "learning_rate": 9.676196644130858e-08,
      "loss": 1.0986,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2934
    },
    {
      "epoch": 1.841279799247177,
      "grad_norm": 1.8701595067977905,
      "learning_rate": 9.600942295372562e-08,
      "loss": 1.2783,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2935
    },
    {
      "epoch": 1.8419071518193224,
      "grad_norm": 1.8206629753112793,
      "learning_rate": 9.525975997138254e-08,
      "loss": 1.2304,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2936
    },
    {
      "epoch": 1.842534504391468,
      "grad_norm": 1.8754552602767944,
      "learning_rate": 9.451297839253915e-08,
      "loss": 1.2556,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2937
    },
    {
      "epoch": 1.8431618569636137,
      "grad_norm": 1.714375376701355,
      "learning_rate": 9.3769079112003e-08,
      "loss": 1.1894,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2938
    },
    {
      "epoch": 1.843789209535759,
      "grad_norm": 1.7647253274917603,
      "learning_rate": 9.302806302112694e-08,
      "loss": 1.1737,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2939
    },
    {
      "epoch": 1.8444165621079045,
      "grad_norm": 1.630103349685669,
      "learning_rate": 9.228993100781097e-08,
      "loss": 1.2667,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2940
    },
    {
      "epoch": 1.8450439146800501,
      "grad_norm": 1.5752959251403809,
      "learning_rate": 9.155468395649764e-08,
      "loss": 1.2148,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2941
    },
    {
      "epoch": 1.8456712672521958,
      "grad_norm": 1.9237995147705078,
      "learning_rate": 9.082232274817332e-08,
      "loss": 1.18,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2942
    },
    {
      "epoch": 1.8462986198243412,
      "grad_norm": 1.7361148595809937,
      "learning_rate": 9.00928482603669e-08,
      "loss": 1.306,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2943
    },
    {
      "epoch": 1.8469259723964868,
      "grad_norm": 1.6503828763961792,
      "learning_rate": 8.936626136714754e-08,
      "loss": 1.4332,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2944
    },
    {
      "epoch": 1.8475533249686324,
      "grad_norm": 2.4304773807525635,
      "learning_rate": 8.864256293912521e-08,
      "loss": 1.346,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2945
    },
    {
      "epoch": 1.8481806775407779,
      "grad_norm": 1.8197497129440308,
      "learning_rate": 8.792175384344848e-08,
      "loss": 1.0988,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2946
    },
    {
      "epoch": 1.8488080301129235,
      "grad_norm": 1.7978287935256958,
      "learning_rate": 8.720383494380397e-08,
      "loss": 1.282,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2947
    },
    {
      "epoch": 1.8494353826850691,
      "grad_norm": 1.7161986827850342,
      "learning_rate": 8.648880710041496e-08,
      "loss": 1.2886,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2948
    },
    {
      "epoch": 1.8500627352572145,
      "grad_norm": 1.6135600805282593,
      "learning_rate": 8.577667117004084e-08,
      "loss": 1.1518,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2949
    },
    {
      "epoch": 1.85069008782936,
      "grad_norm": 1.6712685823440552,
      "learning_rate": 8.506742800597601e-08,
      "loss": 1.3558,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2950
    },
    {
      "epoch": 1.8513174404015058,
      "grad_norm": 1.5128730535507202,
      "learning_rate": 8.436107845804842e-08,
      "loss": 1.1955,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2951
    },
    {
      "epoch": 1.8519447929736512,
      "grad_norm": 1.6145243644714355,
      "learning_rate": 8.365762337261885e-08,
      "loss": 1.2239,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2952
    },
    {
      "epoch": 1.8525721455457966,
      "grad_norm": 1.9505369663238525,
      "learning_rate": 8.295706359257998e-08,
      "loss": 1.1648,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2953
    },
    {
      "epoch": 1.8531994981179423,
      "grad_norm": 1.492814302444458,
      "learning_rate": 8.225939995735593e-08,
      "loss": 1.2614,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2954
    },
    {
      "epoch": 1.853826850690088,
      "grad_norm": 1.5193251371383667,
      "learning_rate": 8.15646333028991e-08,
      "loss": 1.3263,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2955
    },
    {
      "epoch": 1.8544542032622333,
      "grad_norm": 1.7506705522537231,
      "learning_rate": 8.087276446169162e-08,
      "loss": 1.3393,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2956
    },
    {
      "epoch": 1.855081555834379,
      "grad_norm": 1.7187601327896118,
      "learning_rate": 8.018379426274397e-08,
      "loss": 1.2017,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2957
    },
    {
      "epoch": 1.8557089084065246,
      "grad_norm": 1.3737879991531372,
      "learning_rate": 7.949772353159191e-08,
      "loss": 1.104,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2958
    },
    {
      "epoch": 1.85633626097867,
      "grad_norm": 1.9042035341262817,
      "learning_rate": 7.881455309029895e-08,
      "loss": 1.216,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2959
    },
    {
      "epoch": 1.8569636135508154,
      "grad_norm": 1.4430084228515625,
      "learning_rate": 7.813428375745142e-08,
      "loss": 1.2731,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2960
    },
    {
      "epoch": 1.8575909661229613,
      "grad_norm": 1.2014703750610352,
      "learning_rate": 7.745691634816032e-08,
      "loss": 1.1147,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2961
    },
    {
      "epoch": 1.8582183186951067,
      "grad_norm": 1.566699504852295,
      "learning_rate": 7.678245167406061e-08,
      "loss": 1.1803,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2962
    },
    {
      "epoch": 1.858845671267252,
      "grad_norm": 1.5141801834106445,
      "learning_rate": 7.611089054330723e-08,
      "loss": 1.1781,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2963
    },
    {
      "epoch": 1.8594730238393977,
      "grad_norm": 1.6204683780670166,
      "learning_rate": 7.544223376057702e-08,
      "loss": 1.1162,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2964
    },
    {
      "epoch": 1.8601003764115434,
      "grad_norm": 1.9046368598937988,
      "learning_rate": 7.477648212706746e-08,
      "loss": 1.33,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2965
    },
    {
      "epoch": 1.8607277289836888,
      "grad_norm": 1.694454550743103,
      "learning_rate": 7.411363644049346e-08,
      "loss": 1.0354,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2966
    },
    {
      "epoch": 1.8613550815558344,
      "grad_norm": 1.9594048261642456,
      "learning_rate": 7.345369749508891e-08,
      "loss": 1.2649,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2967
    },
    {
      "epoch": 1.86198243412798,
      "grad_norm": 1.9656914472579956,
      "learning_rate": 7.279666608160458e-08,
      "loss": 1.2194,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2968
    },
    {
      "epoch": 1.8626097867001254,
      "grad_norm": 1.7213078737258911,
      "learning_rate": 7.214254298730794e-08,
      "loss": 1.3395,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2969
    },
    {
      "epoch": 1.8632371392722709,
      "grad_norm": 1.6851757764816284,
      "learning_rate": 7.149132899598066e-08,
      "loss": 1.2267,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2970
    },
    {
      "epoch": 1.8638644918444167,
      "grad_norm": 1.0846415758132935,
      "learning_rate": 7.084302488791994e-08,
      "loss": 1.1227,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2971
    },
    {
      "epoch": 1.8644918444165621,
      "grad_norm": 1.6302671432495117,
      "learning_rate": 7.019763143993441e-08,
      "loss": 1.2335,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2972
    },
    {
      "epoch": 1.8651191969887075,
      "grad_norm": 1.5246875286102295,
      "learning_rate": 6.955514942534742e-08,
      "loss": 1.3178,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2973
    },
    {
      "epoch": 1.8657465495608532,
      "grad_norm": 1.6435633897781372,
      "learning_rate": 6.891557961399175e-08,
      "loss": 1.3119,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2974
    },
    {
      "epoch": 1.8663739021329988,
      "grad_norm": 1.7738655805587769,
      "learning_rate": 6.827892277221193e-08,
      "loss": 1.2928,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2975
    },
    {
      "epoch": 1.8670012547051442,
      "grad_norm": 1.7210484743118286,
      "learning_rate": 6.764517966286188e-08,
      "loss": 1.2324,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2976
    },
    {
      "epoch": 1.8676286072772899,
      "grad_norm": 1.6799976825714111,
      "learning_rate": 6.701435104530363e-08,
      "loss": 1.1218,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2977
    },
    {
      "epoch": 1.8682559598494355,
      "grad_norm": 1.6705591678619385,
      "learning_rate": 6.638643767540837e-08,
      "loss": 1.1626,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2978
    },
    {
      "epoch": 1.868883312421581,
      "grad_norm": 1.8721472024917603,
      "learning_rate": 6.576144030555259e-08,
      "loss": 1.3582,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2979
    },
    {
      "epoch": 1.8695106649937263,
      "grad_norm": 1.9409172534942627,
      "learning_rate": 6.513935968461948e-08,
      "loss": 1.3134,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2980
    },
    {
      "epoch": 1.8701380175658722,
      "grad_norm": 1.959789752960205,
      "learning_rate": 6.45201965579978e-08,
      "loss": 1.2429,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2981
    },
    {
      "epoch": 1.8707653701380176,
      "grad_norm": 1.5382444858551025,
      "learning_rate": 6.39039516675799e-08,
      "loss": 1.1036,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2982
    },
    {
      "epoch": 1.871392722710163,
      "grad_norm": 1.5872652530670166,
      "learning_rate": 6.329062575176182e-08,
      "loss": 1.1977,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2983
    },
    {
      "epoch": 1.8720200752823086,
      "grad_norm": 2.0065999031066895,
      "learning_rate": 6.268021954544095e-08,
      "loss": 1.1706,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2984
    },
    {
      "epoch": 1.8726474278544543,
      "grad_norm": 1.7338385581970215,
      "learning_rate": 6.207273378001837e-08,
      "loss": 1.4278,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2985
    },
    {
      "epoch": 1.8732747804265997,
      "grad_norm": 1.912854552268982,
      "learning_rate": 6.14681691833935e-08,
      "loss": 1.2516,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2986
    },
    {
      "epoch": 1.8739021329987453,
      "grad_norm": 1.60573148727417,
      "learning_rate": 6.086652647996738e-08,
      "loss": 1.334,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2987
    },
    {
      "epoch": 1.874529485570891,
      "grad_norm": 2.0827462673187256,
      "learning_rate": 6.026780639063895e-08,
      "loss": 1.1661,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2988
    },
    {
      "epoch": 1.8751568381430364,
      "grad_norm": 1.6736165285110474,
      "learning_rate": 5.967200963280545e-08,
      "loss": 1.3655,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2989
    },
    {
      "epoch": 1.875784190715182,
      "grad_norm": 1.7067066431045532,
      "learning_rate": 5.9079136920361376e-08,
      "loss": 1.1578,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2990
    },
    {
      "epoch": 1.8764115432873276,
      "grad_norm": 1.3567057847976685,
      "learning_rate": 5.848918896369765e-08,
      "loss": 1.1204,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2991
    },
    {
      "epoch": 1.877038895859473,
      "grad_norm": 2.0910308361053467,
      "learning_rate": 5.7902166469701036e-08,
      "loss": 1.2968,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2992
    },
    {
      "epoch": 1.8776662484316184,
      "grad_norm": 1.7153366804122925,
      "learning_rate": 5.731807014175195e-08,
      "loss": 1.2327,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2993
    },
    {
      "epoch": 1.878293601003764,
      "grad_norm": 1.6810442209243774,
      "learning_rate": 5.673690067972554e-08,
      "loss": 1.1654,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2994
    },
    {
      "epoch": 1.8789209535759097,
      "grad_norm": 1.9211241006851196,
      "learning_rate": 5.615865877998977e-08,
      "loss": 1.1672,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2995
    },
    {
      "epoch": 1.8795483061480551,
      "grad_norm": 2.5992534160614014,
      "learning_rate": 5.558334513540403e-08,
      "loss": 1.1507,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2996
    },
    {
      "epoch": 1.8801756587202008,
      "grad_norm": 2.0081636905670166,
      "learning_rate": 5.5010960435320224e-08,
      "loss": 1.2349,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2997
    },
    {
      "epoch": 1.8808030112923464,
      "grad_norm": 1.8557283878326416,
      "learning_rate": 5.444150536558002e-08,
      "loss": 1.2541,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2998
    },
    {
      "epoch": 1.8814303638644918,
      "grad_norm": 2.5331225395202637,
      "learning_rate": 5.3874980608514535e-08,
      "loss": 1.2488,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 2999
    },
    {
      "epoch": 1.8820577164366374,
      "grad_norm": 1.7225735187530518,
      "learning_rate": 5.3311386842944125e-08,
      "loss": 1.279,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3000
    },
    {
      "epoch": 1.882685069008783,
      "grad_norm": 1.9744964838027954,
      "learning_rate": 5.275072474417719e-08,
      "loss": 1.1967,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3001
    },
    {
      "epoch": 1.8833124215809285,
      "grad_norm": 1.9109421968460083,
      "learning_rate": 5.2192994984009425e-08,
      "loss": 1.3013,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3002
    },
    {
      "epoch": 1.883939774153074,
      "grad_norm": 1.41983962059021,
      "learning_rate": 5.163819823072208e-08,
      "loss": 1.1192,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3003
    },
    {
      "epoch": 1.8845671267252195,
      "grad_norm": 1.7811996936798096,
      "learning_rate": 5.108633514908368e-08,
      "loss": 1.2609,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3004
    },
    {
      "epoch": 1.8851944792973652,
      "grad_norm": 1.9444513320922852,
      "learning_rate": 5.053740640034582e-08,
      "loss": 1.2662,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3005
    },
    {
      "epoch": 1.8858218318695106,
      "grad_norm": 2.0892276763916016,
      "learning_rate": 4.999141264224544e-08,
      "loss": 1.2915,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3006
    },
    {
      "epoch": 1.8864491844416562,
      "grad_norm": 1.5862740278244019,
      "learning_rate": 4.944835452900199e-08,
      "loss": 1.1375,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3007
    },
    {
      "epoch": 1.8870765370138018,
      "grad_norm": 1.6205672025680542,
      "learning_rate": 4.890823271131745e-08,
      "loss": 1.2271,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3008
    },
    {
      "epoch": 1.8877038895859473,
      "grad_norm": 1.8633159399032593,
      "learning_rate": 4.8371047836375806e-08,
      "loss": 1.1729,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3009
    },
    {
      "epoch": 1.888331242158093,
      "grad_norm": 1.459846019744873,
      "learning_rate": 4.783680054784162e-08,
      "loss": 1.2904,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3010
    },
    {
      "epoch": 1.8889585947302385,
      "grad_norm": 1.7166321277618408,
      "learning_rate": 4.730549148586006e-08,
      "loss": 1.1306,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3011
    },
    {
      "epoch": 1.889585947302384,
      "grad_norm": 1.547766089439392,
      "learning_rate": 4.677712128705464e-08,
      "loss": 1.1716,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3012
    },
    {
      "epoch": 1.8902132998745294,
      "grad_norm": 1.8805586099624634,
      "learning_rate": 4.6251690584528665e-08,
      "loss": 1.2843,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3013
    },
    {
      "epoch": 1.890840652446675,
      "grad_norm": 1.8120684623718262,
      "learning_rate": 4.5729200007862686e-08,
      "loss": 1.208,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3014
    },
    {
      "epoch": 1.8914680050188206,
      "grad_norm": 1.711977481842041,
      "learning_rate": 4.5209650183114514e-08,
      "loss": 1.2166,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3015
    },
    {
      "epoch": 1.892095357590966,
      "grad_norm": 1.627962589263916,
      "learning_rate": 4.469304173281785e-08,
      "loss": 1.1748,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3016
    },
    {
      "epoch": 1.8927227101631117,
      "grad_norm": 1.7329778671264648,
      "learning_rate": 4.417937527598226e-08,
      "loss": 1.2154,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3017
    },
    {
      "epoch": 1.8933500627352573,
      "grad_norm": 1.8263328075408936,
      "learning_rate": 4.3668651428092625e-08,
      "loss": 1.2522,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3018
    },
    {
      "epoch": 1.8939774153074027,
      "grad_norm": 2.0282092094421387,
      "learning_rate": 4.316087080110748e-08,
      "loss": 1.2865,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3019
    },
    {
      "epoch": 1.8946047678795483,
      "grad_norm": 1.613545298576355,
      "learning_rate": 4.2656034003458746e-08,
      "loss": 1.1478,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3020
    },
    {
      "epoch": 1.895232120451694,
      "grad_norm": 1.3981359004974365,
      "learning_rate": 4.215414164005116e-08,
      "loss": 1.1478,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3021
    },
    {
      "epoch": 1.8958594730238394,
      "grad_norm": 1.705987811088562,
      "learning_rate": 4.165519431226117e-08,
      "loss": 1.0804,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3022
    },
    {
      "epoch": 1.8964868255959848,
      "grad_norm": 1.808300256729126,
      "learning_rate": 4.115919261793638e-08,
      "loss": 1.2832,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3023
    },
    {
      "epoch": 1.8971141781681304,
      "grad_norm": 2.1151375770568848,
      "learning_rate": 4.0666137151395277e-08,
      "loss": 1.2959,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3024
    },
    {
      "epoch": 1.897741530740276,
      "grad_norm": 1.6511573791503906,
      "learning_rate": 4.017602850342584e-08,
      "loss": 1.4428,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3025
    },
    {
      "epoch": 1.8983688833124215,
      "grad_norm": 1.432206630706787,
      "learning_rate": 3.968886726128524e-08,
      "loss": 1.1333,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3026
    },
    {
      "epoch": 1.8989962358845671,
      "grad_norm": 1.6224288940429688,
      "learning_rate": 3.920465400869877e-08,
      "loss": 1.2393,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3027
    },
    {
      "epoch": 1.8996235884567128,
      "grad_norm": 1.6806602478027344,
      "learning_rate": 3.872338932585984e-08,
      "loss": 1.3892,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3028
    },
    {
      "epoch": 1.9002509410288582,
      "grad_norm": 2.271575927734375,
      "learning_rate": 3.8245073789427986e-08,
      "loss": 1.1943,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3029
    },
    {
      "epoch": 1.9008782936010038,
      "grad_norm": 1.6886036396026611,
      "learning_rate": 3.7769707972530046e-08,
      "loss": 1.2705,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3030
    },
    {
      "epoch": 1.9015056461731494,
      "grad_norm": 2.129394292831421,
      "learning_rate": 3.729729244475816e-08,
      "loss": 1.154,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3031
    },
    {
      "epoch": 1.9021329987452948,
      "grad_norm": 1.8195149898529053,
      "learning_rate": 3.682782777216898e-08,
      "loss": 1.2528,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3032
    },
    {
      "epoch": 1.9027603513174403,
      "grad_norm": 1.6862359046936035,
      "learning_rate": 3.6361314517283085e-08,
      "loss": 1.2078,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3033
    },
    {
      "epoch": 1.9033877038895861,
      "grad_norm": 1.1818653345108032,
      "learning_rate": 3.589775323908612e-08,
      "loss": 1.1659,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3034
    },
    {
      "epoch": 1.9040150564617315,
      "grad_norm": 1.4935318231582642,
      "learning_rate": 3.543714449302488e-08,
      "loss": 1.1956,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3035
    },
    {
      "epoch": 1.904642409033877,
      "grad_norm": 1.7930165529251099,
      "learning_rate": 3.497948883100927e-08,
      "loss": 1.2859,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3036
    },
    {
      "epoch": 1.9052697616060226,
      "grad_norm": 1.831816554069519,
      "learning_rate": 3.4524786801411195e-08,
      "loss": 1.2086,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3037
    },
    {
      "epoch": 1.9058971141781682,
      "grad_norm": 1.8915395736694336,
      "learning_rate": 3.407303894906205e-08,
      "loss": 1.2686,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3038
    },
    {
      "epoch": 1.9065244667503136,
      "grad_norm": 1.8743354082107544,
      "learning_rate": 3.362424581525498e-08,
      "loss": 1.4302,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3039
    },
    {
      "epoch": 1.9071518193224593,
      "grad_norm": 1.7570542097091675,
      "learning_rate": 3.317840793774174e-08,
      "loss": 1.3382,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3040
    },
    {
      "epoch": 1.9077791718946049,
      "grad_norm": 1.849229335784912,
      "learning_rate": 3.273552585073364e-08,
      "loss": 1.1924,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3041
    },
    {
      "epoch": 1.9084065244667503,
      "grad_norm": 1.9815819263458252,
      "learning_rate": 3.229560008490007e-08,
      "loss": 1.1692,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3042
    },
    {
      "epoch": 1.9090338770388957,
      "grad_norm": 1.822123646736145,
      "learning_rate": 3.18586311673677e-08,
      "loss": 1.2147,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3043
    },
    {
      "epoch": 1.9096612296110416,
      "grad_norm": 1.6399743556976318,
      "learning_rate": 3.142461962172105e-08,
      "loss": 1.245,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3044
    },
    {
      "epoch": 1.910288582183187,
      "grad_norm": 1.9598350524902344,
      "learning_rate": 3.099356596800024e-08,
      "loss": 1.1548,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3045
    },
    {
      "epoch": 1.9109159347553324,
      "grad_norm": 1.7600895166397095,
      "learning_rate": 3.056547072270183e-08,
      "loss": 1.2108,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3046
    },
    {
      "epoch": 1.911543287327478,
      "grad_norm": 1.9633840322494507,
      "learning_rate": 3.014033439877745e-08,
      "loss": 1.2317,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3047
    },
    {
      "epoch": 1.9121706398996237,
      "grad_norm": 1.5512139797210693,
      "learning_rate": 2.9718157505633226e-08,
      "loss": 1.3363,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3048
    },
    {
      "epoch": 1.912797992471769,
      "grad_norm": 1.947986364364624,
      "learning_rate": 2.9298940549128962e-08,
      "loss": 1.2215,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3049
    },
    {
      "epoch": 1.9134253450439147,
      "grad_norm": 1.9563969373703003,
      "learning_rate": 2.8882684031577845e-08,
      "loss": 1.2041,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3050
    },
    {
      "epoch": 1.9140526976160603,
      "grad_norm": 1.6992594003677368,
      "learning_rate": 2.846938845174646e-08,
      "loss": 1.3089,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3051
    },
    {
      "epoch": 1.9146800501882058,
      "grad_norm": 1.580871343612671,
      "learning_rate": 2.805905430485256e-08,
      "loss": 1.1649,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3052
    },
    {
      "epoch": 1.9153074027603512,
      "grad_norm": 1.5608967542648315,
      "learning_rate": 2.7651682082566743e-08,
      "loss": 1.4416,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3053
    },
    {
      "epoch": 1.915934755332497,
      "grad_norm": 1.9078105688095093,
      "learning_rate": 2.7247272273009108e-08,
      "loss": 1.3189,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3054
    },
    {
      "epoch": 1.9165621079046424,
      "grad_norm": 1.723378300666809,
      "learning_rate": 2.6845825360751198e-08,
      "loss": 1.275,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3055
    },
    {
      "epoch": 1.9171894604767878,
      "grad_norm": 1.582735300064087,
      "learning_rate": 2.6447341826814077e-08,
      "loss": 1.4529,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3056
    },
    {
      "epoch": 1.9178168130489335,
      "grad_norm": 1.8323571681976318,
      "learning_rate": 2.605182214866747e-08,
      "loss": 1.2575,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3057
    },
    {
      "epoch": 1.9184441656210791,
      "grad_norm": 1.9233479499816895,
      "learning_rate": 2.5659266800230897e-08,
      "loss": 1.3383,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3058
    },
    {
      "epoch": 1.9190715181932245,
      "grad_norm": 1.333604335784912,
      "learning_rate": 2.5269676251870878e-08,
      "loss": 1.1025,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3059
    },
    {
      "epoch": 1.9196988707653702,
      "grad_norm": 1.752911925315857,
      "learning_rate": 2.4883050970402334e-08,
      "loss": 1.2992,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3060
    },
    {
      "epoch": 1.9203262233375158,
      "grad_norm": 1.5046241283416748,
      "learning_rate": 2.4499391419086083e-08,
      "loss": 1.2568,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3061
    },
    {
      "epoch": 1.9209535759096612,
      "grad_norm": 1.6993424892425537,
      "learning_rate": 2.411869805763023e-08,
      "loss": 1.2413,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3062
    },
    {
      "epoch": 1.9215809284818066,
      "grad_norm": 1.9174491167068481,
      "learning_rate": 2.3740971342189056e-08,
      "loss": 1.1423,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3063
    },
    {
      "epoch": 1.9222082810539525,
      "grad_norm": 1.737568736076355,
      "learning_rate": 2.33662117253608e-08,
      "loss": 1.1946,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3064
    },
    {
      "epoch": 1.9228356336260979,
      "grad_norm": 1.9473689794540405,
      "learning_rate": 2.2994419656189594e-08,
      "loss": 1.3369,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3065
    },
    {
      "epoch": 1.9234629861982433,
      "grad_norm": 1.6701767444610596,
      "learning_rate": 2.262559558016325e-08,
      "loss": 1.3142,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3066
    },
    {
      "epoch": 1.924090338770389,
      "grad_norm": 1.6051278114318848,
      "learning_rate": 2.225973993921382e-08,
      "loss": 1.4871,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3067
    },
    {
      "epoch": 1.9247176913425346,
      "grad_norm": 1.8993639945983887,
      "learning_rate": 2.1896853171715916e-08,
      "loss": 1.2304,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3068
    },
    {
      "epoch": 1.92534504391468,
      "grad_norm": 1.937630295753479,
      "learning_rate": 2.1536935712486993e-08,
      "loss": 1.1656,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3069
    },
    {
      "epoch": 1.9259723964868256,
      "grad_norm": 1.7782254219055176,
      "learning_rate": 2.117998799278709e-08,
      "loss": 1.3517,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3070
    },
    {
      "epoch": 1.9265997490589712,
      "grad_norm": 1.7657594680786133,
      "learning_rate": 2.0826010440317125e-08,
      "loss": 1.1481,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3071
    },
    {
      "epoch": 1.9272271016311167,
      "grad_norm": 1.789623737335205,
      "learning_rate": 2.0475003479219492e-08,
      "loss": 1.2019,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3072
    },
    {
      "epoch": 1.927854454203262,
      "grad_norm": 1.2043588161468506,
      "learning_rate": 2.01269675300772e-08,
      "loss": 1.2038,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3073
    },
    {
      "epoch": 1.928481806775408,
      "grad_norm": 1.2602628469467163,
      "learning_rate": 1.978190300991334e-08,
      "loss": 1.0944,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3074
    },
    {
      "epoch": 1.9291091593475533,
      "grad_norm": 1.8765314817428589,
      "learning_rate": 1.94398103321905e-08,
      "loss": 1.2142,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3075
    },
    {
      "epoch": 1.9297365119196987,
      "grad_norm": 1.8176100254058838,
      "learning_rate": 1.9100689906809965e-08,
      "loss": 1.1382,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3076
    },
    {
      "epoch": 1.9303638644918444,
      "grad_norm": 1.77439546585083,
      "learning_rate": 1.876454214011253e-08,
      "loss": 1.3485,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3077
    },
    {
      "epoch": 1.93099121706399,
      "grad_norm": 1.075534701347351,
      "learning_rate": 1.8431367434876002e-08,
      "loss": 1.0438,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3078
    },
    {
      "epoch": 1.9316185696361354,
      "grad_norm": 1.7188340425491333,
      "learning_rate": 1.8101166190316876e-08,
      "loss": 1.1317,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3079
    },
    {
      "epoch": 1.932245922208281,
      "grad_norm": 1.5703520774841309,
      "learning_rate": 1.777393880208811e-08,
      "loss": 1.3624,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3080
    },
    {
      "epoch": 1.9328732747804267,
      "grad_norm": 1.767533779144287,
      "learning_rate": 1.744968566227939e-08,
      "loss": 1.0852,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3081
    },
    {
      "epoch": 1.933500627352572,
      "grad_norm": 1.7214370965957642,
      "learning_rate": 1.7128407159416604e-08,
      "loss": 1.2785,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3082
    },
    {
      "epoch": 1.9341279799247177,
      "grad_norm": 1.7513182163238525,
      "learning_rate": 1.6810103678462088e-08,
      "loss": 1.1782,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3083
    },
    {
      "epoch": 1.9347553324968634,
      "grad_norm": 1.5604854822158813,
      "learning_rate": 1.6494775600812418e-08,
      "loss": 1.2548,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3084
    },
    {
      "epoch": 1.9353826850690088,
      "grad_norm": 1.8918966054916382,
      "learning_rate": 1.6182423304299255e-08,
      "loss": 1.2621,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3085
    },
    {
      "epoch": 1.9360100376411542,
      "grad_norm": 1.689467430114746,
      "learning_rate": 1.5873047163189326e-08,
      "loss": 1.2018,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3086
    },
    {
      "epoch": 1.9366373902132998,
      "grad_norm": 1.5074502229690552,
      "learning_rate": 1.556664754818249e-08,
      "loss": 1.2346,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3087
    },
    {
      "epoch": 1.9372647427854455,
      "grad_norm": 1.5012305974960327,
      "learning_rate": 1.5263224826412292e-08,
      "loss": 1.1775,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3088
    },
    {
      "epoch": 1.9378920953575909,
      "grad_norm": 1.8126957416534424,
      "learning_rate": 1.496277936144541e-08,
      "loss": 1.4159,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3089
    },
    {
      "epoch": 1.9385194479297365,
      "grad_norm": 1.548186182975769,
      "learning_rate": 1.4665311513280822e-08,
      "loss": 1.2268,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3090
    },
    {
      "epoch": 1.9391468005018822,
      "grad_norm": 1.7500059604644775,
      "learning_rate": 1.4370821638350353e-08,
      "loss": 1.1917,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3091
    },
    {
      "epoch": 1.9397741530740276,
      "grad_norm": 1.6493536233901978,
      "learning_rate": 1.4079310089516741e-08,
      "loss": 1.0922,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3092
    },
    {
      "epoch": 1.9404015056461732,
      "grad_norm": 1.9024580717086792,
      "learning_rate": 1.3790777216074747e-08,
      "loss": 1.1731,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3093
    },
    {
      "epoch": 1.9410288582183188,
      "grad_norm": 1.912346601486206,
      "learning_rate": 1.3505223363749487e-08,
      "loss": 1.2582,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3094
    },
    {
      "epoch": 1.9416562107904642,
      "grad_norm": 1.685202717781067,
      "learning_rate": 1.3222648874696986e-08,
      "loss": 1.0968,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3095
    },
    {
      "epoch": 1.9422835633626097,
      "grad_norm": 1.895370602607727,
      "learning_rate": 1.294305408750307e-08,
      "loss": 1.2619,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3096
    },
    {
      "epoch": 1.9429109159347553,
      "grad_norm": 1.3072205781936646,
      "learning_rate": 1.2666439337183089e-08,
      "loss": 1.1337,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3097
    },
    {
      "epoch": 1.943538268506901,
      "grad_norm": 1.7828336954116821,
      "learning_rate": 1.2392804955181915e-08,
      "loss": 1.2324,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3098
    },
    {
      "epoch": 1.9441656210790463,
      "grad_norm": 1.8671600818634033,
      "learning_rate": 1.2122151269373383e-08,
      "loss": 1.2573,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3099
    },
    {
      "epoch": 1.944792973651192,
      "grad_norm": 1.743040919303894,
      "learning_rate": 1.185447860405975e-08,
      "loss": 1.2609,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3100
    },
    {
      "epoch": 1.9454203262233376,
      "grad_norm": 1.7714751958847046,
      "learning_rate": 1.158978727997112e-08,
      "loss": 1.1616,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3101
    },
    {
      "epoch": 1.946047678795483,
      "grad_norm": 1.590544581413269,
      "learning_rate": 1.1328077614265465e-08,
      "loss": 1.0432,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3102
    },
    {
      "epoch": 1.9466750313676286,
      "grad_norm": 2.0189173221588135,
      "learning_rate": 1.1069349920528327e-08,
      "loss": 1.3164,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3103
    },
    {
      "epoch": 1.9473023839397743,
      "grad_norm": 1.8704915046691895,
      "learning_rate": 1.081360450877117e-08,
      "loss": 1.186,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3104
    },
    {
      "epoch": 1.9479297365119197,
      "grad_norm": 1.8126301765441895,
      "learning_rate": 1.0560841685433864e-08,
      "loss": 1.1901,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3105
    },
    {
      "epoch": 1.948557089084065,
      "grad_norm": 1.7933433055877686,
      "learning_rate": 1.0311061753380536e-08,
      "loss": 1.1437,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3106
    },
    {
      "epoch": 1.9491844416562107,
      "grad_norm": 1.9376479387283325,
      "learning_rate": 1.006426501190233e-08,
      "loss": 1.3331,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3107
    },
    {
      "epoch": 1.9498117942283564,
      "grad_norm": 1.83939790725708,
      "learning_rate": 9.820451756715754e-09,
      "loss": 1.3347,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3108
    },
    {
      "epoch": 1.9504391468005018,
      "grad_norm": 1.802091360092163,
      "learning_rate": 9.579622279962397e-09,
      "loss": 1.2383,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3109
    },
    {
      "epoch": 1.9510664993726474,
      "grad_norm": 1.7190696001052856,
      "learning_rate": 9.341776870208096e-09,
      "loss": 1.2428,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3110
    },
    {
      "epoch": 1.951693851944793,
      "grad_norm": 1.8278087377548218,
      "learning_rate": 9.106915812443772e-09,
      "loss": 1.2512,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3111
    },
    {
      "epoch": 1.9523212045169385,
      "grad_norm": 2.022813320159912,
      "learning_rate": 8.875039388084317e-09,
      "loss": 1.3083,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3112
    },
    {
      "epoch": 1.952948557089084,
      "grad_norm": 1.781859278678894,
      "learning_rate": 8.646147874968037e-09,
      "loss": 1.2895,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3113
    },
    {
      "epoch": 1.9535759096612297,
      "grad_norm": 1.8502413034439087,
      "learning_rate": 8.420241547356933e-09,
      "loss": 1.3009,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3114
    },
    {
      "epoch": 1.9542032622333751,
      "grad_norm": 1.5541293621063232,
      "learning_rate": 8.197320675936148e-09,
      "loss": 1.1714,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3115
    },
    {
      "epoch": 1.9548306148055206,
      "grad_norm": 1.9019731283187866,
      "learning_rate": 7.977385527813963e-09,
      "loss": 1.16,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3116
    },
    {
      "epoch": 1.9554579673776662,
      "grad_norm": 1.3666876554489136,
      "learning_rate": 7.760436366519853e-09,
      "loss": 1.0896,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3117
    },
    {
      "epoch": 1.9560853199498118,
      "grad_norm": 0.9233667254447937,
      "learning_rate": 7.546473452006708e-09,
      "loss": 1.0859,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3118
    },
    {
      "epoch": 1.9567126725219572,
      "grad_norm": 1.7889835834503174,
      "learning_rate": 7.335497040648898e-09,
      "loss": 1.2028,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3119
    },
    {
      "epoch": 1.9573400250941029,
      "grad_norm": 4.186465740203857,
      "learning_rate": 7.127507385241983e-09,
      "loss": 1.2708,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3120
    },
    {
      "epoch": 1.9579673776662485,
      "grad_norm": 1.7843849658966064,
      "learning_rate": 6.922504735002999e-09,
      "loss": 1.1866,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3121
    },
    {
      "epoch": 1.958594730238394,
      "grad_norm": 1.6640843152999878,
      "learning_rate": 6.7204893355696245e-09,
      "loss": 1.2275,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3122
    },
    {
      "epoch": 1.9592220828105396,
      "grad_norm": 2.044402599334717,
      "learning_rate": 6.5214614290010085e-09,
      "loss": 1.2866,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3123
    },
    {
      "epoch": 1.9598494353826852,
      "grad_norm": 1.519188642501831,
      "learning_rate": 6.325421253775277e-09,
      "loss": 1.2247,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3124
    },
    {
      "epoch": 1.9604767879548306,
      "grad_norm": 1.9582196474075317,
      "learning_rate": 6.1323690447917525e-09,
      "loss": 1.2913,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3125
    },
    {
      "epoch": 1.961104140526976,
      "grad_norm": 2.0063629150390625,
      "learning_rate": 5.942305033369289e-09,
      "loss": 1.2017,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3126
    },
    {
      "epoch": 1.9617314930991219,
      "grad_norm": 1.6914819478988647,
      "learning_rate": 5.755229447245436e-09,
      "loss": 1.2231,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3127
    },
    {
      "epoch": 1.9623588456712673,
      "grad_norm": 1.9196043014526367,
      "learning_rate": 5.5711425105781096e-09,
      "loss": 1.2322,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3128
    },
    {
      "epoch": 1.9629861982434127,
      "grad_norm": 1.8479219675064087,
      "learning_rate": 5.390044443943365e-09,
      "loss": 1.2172,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3129
    },
    {
      "epoch": 1.9636135508155583,
      "grad_norm": 1.501006841659546,
      "learning_rate": 5.211935464336238e-09,
      "loss": 1.0868,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3130
    },
    {
      "epoch": 1.964240903387704,
      "grad_norm": 1.5361695289611816,
      "learning_rate": 5.036815785170179e-09,
      "loss": 1.0671,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3131
    },
    {
      "epoch": 1.9648682559598494,
      "grad_norm": 1.5813148021697998,
      "learning_rate": 4.864685616276788e-09,
      "loss": 1.3122,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3132
    },
    {
      "epoch": 1.965495608531995,
      "grad_norm": 1.9053575992584229,
      "learning_rate": 4.695545163905524e-09,
      "loss": 1.2872,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3133
    },
    {
      "epoch": 1.9661229611041406,
      "grad_norm": 1.71331787109375,
      "learning_rate": 4.529394630723438e-09,
      "loss": 1.2103,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3134
    },
    {
      "epoch": 1.966750313676286,
      "grad_norm": 2.0272982120513916,
      "learning_rate": 4.366234215815446e-09,
      "loss": 1.3642,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3135
    },
    {
      "epoch": 1.9673776662484315,
      "grad_norm": 1.7172526121139526,
      "learning_rate": 4.206064114682939e-09,
      "loss": 1.1586,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3136
    },
    {
      "epoch": 1.9680050188205773,
      "grad_norm": 1.4065996408462524,
      "learning_rate": 4.0488845192449e-09,
      "loss": 1.2617,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3137
    },
    {
      "epoch": 1.9686323713927227,
      "grad_norm": 1.587286353111267,
      "learning_rate": 3.894695617836786e-09,
      "loss": 1.285,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3138
    },
    {
      "epoch": 1.9692597239648681,
      "grad_norm": 1.8322442770004272,
      "learning_rate": 3.743497595210255e-09,
      "loss": 1.2328,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3139
    },
    {
      "epoch": 1.9698870765370138,
      "grad_norm": 1.8742417097091675,
      "learning_rate": 3.5952906325339988e-09,
      "loss": 1.2586,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3140
    },
    {
      "epoch": 1.9705144291091594,
      "grad_norm": 1.5411434173583984,
      "learning_rate": 3.4500749073920757e-09,
      "loss": 1.1538,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3141
    },
    {
      "epoch": 1.9711417816813048,
      "grad_norm": 13.465494155883789,
      "learning_rate": 3.3078505937844674e-09,
      "loss": 1.2148,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3142
    },
    {
      "epoch": 1.9717691342534505,
      "grad_norm": 1.985593318939209,
      "learning_rate": 3.168617862127077e-09,
      "loss": 1.1492,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3143
    },
    {
      "epoch": 1.972396486825596,
      "grad_norm": 1.571092963218689,
      "learning_rate": 3.0323768792508978e-09,
      "loss": 1.1533,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3144
    },
    {
      "epoch": 1.9730238393977415,
      "grad_norm": 1.5139721632003784,
      "learning_rate": 2.8991278084025687e-09,
      "loss": 1.2752,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3145
    },
    {
      "epoch": 1.973651191969887,
      "grad_norm": 1.6269372701644897,
      "learning_rate": 2.7688708092435403e-09,
      "loss": 1.1284,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3146
    },
    {
      "epoch": 1.9742785445420328,
      "grad_norm": 1.8731541633605957,
      "learning_rate": 2.641606037850353e-09,
      "loss": 1.2279,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3147
    },
    {
      "epoch": 1.9749058971141782,
      "grad_norm": 1.8516762256622314,
      "learning_rate": 2.5173336467135266e-09,
      "loss": 1.1847,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3148
    },
    {
      "epoch": 1.9755332496863236,
      "grad_norm": 1.9598283767700195,
      "learning_rate": 2.3960537847383946e-09,
      "loss": 1.1979,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3149
    },
    {
      "epoch": 1.9761606022584692,
      "grad_norm": 1.7261936664581299,
      "learning_rate": 2.2777665972453788e-09,
      "loss": 1.2985,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3150
    },
    {
      "epoch": 1.9767879548306149,
      "grad_norm": 1.8943806886672974,
      "learning_rate": 2.1624722259674934e-09,
      "loss": 1.2649,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3151
    },
    {
      "epoch": 1.9774153074027603,
      "grad_norm": 1.5255440473556519,
      "learning_rate": 2.050170809053398e-09,
      "loss": 1.3109,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3152
    },
    {
      "epoch": 1.978042659974906,
      "grad_norm": 1.7608319520950317,
      "learning_rate": 1.9408624810640654e-09,
      "loss": 1.2154,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3153
    },
    {
      "epoch": 1.9786700125470515,
      "grad_norm": 1.6721044778823853,
      "learning_rate": 1.834547372975004e-09,
      "loss": 1.1618,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3154
    },
    {
      "epoch": 1.979297365119197,
      "grad_norm": 1.690503716468811,
      "learning_rate": 1.7312256121748695e-09,
      "loss": 1.2142,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3155
    },
    {
      "epoch": 1.9799247176913424,
      "grad_norm": 1.3389973640441895,
      "learning_rate": 1.6308973224654634e-09,
      "loss": 1.1176,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3156
    },
    {
      "epoch": 1.9805520702634882,
      "grad_norm": 1.7939364910125732,
      "learning_rate": 1.5335626240622903e-09,
      "loss": 1.3278,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3157
    },
    {
      "epoch": 1.9811794228356336,
      "grad_norm": 2.0031301975250244,
      "learning_rate": 1.4392216335934462e-09,
      "loss": 1.2844,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3158
    },
    {
      "epoch": 1.981806775407779,
      "grad_norm": 1.8132704496383667,
      "learning_rate": 1.3478744640998963e-09,
      "loss": 1.3131,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3159
    },
    {
      "epoch": 1.9824341279799247,
      "grad_norm": 1.3770238161087036,
      "learning_rate": 1.2595212250357536e-09,
      "loss": 1.1148,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3160
    },
    {
      "epoch": 1.9830614805520703,
      "grad_norm": 1.5473389625549316,
      "learning_rate": 1.1741620222671667e-09,
      "loss": 1.0943,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3161
    },
    {
      "epoch": 1.9836888331242157,
      "grad_norm": 1.7902740240097046,
      "learning_rate": 1.0917969580734322e-09,
      "loss": 1.0838,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3162
    },
    {
      "epoch": 1.9843161856963614,
      "grad_norm": 1.8456615209579468,
      "learning_rate": 1.0124261311453276e-09,
      "loss": 1.305,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3163
    },
    {
      "epoch": 1.984943538268507,
      "grad_norm": 1.8739662170410156,
      "learning_rate": 9.360496365870552e-10,
      "loss": 1.2175,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3164
    },
    {
      "epoch": 1.9855708908406524,
      "grad_norm": 1.9121259450912476,
      "learning_rate": 8.62667565913744e-10,
      "loss": 1.2805,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3165
    },
    {
      "epoch": 1.986198243412798,
      "grad_norm": 1.4620251655578613,
      "learning_rate": 7.922800070536696e-10,
      "loss": 1.3768,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3166
    },
    {
      "epoch": 1.9868255959849437,
      "grad_norm": 1.821302890777588,
      "learning_rate": 7.248870443460342e-10,
      "loss": 1.2826,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3167
    },
    {
      "epoch": 1.987452948557089,
      "grad_norm": 2.1709654331207275,
      "learning_rate": 6.604887585426323e-10,
      "loss": 1.227,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3168
    },
    {
      "epoch": 1.9880803011292345,
      "grad_norm": 1.6179709434509277,
      "learning_rate": 5.990852268064617e-10,
      "loss": 1.3695,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3169
    },
    {
      "epoch": 1.9887076537013801,
      "grad_norm": 1.6689754724502563,
      "learning_rate": 5.406765227122801e-10,
      "loss": 1.3144,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3170
    },
    {
      "epoch": 1.9893350062735258,
      "grad_norm": 1.9300893545150757,
      "learning_rate": 4.852627162468814e-10,
      "loss": 1.1671,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3171
    },
    {
      "epoch": 1.9899623588456712,
      "grad_norm": 1.872180700302124,
      "learning_rate": 4.3284387380743145e-10,
      "loss": 1.3608,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3172
    },
    {
      "epoch": 1.9905897114178168,
      "grad_norm": 1.7151243686676025,
      "learning_rate": 3.834200582036873e-10,
      "loss": 1.1362,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3173
    },
    {
      "epoch": 1.9912170639899625,
      "grad_norm": 1.7240897417068481,
      "learning_rate": 3.3699132865605554e-10,
      "loss": 1.317,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3174
    },
    {
      "epoch": 1.9918444165621079,
      "grad_norm": 1.8286082744598389,
      "learning_rate": 2.9355774079614653e-10,
      "loss": 1.2865,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3175
    },
    {
      "epoch": 1.9924717691342535,
      "grad_norm": 2.1468513011932373,
      "learning_rate": 2.5311934666705227e-10,
      "loss": 1.136,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3176
    },
    {
      "epoch": 1.9930991217063991,
      "grad_norm": 1.8591892719268799,
      "learning_rate": 2.1567619472279144e-10,
      "loss": 1.3733,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3177
    },
    {
      "epoch": 1.9937264742785445,
      "grad_norm": 1.981756567955017,
      "learning_rate": 1.812283298280315e-10,
      "loss": 1.4058,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3178
    },
    {
      "epoch": 1.99435382685069,
      "grad_norm": 1.5068249702453613,
      "learning_rate": 1.497757932591992e-10,
      "loss": 1.1924,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3179
    },
    {
      "epoch": 1.9949811794228356,
      "grad_norm": 3.0467026233673096,
      "learning_rate": 1.2131862270337025e-10,
      "loss": 1.2362,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3180
    },
    {
      "epoch": 1.9956085319949812,
      "grad_norm": 1.921818494796753,
      "learning_rate": 9.585685225826924e-11,
      "loss": 1.3118,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3181
    },
    {
      "epoch": 1.9962358845671266,
      "grad_norm": 1.6637516021728516,
      "learning_rate": 7.339051243254735e-11,
      "loss": 1.2209,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3182
    },
    {
      "epoch": 1.9968632371392723,
      "grad_norm": 1.7344951629638672,
      "learning_rate": 5.391963014605983e-11,
      "loss": 1.3776,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3183
    },
    {
      "epoch": 1.997490589711418,
      "grad_norm": 1.964633822441101,
      "learning_rate": 3.744422872875575e-11,
      "loss": 1.2294,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3184
    },
    {
      "epoch": 1.9981179422835633,
      "grad_norm": 1.756880521774292,
      "learning_rate": 2.3964327922343377e-11,
      "loss": 1.2063,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3185
    },
    {
      "epoch": 1.998745294855709,
      "grad_norm": 1.5146030187606812,
      "learning_rate": 1.3479943878347279e-11,
      "loss": 1.1933,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3186
    },
    {
      "epoch": 1.9993726474278546,
      "grad_norm": 1.5605469942092896,
      "learning_rate": 5.9910891592185325e-12,
      "loss": 1.2432,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3187
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.724379062652588,
      "learning_rate": 1.497772738334735e-12,
      "loss": 1.1339,
      "memory/device_mem_reserved(gib)": 67.56,
      "memory/max_mem_active(gib)": 62.26,
      "memory/max_mem_allocated(gib)": 61.07,
      "step": 3188
    }
  ],
  "logging_steps": 1,
  "max_steps": 3188,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 797,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.2702356450110027e+20,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}