opus_sft_dpsk_overlap / trainer_state.json
StephYang's picture
Upload opus_sft_dpsk_overlap qwen3-32b full SFT model
3736465 verified
Raw
History Blame Contribute Delete
135 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 780,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0038461538461538464,
"grad_norm": 3.9238964043520346,
"learning_rate": 0.0,
"loss": 0.5171,
"step": 1
},
{
"epoch": 0.007692307692307693,
"grad_norm": 3.865041716718645,
"learning_rate": 2.564102564102564e-07,
"loss": 0.5361,
"step": 2
},
{
"epoch": 0.011538461538461539,
"grad_norm": 3.558480571282935,
"learning_rate": 5.128205128205128e-07,
"loss": 0.5043,
"step": 3
},
{
"epoch": 0.015384615384615385,
"grad_norm": 3.7998614087292806,
"learning_rate": 7.692307692307694e-07,
"loss": 0.5031,
"step": 4
},
{
"epoch": 0.019230769230769232,
"grad_norm": 3.5689428151249936,
"learning_rate": 1.0256410256410257e-06,
"loss": 0.509,
"step": 5
},
{
"epoch": 0.023076923076923078,
"grad_norm": 2.9879599665303824,
"learning_rate": 1.282051282051282e-06,
"loss": 0.481,
"step": 6
},
{
"epoch": 0.026923076923076925,
"grad_norm": 3.2526580242949326,
"learning_rate": 1.5384615384615387e-06,
"loss": 0.4584,
"step": 7
},
{
"epoch": 0.03076923076923077,
"grad_norm": 2.613802020605672,
"learning_rate": 1.794871794871795e-06,
"loss": 0.4903,
"step": 8
},
{
"epoch": 0.03461538461538462,
"grad_norm": 2.601165702909604,
"learning_rate": 2.0512820512820513e-06,
"loss": 0.485,
"step": 9
},
{
"epoch": 0.038461538461538464,
"grad_norm": 1.5763901158285847,
"learning_rate": 2.307692307692308e-06,
"loss": 0.398,
"step": 10
},
{
"epoch": 0.04230769230769231,
"grad_norm": 2.1261184639761153,
"learning_rate": 2.564102564102564e-06,
"loss": 0.4068,
"step": 11
},
{
"epoch": 0.046153846153846156,
"grad_norm": 1.8665392855577536,
"learning_rate": 2.8205128205128207e-06,
"loss": 0.3544,
"step": 12
},
{
"epoch": 0.05,
"grad_norm": 1.318728933171513,
"learning_rate": 3.0769230769230774e-06,
"loss": 0.3043,
"step": 13
},
{
"epoch": 0.05384615384615385,
"grad_norm": 1.132124697037767,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.3178,
"step": 14
},
{
"epoch": 0.057692307692307696,
"grad_norm": 1.1202543530180564,
"learning_rate": 3.58974358974359e-06,
"loss": 0.3407,
"step": 15
},
{
"epoch": 0.06153846153846154,
"grad_norm": 1.291126276825844,
"learning_rate": 3.846153846153847e-06,
"loss": 0.3616,
"step": 16
},
{
"epoch": 0.06538461538461539,
"grad_norm": 1.095814725126234,
"learning_rate": 4.102564102564103e-06,
"loss": 0.3443,
"step": 17
},
{
"epoch": 0.06923076923076923,
"grad_norm": 1.0636216017507398,
"learning_rate": 4.358974358974359e-06,
"loss": 0.3105,
"step": 18
},
{
"epoch": 0.07307692307692308,
"grad_norm": 1.179492077225723,
"learning_rate": 4.615384615384616e-06,
"loss": 0.3157,
"step": 19
},
{
"epoch": 0.07692307692307693,
"grad_norm": 1.2071397233144883,
"learning_rate": 4.871794871794872e-06,
"loss": 0.3196,
"step": 20
},
{
"epoch": 0.08076923076923077,
"grad_norm": 1.0045930089826436,
"learning_rate": 5.128205128205128e-06,
"loss": 0.2872,
"step": 21
},
{
"epoch": 0.08461538461538462,
"grad_norm": 1.274056377267142,
"learning_rate": 5.384615384615385e-06,
"loss": 0.3679,
"step": 22
},
{
"epoch": 0.08846153846153847,
"grad_norm": 1.000453778201061,
"learning_rate": 5.641025641025641e-06,
"loss": 0.323,
"step": 23
},
{
"epoch": 0.09230769230769231,
"grad_norm": 0.8897229070965105,
"learning_rate": 5.897435897435898e-06,
"loss": 0.324,
"step": 24
},
{
"epoch": 0.09615384615384616,
"grad_norm": 0.7541504685432434,
"learning_rate": 6.153846153846155e-06,
"loss": 0.3176,
"step": 25
},
{
"epoch": 0.1,
"grad_norm": 0.7956017971388913,
"learning_rate": 6.410256410256412e-06,
"loss": 0.3318,
"step": 26
},
{
"epoch": 0.10384615384615385,
"grad_norm": 0.922503309020762,
"learning_rate": 6.666666666666667e-06,
"loss": 0.3187,
"step": 27
},
{
"epoch": 0.1076923076923077,
"grad_norm": 0.7532018031576193,
"learning_rate": 6.923076923076923e-06,
"loss": 0.2855,
"step": 28
},
{
"epoch": 0.11153846153846154,
"grad_norm": 0.700631053805919,
"learning_rate": 7.17948717948718e-06,
"loss": 0.2672,
"step": 29
},
{
"epoch": 0.11538461538461539,
"grad_norm": 0.7195542734195851,
"learning_rate": 7.435897435897437e-06,
"loss": 0.2723,
"step": 30
},
{
"epoch": 0.11923076923076924,
"grad_norm": 0.6723433830688036,
"learning_rate": 7.692307692307694e-06,
"loss": 0.2235,
"step": 31
},
{
"epoch": 0.12307692307692308,
"grad_norm": 0.7500173971223075,
"learning_rate": 7.948717948717949e-06,
"loss": 0.2564,
"step": 32
},
{
"epoch": 0.12692307692307692,
"grad_norm": 0.5392107362297897,
"learning_rate": 8.205128205128205e-06,
"loss": 0.2605,
"step": 33
},
{
"epoch": 0.13076923076923078,
"grad_norm": 0.6513320801692216,
"learning_rate": 8.461538461538462e-06,
"loss": 0.2419,
"step": 34
},
{
"epoch": 0.1346153846153846,
"grad_norm": 0.7060740174436259,
"learning_rate": 8.717948717948719e-06,
"loss": 0.2697,
"step": 35
},
{
"epoch": 0.13846153846153847,
"grad_norm": 0.5850462622222398,
"learning_rate": 8.974358974358976e-06,
"loss": 0.2624,
"step": 36
},
{
"epoch": 0.1423076923076923,
"grad_norm": 0.6685505320489104,
"learning_rate": 9.230769230769232e-06,
"loss": 0.2475,
"step": 37
},
{
"epoch": 0.14615384615384616,
"grad_norm": 0.6053524432955536,
"learning_rate": 9.487179487179487e-06,
"loss": 0.2864,
"step": 38
},
{
"epoch": 0.15,
"grad_norm": 0.7251770133590191,
"learning_rate": 9.743589743589744e-06,
"loss": 0.2777,
"step": 39
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.6133734638624586,
"learning_rate": 1e-05,
"loss": 0.2758,
"step": 40
},
{
"epoch": 0.1576923076923077,
"grad_norm": 0.8478498938418025,
"learning_rate": 9.99995506314361e-06,
"loss": 0.2817,
"step": 41
},
{
"epoch": 0.16153846153846155,
"grad_norm": 0.8683596484373798,
"learning_rate": 9.99982025338217e-06,
"loss": 0.2599,
"step": 42
},
{
"epoch": 0.16538461538461538,
"grad_norm": 0.47646754300069805,
"learning_rate": 9.999595573138845e-06,
"loss": 0.2238,
"step": 43
},
{
"epoch": 0.16923076923076924,
"grad_norm": 0.5006834529361883,
"learning_rate": 9.99928102645221e-06,
"loss": 0.3349,
"step": 44
},
{
"epoch": 0.17307692307692307,
"grad_norm": 0.5862107789466642,
"learning_rate": 9.99887661897616e-06,
"loss": 0.3636,
"step": 45
},
{
"epoch": 0.17692307692307693,
"grad_norm": 0.5551999971664277,
"learning_rate": 9.99838235797981e-06,
"loss": 0.2511,
"step": 46
},
{
"epoch": 0.18076923076923077,
"grad_norm": 0.6431444132906169,
"learning_rate": 9.997798252347382e-06,
"loss": 0.2546,
"step": 47
},
{
"epoch": 0.18461538461538463,
"grad_norm": 0.7016538602092144,
"learning_rate": 9.99712431257802e-06,
"loss": 0.2432,
"step": 48
},
{
"epoch": 0.18846153846153846,
"grad_norm": 0.6436268838669431,
"learning_rate": 9.996360550785619e-06,
"loss": 0.2238,
"step": 49
},
{
"epoch": 0.19230769230769232,
"grad_norm": 0.8154568253550334,
"learning_rate": 9.9955069806986e-06,
"loss": 0.1921,
"step": 50
},
{
"epoch": 0.19615384615384615,
"grad_norm": 0.6867504474244611,
"learning_rate": 9.994563617659665e-06,
"loss": 0.3021,
"step": 51
},
{
"epoch": 0.2,
"grad_norm": 0.542407945905818,
"learning_rate": 9.993530478625524e-06,
"loss": 0.2838,
"step": 52
},
{
"epoch": 0.20384615384615384,
"grad_norm": 0.6635448135497594,
"learning_rate": 9.992407582166582e-06,
"loss": 0.396,
"step": 53
},
{
"epoch": 0.2076923076923077,
"grad_norm": 0.6239678658264209,
"learning_rate": 9.991194948466615e-06,
"loss": 0.2537,
"step": 54
},
{
"epoch": 0.21153846153846154,
"grad_norm": 0.572657279292308,
"learning_rate": 9.989892599322404e-06,
"loss": 0.2756,
"step": 55
},
{
"epoch": 0.2153846153846154,
"grad_norm": 0.5046398762256985,
"learning_rate": 9.988500558143337e-06,
"loss": 0.2282,
"step": 56
},
{
"epoch": 0.21923076923076923,
"grad_norm": 0.6106871038894085,
"learning_rate": 9.987018849950996e-06,
"loss": 0.2621,
"step": 57
},
{
"epoch": 0.2230769230769231,
"grad_norm": 0.6667434956957957,
"learning_rate": 9.985447501378706e-06,
"loss": 0.2581,
"step": 58
},
{
"epoch": 0.22692307692307692,
"grad_norm": 0.5707877283877809,
"learning_rate": 9.983786540671052e-06,
"loss": 0.2493,
"step": 59
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.5411517461500795,
"learning_rate": 9.982035997683372e-06,
"loss": 0.2592,
"step": 60
},
{
"epoch": 0.23461538461538461,
"grad_norm": 1.255284755689992,
"learning_rate": 9.980195903881231e-06,
"loss": 0.2601,
"step": 61
},
{
"epoch": 0.23846153846153847,
"grad_norm": 0.5784187044081806,
"learning_rate": 9.978266292339838e-06,
"loss": 0.269,
"step": 62
},
{
"epoch": 0.2423076923076923,
"grad_norm": 0.5251037889321888,
"learning_rate": 9.976247197743465e-06,
"loss": 0.2701,
"step": 63
},
{
"epoch": 0.24615384615384617,
"grad_norm": 0.6700677521444053,
"learning_rate": 9.974138656384815e-06,
"loss": 0.2349,
"step": 64
},
{
"epoch": 0.25,
"grad_norm": 0.5453274857953114,
"learning_rate": 9.97194070616438e-06,
"loss": 0.2592,
"step": 65
},
{
"epoch": 0.25384615384615383,
"grad_norm": 0.4841168618676258,
"learning_rate": 9.969653386589749e-06,
"loss": 0.3129,
"step": 66
},
{
"epoch": 0.25769230769230766,
"grad_norm": 0.6008451041239414,
"learning_rate": 9.967276738774897e-06,
"loss": 0.3239,
"step": 67
},
{
"epoch": 0.26153846153846155,
"grad_norm": 0.5088532515945503,
"learning_rate": 9.964810805439464e-06,
"loss": 0.259,
"step": 68
},
{
"epoch": 0.2653846153846154,
"grad_norm": 0.5054084340402627,
"learning_rate": 9.962255630907964e-06,
"loss": 0.2225,
"step": 69
},
{
"epoch": 0.2692307692307692,
"grad_norm": 0.5368741406906062,
"learning_rate": 9.959611261108999e-06,
"loss": 0.2794,
"step": 70
},
{
"epoch": 0.27307692307692305,
"grad_norm": 0.6421423292934936,
"learning_rate": 9.956877743574437e-06,
"loss": 0.2349,
"step": 71
},
{
"epoch": 0.27692307692307694,
"grad_norm": 0.48170209929410346,
"learning_rate": 9.954055127438554e-06,
"loss": 0.2289,
"step": 72
},
{
"epoch": 0.28076923076923077,
"grad_norm": 0.538709956261295,
"learning_rate": 9.951143463437145e-06,
"loss": 0.2184,
"step": 73
},
{
"epoch": 0.2846153846153846,
"grad_norm": 0.5415562971542437,
"learning_rate": 9.948142803906623e-06,
"loss": 0.2402,
"step": 74
},
{
"epoch": 0.28846153846153844,
"grad_norm": 0.3693718172979592,
"learning_rate": 9.94505320278307e-06,
"loss": 0.1762,
"step": 75
},
{
"epoch": 0.2923076923076923,
"grad_norm": 0.5636531788448021,
"learning_rate": 9.94187471560127e-06,
"loss": 0.2499,
"step": 76
},
{
"epoch": 0.29615384615384616,
"grad_norm": 0.6416879363388094,
"learning_rate": 9.938607399493714e-06,
"loss": 0.2385,
"step": 77
},
{
"epoch": 0.3,
"grad_norm": 0.7496386304718934,
"learning_rate": 9.935251313189564e-06,
"loss": 0.239,
"step": 78
},
{
"epoch": 0.3038461538461538,
"grad_norm": 0.514510908906308,
"learning_rate": 9.931806517013612e-06,
"loss": 0.2853,
"step": 79
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.4740375973334633,
"learning_rate": 9.92827307288518e-06,
"loss": 0.2354,
"step": 80
},
{
"epoch": 0.31153846153846154,
"grad_norm": 0.5591654441743603,
"learning_rate": 9.924651044317017e-06,
"loss": 0.1946,
"step": 81
},
{
"epoch": 0.3153846153846154,
"grad_norm": 0.5061643025705186,
"learning_rate": 9.920940496414153e-06,
"loss": 0.2493,
"step": 82
},
{
"epoch": 0.3192307692307692,
"grad_norm": 0.6818593286674031,
"learning_rate": 9.917141495872733e-06,
"loss": 0.2222,
"step": 83
},
{
"epoch": 0.3230769230769231,
"grad_norm": 0.5218088332503258,
"learning_rate": 9.913254110978812e-06,
"loss": 0.1953,
"step": 84
},
{
"epoch": 0.3269230769230769,
"grad_norm": 0.6356033435216335,
"learning_rate": 9.909278411607134e-06,
"loss": 0.2251,
"step": 85
},
{
"epoch": 0.33076923076923076,
"grad_norm": 0.6624665627199435,
"learning_rate": 9.90521446921987e-06,
"loss": 0.2779,
"step": 86
},
{
"epoch": 0.3346153846153846,
"grad_norm": 0.5410383443932127,
"learning_rate": 9.90106235686534e-06,
"loss": 0.199,
"step": 87
},
{
"epoch": 0.3384615384615385,
"grad_norm": 0.7191836401826213,
"learning_rate": 9.896822149176695e-06,
"loss": 0.1948,
"step": 88
},
{
"epoch": 0.3423076923076923,
"grad_norm": 0.5801324782802036,
"learning_rate": 9.892493922370575e-06,
"loss": 0.2115,
"step": 89
},
{
"epoch": 0.34615384615384615,
"grad_norm": 0.6848540278900523,
"learning_rate": 9.888077754245741e-06,
"loss": 0.2351,
"step": 90
},
{
"epoch": 0.35,
"grad_norm": 0.5994634992164671,
"learning_rate": 9.883573724181683e-06,
"loss": 0.2349,
"step": 91
},
{
"epoch": 0.35384615384615387,
"grad_norm": 0.5718311795464627,
"learning_rate": 9.878981913137178e-06,
"loss": 0.2543,
"step": 92
},
{
"epoch": 0.3576923076923077,
"grad_norm": 0.5084484455832017,
"learning_rate": 9.87430240364885e-06,
"loss": 0.2103,
"step": 93
},
{
"epoch": 0.36153846153846153,
"grad_norm": 0.622962170870338,
"learning_rate": 9.869535279829674e-06,
"loss": 0.2451,
"step": 94
},
{
"epoch": 0.36538461538461536,
"grad_norm": 0.8011484280722398,
"learning_rate": 9.864680627367476e-06,
"loss": 0.2263,
"step": 95
},
{
"epoch": 0.36923076923076925,
"grad_norm": 0.6723527978962419,
"learning_rate": 9.859738533523384e-06,
"loss": 0.2209,
"step": 96
},
{
"epoch": 0.3730769230769231,
"grad_norm": 0.6388659975732571,
"learning_rate": 9.854709087130261e-06,
"loss": 0.2764,
"step": 97
},
{
"epoch": 0.3769230769230769,
"grad_norm": 0.5809471981558344,
"learning_rate": 9.849592378591113e-06,
"loss": 0.2076,
"step": 98
},
{
"epoch": 0.38076923076923075,
"grad_norm": 0.9373923650289692,
"learning_rate": 9.844388499877457e-06,
"loss": 0.2285,
"step": 99
},
{
"epoch": 0.38461538461538464,
"grad_norm": 0.5892783197354109,
"learning_rate": 9.839097544527674e-06,
"loss": 0.3147,
"step": 100
},
{
"epoch": 0.38846153846153847,
"grad_norm": 0.5229355872601981,
"learning_rate": 9.833719607645325e-06,
"loss": 0.236,
"step": 101
},
{
"epoch": 0.3923076923076923,
"grad_norm": 0.5496958778497968,
"learning_rate": 9.82825478589744e-06,
"loss": 0.2259,
"step": 102
},
{
"epoch": 0.39615384615384613,
"grad_norm": 0.4997807501259826,
"learning_rate": 9.822703177512783e-06,
"loss": 0.2772,
"step": 103
},
{
"epoch": 0.4,
"grad_norm": 0.48130314935858876,
"learning_rate": 9.817064882280085e-06,
"loss": 0.2112,
"step": 104
},
{
"epoch": 0.40384615384615385,
"grad_norm": 0.5611229612776816,
"learning_rate": 9.811340001546252e-06,
"loss": 0.2779,
"step": 105
},
{
"epoch": 0.4076923076923077,
"grad_norm": 0.5121078157860088,
"learning_rate": 9.805528638214543e-06,
"loss": 0.1909,
"step": 106
},
{
"epoch": 0.4115384615384615,
"grad_norm": 0.43303744635314717,
"learning_rate": 9.799630896742716e-06,
"loss": 0.2063,
"step": 107
},
{
"epoch": 0.4153846153846154,
"grad_norm": 0.5309729488848974,
"learning_rate": 9.793646883141155e-06,
"loss": 0.2502,
"step": 108
},
{
"epoch": 0.41923076923076924,
"grad_norm": 0.504019838761931,
"learning_rate": 9.787576704970965e-06,
"loss": 0.2684,
"step": 109
},
{
"epoch": 0.4230769230769231,
"grad_norm": 0.5188716601426719,
"learning_rate": 9.781420471342035e-06,
"loss": 0.2212,
"step": 110
},
{
"epoch": 0.4269230769230769,
"grad_norm": 0.5676741564045472,
"learning_rate": 9.77517829291108e-06,
"loss": 0.2683,
"step": 111
},
{
"epoch": 0.4307692307692308,
"grad_norm": 0.6653028482418869,
"learning_rate": 9.768850281879651e-06,
"loss": 0.2221,
"step": 112
},
{
"epoch": 0.4346153846153846,
"grad_norm": 0.6108832847466867,
"learning_rate": 9.762436551992117e-06,
"loss": 0.2242,
"step": 113
},
{
"epoch": 0.43846153846153846,
"grad_norm": 0.5357996955018922,
"learning_rate": 9.755937218533622e-06,
"loss": 0.1966,
"step": 114
},
{
"epoch": 0.4423076923076923,
"grad_norm": 0.46061799718993846,
"learning_rate": 9.74935239832801e-06,
"loss": 0.2533,
"step": 115
},
{
"epoch": 0.4461538461538462,
"grad_norm": 0.47312594355675086,
"learning_rate": 9.742682209735727e-06,
"loss": 0.2282,
"step": 116
},
{
"epoch": 0.45,
"grad_norm": 0.5626675677514869,
"learning_rate": 9.735926772651703e-06,
"loss": 0.2181,
"step": 117
},
{
"epoch": 0.45384615384615384,
"grad_norm": 0.6723379929283086,
"learning_rate": 9.729086208503174e-06,
"loss": 0.2343,
"step": 118
},
{
"epoch": 0.4576923076923077,
"grad_norm": 0.6323186080541812,
"learning_rate": 9.722160640247523e-06,
"loss": 0.2355,
"step": 119
},
{
"epoch": 0.46153846153846156,
"grad_norm": 0.6602878154578612,
"learning_rate": 9.715150192370054e-06,
"loss": 0.2388,
"step": 120
},
{
"epoch": 0.4653846153846154,
"grad_norm": 0.577240989795038,
"learning_rate": 9.708054990881763e-06,
"loss": 0.2737,
"step": 121
},
{
"epoch": 0.46923076923076923,
"grad_norm": 0.5400295584803867,
"learning_rate": 9.700875163317072e-06,
"loss": 0.2841,
"step": 122
},
{
"epoch": 0.47307692307692306,
"grad_norm": 0.5503579009887059,
"learning_rate": 9.693610838731532e-06,
"loss": 0.2369,
"step": 123
},
{
"epoch": 0.47692307692307695,
"grad_norm": 0.5551607237688418,
"learning_rate": 9.686262147699507e-06,
"loss": 0.2566,
"step": 124
},
{
"epoch": 0.4807692307692308,
"grad_norm": 0.45715780432864744,
"learning_rate": 9.678829222311827e-06,
"loss": 0.1868,
"step": 125
},
{
"epoch": 0.4846153846153846,
"grad_norm": 0.5743006310889183,
"learning_rate": 9.671312196173413e-06,
"loss": 0.2235,
"step": 126
},
{
"epoch": 0.48846153846153845,
"grad_norm": 0.5547219023660661,
"learning_rate": 9.663711204400872e-06,
"loss": 0.2359,
"step": 127
},
{
"epoch": 0.49230769230769234,
"grad_norm": 0.5404451860104633,
"learning_rate": 9.656026383620076e-06,
"loss": 0.2585,
"step": 128
},
{
"epoch": 0.49615384615384617,
"grad_norm": 0.5141545423609298,
"learning_rate": 9.6482578719637e-06,
"loss": 0.2207,
"step": 129
},
{
"epoch": 0.5,
"grad_norm": 0.5819489421942832,
"learning_rate": 9.640405809068743e-06,
"loss": 0.185,
"step": 130
},
{
"epoch": 0.5038461538461538,
"grad_norm": 0.5453700561610013,
"learning_rate": 9.632470336074009e-06,
"loss": 0.2538,
"step": 131
},
{
"epoch": 0.5076923076923077,
"grad_norm": 0.5268074098292405,
"learning_rate": 9.624451595617588e-06,
"loss": 0.2206,
"step": 132
},
{
"epoch": 0.5115384615384615,
"grad_norm": 0.5151242484270426,
"learning_rate": 9.616349731834271e-06,
"loss": 0.1973,
"step": 133
},
{
"epoch": 0.5153846153846153,
"grad_norm": 0.7186109482767982,
"learning_rate": 9.608164890352977e-06,
"loss": 0.2197,
"step": 134
},
{
"epoch": 0.5192307692307693,
"grad_norm": 0.5069905518609401,
"learning_rate": 9.599897218294122e-06,
"loss": 0.1997,
"step": 135
},
{
"epoch": 0.5230769230769231,
"grad_norm": 0.5281677650240848,
"learning_rate": 9.591546864266983e-06,
"loss": 0.2114,
"step": 136
},
{
"epoch": 0.5269230769230769,
"grad_norm": 0.6279171503407216,
"learning_rate": 9.583113978367026e-06,
"loss": 0.2347,
"step": 137
},
{
"epoch": 0.5307692307692308,
"grad_norm": 0.502812100216165,
"learning_rate": 9.574598712173202e-06,
"loss": 0.2121,
"step": 138
},
{
"epoch": 0.5346153846153846,
"grad_norm": 0.46388349097977605,
"learning_rate": 9.56600121874523e-06,
"loss": 0.2697,
"step": 139
},
{
"epoch": 0.5384615384615384,
"grad_norm": 0.8558413327232558,
"learning_rate": 9.557321652620839e-06,
"loss": 0.1773,
"step": 140
},
{
"epoch": 0.5423076923076923,
"grad_norm": 0.985345248669586,
"learning_rate": 9.548560169812997e-06,
"loss": 0.2561,
"step": 141
},
{
"epoch": 0.5461538461538461,
"grad_norm": 0.7524493863035334,
"learning_rate": 9.539716927807102e-06,
"loss": 0.2464,
"step": 142
},
{
"epoch": 0.55,
"grad_norm": 0.5111825715315346,
"learning_rate": 9.530792085558151e-06,
"loss": 0.232,
"step": 143
},
{
"epoch": 0.5538461538461539,
"grad_norm": 0.5056580001614729,
"learning_rate": 9.521785803487888e-06,
"loss": 0.1731,
"step": 144
},
{
"epoch": 0.5576923076923077,
"grad_norm": 1.2585642666961603,
"learning_rate": 9.512698243481914e-06,
"loss": 0.2044,
"step": 145
},
{
"epoch": 0.5615384615384615,
"grad_norm": 0.7913413567384371,
"learning_rate": 9.50352956888678e-06,
"loss": 0.231,
"step": 146
},
{
"epoch": 0.5653846153846154,
"grad_norm": 0.45252324024199986,
"learning_rate": 9.49427994450705e-06,
"loss": 0.205,
"step": 147
},
{
"epoch": 0.5692307692307692,
"grad_norm": 0.5542049816931011,
"learning_rate": 9.484949536602343e-06,
"loss": 0.2366,
"step": 148
},
{
"epoch": 0.573076923076923,
"grad_norm": 0.6233470226885651,
"learning_rate": 9.47553851288434e-06,
"loss": 0.2538,
"step": 149
},
{
"epoch": 0.5769230769230769,
"grad_norm": 0.8311270070122355,
"learning_rate": 9.466047042513767e-06,
"loss": 0.2817,
"step": 150
},
{
"epoch": 0.5807692307692308,
"grad_norm": 0.4129513701760548,
"learning_rate": 9.45647529609736e-06,
"loss": 0.2203,
"step": 151
},
{
"epoch": 0.5846153846153846,
"grad_norm": 0.7219385594298195,
"learning_rate": 9.4468234456848e-06,
"loss": 0.3649,
"step": 152
},
{
"epoch": 0.5884615384615385,
"grad_norm": 0.5242222250849599,
"learning_rate": 9.437091664765611e-06,
"loss": 0.2133,
"step": 153
},
{
"epoch": 0.5923076923076923,
"grad_norm": 0.5153401533175287,
"learning_rate": 9.427280128266049e-06,
"loss": 0.2033,
"step": 154
},
{
"epoch": 0.5961538461538461,
"grad_norm": 0.5588229385879334,
"learning_rate": 9.41738901254596e-06,
"loss": 0.2091,
"step": 155
},
{
"epoch": 0.6,
"grad_norm": 0.4330400405855999,
"learning_rate": 9.4074184953956e-06,
"loss": 0.2198,
"step": 156
},
{
"epoch": 0.6038461538461538,
"grad_norm": 0.48136666750584844,
"learning_rate": 9.397368756032445e-06,
"loss": 0.2549,
"step": 157
},
{
"epoch": 0.6076923076923076,
"grad_norm": 0.5491403765871278,
"learning_rate": 9.38723997509798e-06,
"loss": 0.2231,
"step": 158
},
{
"epoch": 0.6115384615384616,
"grad_norm": 0.47159123955445276,
"learning_rate": 9.37703233465443e-06,
"loss": 0.2012,
"step": 159
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.4707541560308235,
"learning_rate": 9.366746018181503e-06,
"loss": 0.2695,
"step": 160
},
{
"epoch": 0.6192307692307693,
"grad_norm": 0.4448318545569148,
"learning_rate": 9.356381210573092e-06,
"loss": 0.2457,
"step": 161
},
{
"epoch": 0.6230769230769231,
"grad_norm": 0.4916189237805617,
"learning_rate": 9.345938098133946e-06,
"loss": 0.2972,
"step": 162
},
{
"epoch": 0.6269230769230769,
"grad_norm": 0.4965526632348685,
"learning_rate": 9.33541686857632e-06,
"loss": 0.2265,
"step": 163
},
{
"epoch": 0.6307692307692307,
"grad_norm": 0.42646023358197266,
"learning_rate": 9.324817711016609e-06,
"loss": 0.2067,
"step": 164
},
{
"epoch": 0.6346153846153846,
"grad_norm": 0.44406548353079006,
"learning_rate": 9.31414081597194e-06,
"loss": 0.1979,
"step": 165
},
{
"epoch": 0.6384615384615384,
"grad_norm": 0.5002862988710794,
"learning_rate": 9.303386375356752e-06,
"loss": 0.1812,
"step": 166
},
{
"epoch": 0.6423076923076924,
"grad_norm": 0.47880788650797806,
"learning_rate": 9.292554582479349e-06,
"loss": 0.2277,
"step": 167
},
{
"epoch": 0.6461538461538462,
"grad_norm": 0.6015012859011786,
"learning_rate": 9.281645632038417e-06,
"loss": 0.2246,
"step": 168
},
{
"epoch": 0.65,
"grad_norm": 0.5267702738321232,
"learning_rate": 9.270659720119533e-06,
"loss": 0.2482,
"step": 169
},
{
"epoch": 0.6538461538461539,
"grad_norm": 0.4208996980678987,
"learning_rate": 9.259597044191635e-06,
"loss": 0.1671,
"step": 170
},
{
"epoch": 0.6576923076923077,
"grad_norm": 0.5199810465911215,
"learning_rate": 9.248457803103476e-06,
"loss": 0.2239,
"step": 171
},
{
"epoch": 0.6615384615384615,
"grad_norm": 0.4091312438100926,
"learning_rate": 9.237242197080045e-06,
"loss": 0.2211,
"step": 172
},
{
"epoch": 0.6653846153846154,
"grad_norm": 0.4987472970221044,
"learning_rate": 9.225950427718974e-06,
"loss": 0.1747,
"step": 173
},
{
"epoch": 0.6692307692307692,
"grad_norm": 0.4695095646959319,
"learning_rate": 9.21458269798691e-06,
"loss": 0.2873,
"step": 174
},
{
"epoch": 0.6730769230769231,
"grad_norm": 0.4916670929902003,
"learning_rate": 9.203139212215868e-06,
"loss": 0.2145,
"step": 175
},
{
"epoch": 0.676923076923077,
"grad_norm": 0.4452053491268018,
"learning_rate": 9.191620176099559e-06,
"loss": 0.1928,
"step": 176
},
{
"epoch": 0.6807692307692308,
"grad_norm": 0.4678102288496444,
"learning_rate": 9.180025796689692e-06,
"loss": 0.2179,
"step": 177
},
{
"epoch": 0.6846153846153846,
"grad_norm": 0.5771093006117648,
"learning_rate": 9.168356282392253e-06,
"loss": 0.2469,
"step": 178
},
{
"epoch": 0.6884615384615385,
"grad_norm": 0.48676646503712,
"learning_rate": 9.156611842963753e-06,
"loss": 0.2007,
"step": 179
},
{
"epoch": 0.6923076923076923,
"grad_norm": 1.0169765647926068,
"learning_rate": 9.144792689507471e-06,
"loss": 0.1827,
"step": 180
},
{
"epoch": 0.6961538461538461,
"grad_norm": 0.4326673713362419,
"learning_rate": 9.132899034469648e-06,
"loss": 0.3295,
"step": 181
},
{
"epoch": 0.7,
"grad_norm": 0.44765902162208127,
"learning_rate": 9.120931091635669e-06,
"loss": 0.2724,
"step": 182
},
{
"epoch": 0.7038461538461539,
"grad_norm": 0.44552301472500505,
"learning_rate": 9.108889076126226e-06,
"loss": 0.1967,
"step": 183
},
{
"epoch": 0.7076923076923077,
"grad_norm": 0.5133885952136333,
"learning_rate": 9.09677320439345e-06,
"loss": 0.2821,
"step": 184
},
{
"epoch": 0.7115384615384616,
"grad_norm": 0.47734530904630457,
"learning_rate": 9.084583694217012e-06,
"loss": 0.1839,
"step": 185
},
{
"epoch": 0.7153846153846154,
"grad_norm": 0.4667530787349204,
"learning_rate": 9.072320764700223e-06,
"loss": 0.2002,
"step": 186
},
{
"epoch": 0.7192307692307692,
"grad_norm": 0.5511264226614181,
"learning_rate": 9.059984636266082e-06,
"loss": 0.2684,
"step": 187
},
{
"epoch": 0.7230769230769231,
"grad_norm": 0.5610201978781818,
"learning_rate": 9.047575530653324e-06,
"loss": 0.2172,
"step": 188
},
{
"epoch": 0.7269230769230769,
"grad_norm": 0.3968810105283452,
"learning_rate": 9.035093670912424e-06,
"loss": 0.2404,
"step": 189
},
{
"epoch": 0.7307692307692307,
"grad_norm": 0.6810685432042747,
"learning_rate": 9.022539281401601e-06,
"loss": 0.2426,
"step": 190
},
{
"epoch": 0.7346153846153847,
"grad_norm": 0.4112909424321068,
"learning_rate": 9.009912587782772e-06,
"loss": 0.2368,
"step": 191
},
{
"epoch": 0.7384615384615385,
"grad_norm": 0.45870323572761434,
"learning_rate": 8.997213817017508e-06,
"loss": 0.2221,
"step": 192
},
{
"epoch": 0.7423076923076923,
"grad_norm": 0.5783674430771019,
"learning_rate": 8.984443197362938e-06,
"loss": 0.2598,
"step": 193
},
{
"epoch": 0.7461538461538462,
"grad_norm": 0.3795442220283806,
"learning_rate": 8.971600958367668e-06,
"loss": 0.1971,
"step": 194
},
{
"epoch": 0.75,
"grad_norm": 0.42112062018023,
"learning_rate": 8.958687330867634e-06,
"loss": 0.2286,
"step": 195
},
{
"epoch": 0.7538461538461538,
"grad_norm": 0.6615339096877597,
"learning_rate": 8.94570254698197e-06,
"loss": 0.2364,
"step": 196
},
{
"epoch": 0.7576923076923077,
"grad_norm": 0.6200718199334362,
"learning_rate": 8.932646840108818e-06,
"loss": 0.2353,
"step": 197
},
{
"epoch": 0.7615384615384615,
"grad_norm": 0.506279015102114,
"learning_rate": 8.919520444921153e-06,
"loss": 0.2329,
"step": 198
},
{
"epoch": 0.7653846153846153,
"grad_norm": 0.5903129259902322,
"learning_rate": 8.906323597362547e-06,
"loss": 0.1961,
"step": 199
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.4346371872564603,
"learning_rate": 8.893056534642938e-06,
"loss": 0.247,
"step": 200
},
{
"epoch": 0.7730769230769231,
"grad_norm": 0.4972341789154317,
"learning_rate": 8.879719495234363e-06,
"loss": 0.1995,
"step": 201
},
{
"epoch": 0.7769230769230769,
"grad_norm": 0.6134809334375436,
"learning_rate": 8.866312718866669e-06,
"loss": 0.2421,
"step": 202
},
{
"epoch": 0.7807692307692308,
"grad_norm": 0.46117265343724284,
"learning_rate": 8.852836446523213e-06,
"loss": 0.2599,
"step": 203
},
{
"epoch": 0.7846153846153846,
"grad_norm": 0.3934200135027767,
"learning_rate": 8.83929092043652e-06,
"loss": 0.236,
"step": 204
},
{
"epoch": 0.7884615384615384,
"grad_norm": 0.49922768199565765,
"learning_rate": 8.825676384083936e-06,
"loss": 0.1982,
"step": 205
},
{
"epoch": 0.7923076923076923,
"grad_norm": 0.49668967819118254,
"learning_rate": 8.811993082183243e-06,
"loss": 0.2354,
"step": 206
},
{
"epoch": 0.7961538461538461,
"grad_norm": 0.46619350224576994,
"learning_rate": 8.798241260688273e-06,
"loss": 0.2228,
"step": 207
},
{
"epoch": 0.8,
"grad_norm": 0.4939191727625512,
"learning_rate": 8.784421166784476e-06,
"loss": 0.211,
"step": 208
},
{
"epoch": 0.8038461538461539,
"grad_norm": 0.479154668275832,
"learning_rate": 8.770533048884483e-06,
"loss": 0.2287,
"step": 209
},
{
"epoch": 0.8076923076923077,
"grad_norm": 0.8567918551030094,
"learning_rate": 8.756577156623636e-06,
"loss": 0.2178,
"step": 210
},
{
"epoch": 0.8115384615384615,
"grad_norm": 1.212916158387493,
"learning_rate": 8.742553740855507e-06,
"loss": 0.2222,
"step": 211
},
{
"epoch": 0.8153846153846154,
"grad_norm": 1.1009596903577805,
"learning_rate": 8.728463053647382e-06,
"loss": 0.2376,
"step": 212
},
{
"epoch": 0.8192307692307692,
"grad_norm": 0.4096144715960063,
"learning_rate": 8.71430534827574e-06,
"loss": 0.2019,
"step": 213
},
{
"epoch": 0.823076923076923,
"grad_norm": 0.4018141193635861,
"learning_rate": 8.700080879221689e-06,
"loss": 0.2223,
"step": 214
},
{
"epoch": 0.8269230769230769,
"grad_norm": 0.40333375504436675,
"learning_rate": 8.685789902166395e-06,
"loss": 0.2402,
"step": 215
},
{
"epoch": 0.8307692307692308,
"grad_norm": 0.42357746203750907,
"learning_rate": 8.671432673986493e-06,
"loss": 0.2472,
"step": 216
},
{
"epoch": 0.8346153846153846,
"grad_norm": 0.44559651148780177,
"learning_rate": 8.657009452749466e-06,
"loss": 0.2645,
"step": 217
},
{
"epoch": 0.8384615384615385,
"grad_norm": 0.519415713850819,
"learning_rate": 8.642520497709001e-06,
"loss": 0.1613,
"step": 218
},
{
"epoch": 0.8423076923076923,
"grad_norm": 0.3196349047621588,
"learning_rate": 8.627966069300332e-06,
"loss": 0.1721,
"step": 219
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.4371434957471117,
"learning_rate": 8.613346429135567e-06,
"loss": 0.1922,
"step": 220
},
{
"epoch": 0.85,
"grad_norm": 0.49670428831216723,
"learning_rate": 8.598661839998972e-06,
"loss": 0.2369,
"step": 221
},
{
"epoch": 0.8538461538461538,
"grad_norm": 0.47011699524494455,
"learning_rate": 8.583912565842258e-06,
"loss": 0.2568,
"step": 222
},
{
"epoch": 0.8576923076923076,
"grad_norm": 0.46388843220602094,
"learning_rate": 8.569098871779828e-06,
"loss": 0.2347,
"step": 223
},
{
"epoch": 0.8615384615384616,
"grad_norm": 0.678321603299092,
"learning_rate": 8.554221024084019e-06,
"loss": 0.244,
"step": 224
},
{
"epoch": 0.8653846153846154,
"grad_norm": 0.43429213869514505,
"learning_rate": 8.539279290180315e-06,
"loss": 0.2213,
"step": 225
},
{
"epoch": 0.8692307692307693,
"grad_norm": 0.4396837172284673,
"learning_rate": 8.524273938642539e-06,
"loss": 0.1862,
"step": 226
},
{
"epoch": 0.8730769230769231,
"grad_norm": 0.44344780892992325,
"learning_rate": 8.509205239188017e-06,
"loss": 0.1938,
"step": 227
},
{
"epoch": 0.8769230769230769,
"grad_norm": 0.5965215691655824,
"learning_rate": 8.494073462672743e-06,
"loss": 0.183,
"step": 228
},
{
"epoch": 0.8807692307692307,
"grad_norm": 0.49337040503318447,
"learning_rate": 8.478878881086505e-06,
"loss": 0.1882,
"step": 229
},
{
"epoch": 0.8846153846153846,
"grad_norm": 0.48920987886268164,
"learning_rate": 8.463621767547998e-06,
"loss": 0.2318,
"step": 230
},
{
"epoch": 0.8884615384615384,
"grad_norm": 0.4152161564123169,
"learning_rate": 8.448302396299906e-06,
"loss": 0.2224,
"step": 231
},
{
"epoch": 0.8923076923076924,
"grad_norm": 0.5530919470650333,
"learning_rate": 8.432921042703985e-06,
"loss": 0.2375,
"step": 232
},
{
"epoch": 0.8961538461538462,
"grad_norm": 0.5317664844197529,
"learning_rate": 8.417477983236107e-06,
"loss": 0.2579,
"step": 233
},
{
"epoch": 0.9,
"grad_norm": 0.489762176370703,
"learning_rate": 8.401973495481289e-06,
"loss": 0.2488,
"step": 234
},
{
"epoch": 0.9038461538461539,
"grad_norm": 0.35713820607995794,
"learning_rate": 8.386407858128707e-06,
"loss": 0.1656,
"step": 235
},
{
"epoch": 0.9076923076923077,
"grad_norm": 0.49313248662552256,
"learning_rate": 8.370781350966683e-06,
"loss": 0.2361,
"step": 236
},
{
"epoch": 0.9115384615384615,
"grad_norm": 0.5003814400874651,
"learning_rate": 8.355094254877665e-06,
"loss": 0.2239,
"step": 237
},
{
"epoch": 0.9153846153846154,
"grad_norm": 0.43214736421308636,
"learning_rate": 8.339346851833163e-06,
"loss": 0.2469,
"step": 238
},
{
"epoch": 0.9192307692307692,
"grad_norm": 0.5631615370198494,
"learning_rate": 8.323539424888695e-06,
"loss": 0.2308,
"step": 239
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.48058073108351584,
"learning_rate": 8.30767225817869e-06,
"loss": 0.2211,
"step": 240
},
{
"epoch": 0.926923076923077,
"grad_norm": 0.48550833774790864,
"learning_rate": 8.291745636911382e-06,
"loss": 0.2373,
"step": 241
},
{
"epoch": 0.9307692307692308,
"grad_norm": 0.40886932771514545,
"learning_rate": 8.27575984736369e-06,
"loss": 0.2717,
"step": 242
},
{
"epoch": 0.9346153846153846,
"grad_norm": 0.48366391436080175,
"learning_rate": 8.259715176876069e-06,
"loss": 0.2335,
"step": 243
},
{
"epoch": 0.9384615384615385,
"grad_norm": 0.578973582902094,
"learning_rate": 8.243611913847337e-06,
"loss": 0.2593,
"step": 244
},
{
"epoch": 0.9423076923076923,
"grad_norm": 0.4806012108443053,
"learning_rate": 8.2274503477295e-06,
"loss": 0.2085,
"step": 245
},
{
"epoch": 0.9461538461538461,
"grad_norm": 0.5074347022144132,
"learning_rate": 8.211230769022552e-06,
"loss": 0.2102,
"step": 246
},
{
"epoch": 0.95,
"grad_norm": 0.48359156969936595,
"learning_rate": 8.19495346926924e-06,
"loss": 0.2667,
"step": 247
},
{
"epoch": 0.9538461538461539,
"grad_norm": 0.49161907325464627,
"learning_rate": 8.178618741049841e-06,
"loss": 0.2257,
"step": 248
},
{
"epoch": 0.9576923076923077,
"grad_norm": 0.44631232846642616,
"learning_rate": 8.162226877976886e-06,
"loss": 0.225,
"step": 249
},
{
"epoch": 0.9615384615384616,
"grad_norm": 0.5217563418894521,
"learning_rate": 8.145778174689897e-06,
"loss": 0.2232,
"step": 250
},
{
"epoch": 0.9653846153846154,
"grad_norm": 0.4542249463253952,
"learning_rate": 8.129272926850079e-06,
"loss": 0.221,
"step": 251
},
{
"epoch": 0.9692307692307692,
"grad_norm": 0.48652365861017294,
"learning_rate": 8.112711431135014e-06,
"loss": 0.2047,
"step": 252
},
{
"epoch": 0.9730769230769231,
"grad_norm": 0.42380064512234344,
"learning_rate": 8.096093985233323e-06,
"loss": 0.2437,
"step": 253
},
{
"epoch": 0.9769230769230769,
"grad_norm": 0.5277564327944613,
"learning_rate": 8.079420887839316e-06,
"loss": 0.219,
"step": 254
},
{
"epoch": 0.9807692307692307,
"grad_norm": 0.44095526644873456,
"learning_rate": 8.062692438647628e-06,
"loss": 0.27,
"step": 255
},
{
"epoch": 0.9846153846153847,
"grad_norm": 0.49736917778282513,
"learning_rate": 8.045908938347828e-06,
"loss": 0.192,
"step": 256
},
{
"epoch": 0.9884615384615385,
"grad_norm": 0.4833705267970827,
"learning_rate": 8.029070688619013e-06,
"loss": 0.2102,
"step": 257
},
{
"epoch": 0.9923076923076923,
"grad_norm": 0.43294329075993493,
"learning_rate": 8.012177992124385e-06,
"loss": 0.1661,
"step": 258
},
{
"epoch": 0.9961538461538462,
"grad_norm": 0.5385777287068039,
"learning_rate": 7.995231152505815e-06,
"loss": 0.2171,
"step": 259
},
{
"epoch": 1.0,
"grad_norm": 0.5320402254627459,
"learning_rate": 7.978230474378383e-06,
"loss": 0.2135,
"step": 260
},
{
"epoch": 1.0038461538461538,
"grad_norm": 0.5071165995323762,
"learning_rate": 7.961176263324902e-06,
"loss": 0.1569,
"step": 261
},
{
"epoch": 1.0076923076923077,
"grad_norm": 0.4226589341358406,
"learning_rate": 7.944068825890424e-06,
"loss": 0.1232,
"step": 262
},
{
"epoch": 1.0115384615384615,
"grad_norm": 0.48626949679263576,
"learning_rate": 7.92690846957673e-06,
"loss": 0.1356,
"step": 263
},
{
"epoch": 1.0153846153846153,
"grad_norm": 0.41094635173855426,
"learning_rate": 7.909695502836814e-06,
"loss": 0.1676,
"step": 264
},
{
"epoch": 1.0192307692307692,
"grad_norm": 0.3792296822422694,
"learning_rate": 7.892430235069317e-06,
"loss": 0.1584,
"step": 265
},
{
"epoch": 1.023076923076923,
"grad_norm": 0.49140873552207925,
"learning_rate": 7.875112976612984e-06,
"loss": 0.1597,
"step": 266
},
{
"epoch": 1.0269230769230768,
"grad_norm": 0.4733283629024343,
"learning_rate": 7.857744038741076e-06,
"loss": 0.1154,
"step": 267
},
{
"epoch": 1.0307692307692307,
"grad_norm": 0.41654646486528935,
"learning_rate": 7.84032373365578e-06,
"loss": 0.129,
"step": 268
},
{
"epoch": 1.0346153846153847,
"grad_norm": 0.5044093576664891,
"learning_rate": 7.822852374482597e-06,
"loss": 0.1501,
"step": 269
},
{
"epoch": 1.0384615384615385,
"grad_norm": 0.5702977703791314,
"learning_rate": 7.805330275264707e-06,
"loss": 0.1213,
"step": 270
},
{
"epoch": 1.0423076923076924,
"grad_norm": 0.55240421409857,
"learning_rate": 7.787757750957335e-06,
"loss": 0.1299,
"step": 271
},
{
"epoch": 1.0461538461538462,
"grad_norm": 0.5415442577279818,
"learning_rate": 7.77013511742208e-06,
"loss": 0.1818,
"step": 272
},
{
"epoch": 1.05,
"grad_norm": 0.48874571723181204,
"learning_rate": 7.752462691421245e-06,
"loss": 0.1508,
"step": 273
},
{
"epoch": 1.0538461538461539,
"grad_norm": 0.6703304556936007,
"learning_rate": 7.734740790612137e-06,
"loss": 0.1307,
"step": 274
},
{
"epoch": 1.0576923076923077,
"grad_norm": 0.4381207581021234,
"learning_rate": 7.716969733541357e-06,
"loss": 0.1274,
"step": 275
},
{
"epoch": 1.0615384615384615,
"grad_norm": 0.49204657458030604,
"learning_rate": 7.699149839639086e-06,
"loss": 0.1229,
"step": 276
},
{
"epoch": 1.0653846153846154,
"grad_norm": 0.42262544446720934,
"learning_rate": 7.681281429213328e-06,
"loss": 0.1396,
"step": 277
},
{
"epoch": 1.0692307692307692,
"grad_norm": 0.4381876325594343,
"learning_rate": 7.663364823444157e-06,
"loss": 0.1298,
"step": 278
},
{
"epoch": 1.073076923076923,
"grad_norm": 0.9355501271207461,
"learning_rate": 7.645400344377953e-06,
"loss": 0.1131,
"step": 279
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.8674112741381289,
"learning_rate": 7.627388314921602e-06,
"loss": 0.1295,
"step": 280
},
{
"epoch": 1.0807692307692307,
"grad_norm": 0.6797718943602417,
"learning_rate": 7.609329058836694e-06,
"loss": 0.2985,
"step": 281
},
{
"epoch": 1.0846153846153845,
"grad_norm": 0.45317590070910924,
"learning_rate": 7.59122290073371e-06,
"loss": 0.1434,
"step": 282
},
{
"epoch": 1.0884615384615384,
"grad_norm": 0.43649632201037647,
"learning_rate": 7.5730701660661795e-06,
"loss": 0.1299,
"step": 283
},
{
"epoch": 1.0923076923076924,
"grad_norm": 0.500871531536912,
"learning_rate": 7.554871181124836e-06,
"loss": 0.1372,
"step": 284
},
{
"epoch": 1.0961538461538463,
"grad_norm": 0.5862293449382304,
"learning_rate": 7.536626273031747e-06,
"loss": 0.1001,
"step": 285
},
{
"epoch": 1.1,
"grad_norm": 0.4170632530454472,
"learning_rate": 7.5183357697344395e-06,
"loss": 0.1346,
"step": 286
},
{
"epoch": 1.103846153846154,
"grad_norm": 0.5112039733503784,
"learning_rate": 7.500000000000001e-06,
"loss": 0.1394,
"step": 287
},
{
"epoch": 1.1076923076923078,
"grad_norm": 0.45648359391075244,
"learning_rate": 7.481619293409173e-06,
"loss": 0.1539,
"step": 288
},
{
"epoch": 1.1115384615384616,
"grad_norm": 0.46585894477174705,
"learning_rate": 7.4631939803504215e-06,
"loss": 0.1794,
"step": 289
},
{
"epoch": 1.1153846153846154,
"grad_norm": 0.49871760564039663,
"learning_rate": 7.44472439201401e-06,
"loss": 0.1237,
"step": 290
},
{
"epoch": 1.1192307692307693,
"grad_norm": 0.553612166406509,
"learning_rate": 7.426210860386032e-06,
"loss": 0.1643,
"step": 291
},
{
"epoch": 1.123076923076923,
"grad_norm": 0.4835393537530187,
"learning_rate": 7.407653718242449e-06,
"loss": 0.1366,
"step": 292
},
{
"epoch": 1.126923076923077,
"grad_norm": 0.4744254362776649,
"learning_rate": 7.3890532991431174e-06,
"loss": 0.1124,
"step": 293
},
{
"epoch": 1.1307692307692307,
"grad_norm": 0.47852373260076353,
"learning_rate": 7.370409937425781e-06,
"loss": 0.1328,
"step": 294
},
{
"epoch": 1.1346153846153846,
"grad_norm": 0.4801549458623991,
"learning_rate": 7.3517239682000675e-06,
"loss": 0.1342,
"step": 295
},
{
"epoch": 1.1384615384615384,
"grad_norm": 0.45804435584425557,
"learning_rate": 7.332995727341462e-06,
"loss": 0.092,
"step": 296
},
{
"epoch": 1.1423076923076922,
"grad_norm": 0.35402305936936,
"learning_rate": 7.314225551485273e-06,
"loss": 0.112,
"step": 297
},
{
"epoch": 1.146153846153846,
"grad_norm": 0.487832828265645,
"learning_rate": 7.295413778020579e-06,
"loss": 0.1648,
"step": 298
},
{
"epoch": 1.15,
"grad_norm": 0.5169037119430312,
"learning_rate": 7.276560745084167e-06,
"loss": 0.1344,
"step": 299
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.45466619053860025,
"learning_rate": 7.257666791554448e-06,
"loss": 0.1416,
"step": 300
},
{
"epoch": 1.1576923076923076,
"grad_norm": 0.5782640292313681,
"learning_rate": 7.2387322570453724e-06,
"loss": 0.1102,
"step": 301
},
{
"epoch": 1.1615384615384616,
"grad_norm": 0.5639586474563567,
"learning_rate": 7.219757481900325e-06,
"loss": 0.1495,
"step": 302
},
{
"epoch": 1.1653846153846155,
"grad_norm": 0.6233700180987031,
"learning_rate": 7.2007428071860045e-06,
"loss": 0.1458,
"step": 303
},
{
"epoch": 1.1692307692307693,
"grad_norm": 0.6128584948161208,
"learning_rate": 7.181688574686292e-06,
"loss": 0.1481,
"step": 304
},
{
"epoch": 1.1730769230769231,
"grad_norm": 0.4965091483488049,
"learning_rate": 7.162595126896111e-06,
"loss": 0.1671,
"step": 305
},
{
"epoch": 1.176923076923077,
"grad_norm": 0.46669412021893986,
"learning_rate": 7.143462807015271e-06,
"loss": 0.1337,
"step": 306
},
{
"epoch": 1.1807692307692308,
"grad_norm": 0.559596599545597,
"learning_rate": 7.1242919589422974e-06,
"loss": 0.1399,
"step": 307
},
{
"epoch": 1.1846153846153846,
"grad_norm": 0.48106740106446627,
"learning_rate": 7.105082927268247e-06,
"loss": 0.1253,
"step": 308
},
{
"epoch": 1.1884615384615385,
"grad_norm": 0.510533996852331,
"learning_rate": 7.085836057270521e-06,
"loss": 0.1568,
"step": 309
},
{
"epoch": 1.1923076923076923,
"grad_norm": 0.42243566958325196,
"learning_rate": 7.066551694906651e-06,
"loss": 0.1322,
"step": 310
},
{
"epoch": 1.1961538461538461,
"grad_norm": 0.4280004046815802,
"learning_rate": 7.047230186808085e-06,
"loss": 0.1177,
"step": 311
},
{
"epoch": 1.2,
"grad_norm": 0.48366882870473016,
"learning_rate": 7.027871880273959e-06,
"loss": 0.117,
"step": 312
},
{
"epoch": 1.2038461538461538,
"grad_norm": 0.5151888263425758,
"learning_rate": 7.008477123264849e-06,
"loss": 0.1207,
"step": 313
},
{
"epoch": 1.2076923076923076,
"grad_norm": 0.3731825241165279,
"learning_rate": 6.989046264396516e-06,
"loss": 0.1046,
"step": 314
},
{
"epoch": 1.2115384615384615,
"grad_norm": 0.4694686758162351,
"learning_rate": 6.96957965293365e-06,
"loss": 0.1422,
"step": 315
},
{
"epoch": 1.2153846153846155,
"grad_norm": 0.5253107123426088,
"learning_rate": 6.9500776387835785e-06,
"loss": 0.1385,
"step": 316
},
{
"epoch": 1.2192307692307693,
"grad_norm": 0.41982974237167925,
"learning_rate": 6.9305405724899876e-06,
"loss": 0.1509,
"step": 317
},
{
"epoch": 1.2230769230769232,
"grad_norm": 0.6093877662241833,
"learning_rate": 6.91096880522661e-06,
"loss": 0.1276,
"step": 318
},
{
"epoch": 1.226923076923077,
"grad_norm": 0.39667627036875003,
"learning_rate": 6.891362688790925e-06,
"loss": 0.1361,
"step": 319
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.47936767790965173,
"learning_rate": 6.871722575597829e-06,
"loss": 0.1155,
"step": 320
},
{
"epoch": 1.2346153846153847,
"grad_norm": 0.47001064738519066,
"learning_rate": 6.8520488186733e-06,
"loss": 0.1394,
"step": 321
},
{
"epoch": 1.2384615384615385,
"grad_norm": 0.5869097373832258,
"learning_rate": 6.832341771648057e-06,
"loss": 0.131,
"step": 322
},
{
"epoch": 1.2423076923076923,
"grad_norm": 0.4928050688386806,
"learning_rate": 6.812601788751192e-06,
"loss": 0.1429,
"step": 323
},
{
"epoch": 1.2461538461538462,
"grad_norm": 0.4733841544242154,
"learning_rate": 6.792829224803816e-06,
"loss": 0.1186,
"step": 324
},
{
"epoch": 1.25,
"grad_norm": 0.5056515410954978,
"learning_rate": 6.773024435212678e-06,
"loss": 0.1385,
"step": 325
},
{
"epoch": 1.2538461538461538,
"grad_norm": 0.48015500052960264,
"learning_rate": 6.753187775963773e-06,
"loss": 0.1262,
"step": 326
},
{
"epoch": 1.2576923076923077,
"grad_norm": 0.5362971868851906,
"learning_rate": 6.733319603615941e-06,
"loss": 0.2009,
"step": 327
},
{
"epoch": 1.2615384615384615,
"grad_norm": 0.4516328849004065,
"learning_rate": 6.713420275294467e-06,
"loss": 0.1317,
"step": 328
},
{
"epoch": 1.2653846153846153,
"grad_norm": 0.437881596283669,
"learning_rate": 6.693490148684654e-06,
"loss": 0.1363,
"step": 329
},
{
"epoch": 1.2692307692307692,
"grad_norm": 0.45922609343053794,
"learning_rate": 6.673529582025398e-06,
"loss": 0.1531,
"step": 330
},
{
"epoch": 1.273076923076923,
"grad_norm": 0.4567443843664161,
"learning_rate": 6.653538934102743e-06,
"loss": 0.123,
"step": 331
},
{
"epoch": 1.2769230769230768,
"grad_norm": 0.5534310873771628,
"learning_rate": 6.633518564243442e-06,
"loss": 0.1314,
"step": 332
},
{
"epoch": 1.2807692307692307,
"grad_norm": 0.5290961786713094,
"learning_rate": 6.6134688323084884e-06,
"loss": 0.1514,
"step": 333
},
{
"epoch": 1.2846153846153845,
"grad_norm": 0.5198899424358424,
"learning_rate": 6.593390098686653e-06,
"loss": 0.151,
"step": 334
},
{
"epoch": 1.2884615384615383,
"grad_norm": 0.4751879865495158,
"learning_rate": 6.573282724288001e-06,
"loss": 0.1714,
"step": 335
},
{
"epoch": 1.2923076923076924,
"grad_norm": 0.4145388308561152,
"learning_rate": 6.553147070537413e-06,
"loss": 0.1053,
"step": 336
},
{
"epoch": 1.2961538461538462,
"grad_norm": 0.5368112386277263,
"learning_rate": 6.532983499368078e-06,
"loss": 0.1331,
"step": 337
},
{
"epoch": 1.3,
"grad_norm": 0.47033089006063905,
"learning_rate": 6.512792373215e-06,
"loss": 0.108,
"step": 338
},
{
"epoch": 1.3038461538461539,
"grad_norm": 0.4400959356637256,
"learning_rate": 6.492574055008474e-06,
"loss": 0.1502,
"step": 339
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.46434470648689585,
"learning_rate": 6.472328908167562e-06,
"loss": 0.1505,
"step": 340
},
{
"epoch": 1.3115384615384615,
"grad_norm": 0.5020949398975684,
"learning_rate": 6.452057296593568e-06,
"loss": 0.1244,
"step": 341
},
{
"epoch": 1.3153846153846154,
"grad_norm": 0.3712441688320081,
"learning_rate": 6.431759584663492e-06,
"loss": 0.1209,
"step": 342
},
{
"epoch": 1.3192307692307692,
"grad_norm": 0.4943645793739699,
"learning_rate": 6.411436137223479e-06,
"loss": 0.1435,
"step": 343
},
{
"epoch": 1.323076923076923,
"grad_norm": 0.47680920581232183,
"learning_rate": 6.391087319582264e-06,
"loss": 0.1059,
"step": 344
},
{
"epoch": 1.3269230769230769,
"grad_norm": 0.426335136650958,
"learning_rate": 6.370713497504607e-06,
"loss": 0.1251,
"step": 345
},
{
"epoch": 1.3307692307692307,
"grad_norm": 0.39084146456370533,
"learning_rate": 6.350315037204714e-06,
"loss": 0.1278,
"step": 346
},
{
"epoch": 1.3346153846153845,
"grad_norm": 0.5321053101222896,
"learning_rate": 6.329892305339659e-06,
"loss": 0.2413,
"step": 347
},
{
"epoch": 1.3384615384615386,
"grad_norm": 0.4071435455044252,
"learning_rate": 6.309445669002787e-06,
"loss": 0.122,
"step": 348
},
{
"epoch": 1.3423076923076924,
"grad_norm": 0.48668363499886896,
"learning_rate": 6.288975495717124e-06,
"loss": 0.1393,
"step": 349
},
{
"epoch": 1.3461538461538463,
"grad_norm": 0.5888029545894023,
"learning_rate": 6.268482153428763e-06,
"loss": 0.1176,
"step": 350
},
{
"epoch": 1.35,
"grad_norm": 0.38362834777759924,
"learning_rate": 6.247966010500258e-06,
"loss": 0.1424,
"step": 351
},
{
"epoch": 1.353846153846154,
"grad_norm": 0.42352455188457866,
"learning_rate": 6.227427435703997e-06,
"loss": 0.1135,
"step": 352
},
{
"epoch": 1.3576923076923078,
"grad_norm": 0.5947937278874783,
"learning_rate": 6.206866798215571e-06,
"loss": 0.1289,
"step": 353
},
{
"epoch": 1.3615384615384616,
"grad_norm": 0.5461049245178595,
"learning_rate": 6.186284467607149e-06,
"loss": 0.135,
"step": 354
},
{
"epoch": 1.3653846153846154,
"grad_norm": 0.3846624686030977,
"learning_rate": 6.165680813840822e-06,
"loss": 0.1288,
"step": 355
},
{
"epoch": 1.3692307692307693,
"grad_norm": 0.48492481519069053,
"learning_rate": 6.1450562072619635e-06,
"loss": 0.1227,
"step": 356
},
{
"epoch": 1.373076923076923,
"grad_norm": 0.6782174124693306,
"learning_rate": 6.124411018592568e-06,
"loss": 0.1514,
"step": 357
},
{
"epoch": 1.376923076923077,
"grad_norm": 0.47710544973857055,
"learning_rate": 6.103745618924587e-06,
"loss": 0.1365,
"step": 358
},
{
"epoch": 1.3807692307692307,
"grad_norm": 0.46533380021713566,
"learning_rate": 6.0830603797132574e-06,
"loss": 0.2014,
"step": 359
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.45504398816858743,
"learning_rate": 6.0623556727704306e-06,
"loss": 0.1353,
"step": 360
},
{
"epoch": 1.3884615384615384,
"grad_norm": 0.45086655342517956,
"learning_rate": 6.041631870257882e-06,
"loss": 0.1294,
"step": 361
},
{
"epoch": 1.3923076923076922,
"grad_norm": 0.40924394284714194,
"learning_rate": 6.020889344680627e-06,
"loss": 0.1432,
"step": 362
},
{
"epoch": 1.396153846153846,
"grad_norm": 0.6175350901418594,
"learning_rate": 6.000128468880223e-06,
"loss": 0.1375,
"step": 363
},
{
"epoch": 1.4,
"grad_norm": 0.7397207636843852,
"learning_rate": 5.979349616028067e-06,
"loss": 0.1353,
"step": 364
},
{
"epoch": 1.4038461538461537,
"grad_norm": 0.45515175476691644,
"learning_rate": 5.958553159618693e-06,
"loss": 0.128,
"step": 365
},
{
"epoch": 1.4076923076923076,
"grad_norm": 1.6609191619261732,
"learning_rate": 5.937739473463047e-06,
"loss": 0.1276,
"step": 366
},
{
"epoch": 1.4115384615384614,
"grad_norm": 0.4420710285000945,
"learning_rate": 5.916908931681781e-06,
"loss": 0.1579,
"step": 367
},
{
"epoch": 1.4153846153846155,
"grad_norm": 0.49415766950937573,
"learning_rate": 5.896061908698521e-06,
"loss": 0.1626,
"step": 368
},
{
"epoch": 1.4192307692307693,
"grad_norm": 0.45892046279925974,
"learning_rate": 5.8751987792331365e-06,
"loss": 0.1602,
"step": 369
},
{
"epoch": 1.4230769230769231,
"grad_norm": 0.6948998840071178,
"learning_rate": 5.854319918295012e-06,
"loss": 0.1579,
"step": 370
},
{
"epoch": 1.426923076923077,
"grad_norm": 0.4155348215155468,
"learning_rate": 5.833425701176294e-06,
"loss": 0.1387,
"step": 371
},
{
"epoch": 1.4307692307692308,
"grad_norm": 0.5529369524657344,
"learning_rate": 5.812516503445158e-06,
"loss": 0.15,
"step": 372
},
{
"epoch": 1.4346153846153846,
"grad_norm": 0.5135494590526845,
"learning_rate": 5.79159270093905e-06,
"loss": 0.1641,
"step": 373
},
{
"epoch": 1.4384615384615385,
"grad_norm": 0.6470992919033973,
"learning_rate": 5.770654669757935e-06,
"loss": 0.1374,
"step": 374
},
{
"epoch": 1.4423076923076923,
"grad_norm": 0.42880641227009764,
"learning_rate": 5.749702786257529e-06,
"loss": 0.1603,
"step": 375
},
{
"epoch": 1.4461538461538461,
"grad_norm": 0.47057049039010784,
"learning_rate": 5.7287374270425475e-06,
"loss": 0.1419,
"step": 376
},
{
"epoch": 1.45,
"grad_norm": 0.6574331808013163,
"learning_rate": 5.707758968959923e-06,
"loss": 0.1341,
"step": 377
},
{
"epoch": 1.4538461538461538,
"grad_norm": 0.4997633251872264,
"learning_rate": 5.686767789092041e-06,
"loss": 0.1665,
"step": 378
},
{
"epoch": 1.4576923076923076,
"grad_norm": 0.4859732351095482,
"learning_rate": 5.6657642647499545e-06,
"loss": 0.1551,
"step": 379
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.4388900288302135,
"learning_rate": 5.644748773466606e-06,
"loss": 0.114,
"step": 380
},
{
"epoch": 1.4653846153846155,
"grad_norm": 0.39167241013627174,
"learning_rate": 5.62372169299004e-06,
"loss": 0.0945,
"step": 381
},
{
"epoch": 1.4692307692307693,
"grad_norm": 0.7014801912618394,
"learning_rate": 5.6026834012766155e-06,
"loss": 0.1419,
"step": 382
},
{
"epoch": 1.4730769230769232,
"grad_norm": 0.47550996095079856,
"learning_rate": 5.581634276484211e-06,
"loss": 0.1334,
"step": 383
},
{
"epoch": 1.476923076923077,
"grad_norm": 0.488416159490385,
"learning_rate": 5.560574696965425e-06,
"loss": 0.132,
"step": 384
},
{
"epoch": 1.4807692307692308,
"grad_norm": 0.6771287939866159,
"learning_rate": 5.539505041260779e-06,
"loss": 0.1239,
"step": 385
},
{
"epoch": 1.4846153846153847,
"grad_norm": 0.5465261989484794,
"learning_rate": 5.518425688091906e-06,
"loss": 0.1225,
"step": 386
},
{
"epoch": 1.4884615384615385,
"grad_norm": 0.4507458038569731,
"learning_rate": 5.497337016354757e-06,
"loss": 0.1327,
"step": 387
},
{
"epoch": 1.4923076923076923,
"grad_norm": 0.43625367621285627,
"learning_rate": 5.476239405112775e-06,
"loss": 0.1272,
"step": 388
},
{
"epoch": 1.4961538461538462,
"grad_norm": 0.5048724623365861,
"learning_rate": 5.45513323359009e-06,
"loss": 0.1494,
"step": 389
},
{
"epoch": 1.5,
"grad_norm": 0.9170246454520701,
"learning_rate": 5.434018881164702e-06,
"loss": 0.1502,
"step": 390
},
{
"epoch": 1.5038461538461538,
"grad_norm": 0.44696965770784824,
"learning_rate": 5.412896727361663e-06,
"loss": 0.1356,
"step": 391
},
{
"epoch": 1.5076923076923077,
"grad_norm": 0.4989438723845914,
"learning_rate": 5.391767151846247e-06,
"loss": 0.1519,
"step": 392
},
{
"epoch": 1.5115384615384615,
"grad_norm": 0.6798891990262513,
"learning_rate": 5.370630534417133e-06,
"loss": 0.1359,
"step": 393
},
{
"epoch": 1.5153846153846153,
"grad_norm": 0.41423626621106613,
"learning_rate": 5.349487254999579e-06,
"loss": 0.1391,
"step": 394
},
{
"epoch": 1.5192307692307692,
"grad_norm": 1.2140564839112706,
"learning_rate": 5.328337693638591e-06,
"loss": 0.1215,
"step": 395
},
{
"epoch": 1.523076923076923,
"grad_norm": 0.4559584710573788,
"learning_rate": 5.307182230492089e-06,
"loss": 0.1174,
"step": 396
},
{
"epoch": 1.5269230769230768,
"grad_norm": 0.44247036522684585,
"learning_rate": 5.286021245824075e-06,
"loss": 0.1511,
"step": 397
},
{
"epoch": 1.5307692307692307,
"grad_norm": 0.49800448638046146,
"learning_rate": 5.264855119997803e-06,
"loss": 0.1346,
"step": 398
},
{
"epoch": 1.5346153846153845,
"grad_norm": 0.44949043382009174,
"learning_rate": 5.243684233468933e-06,
"loss": 0.1202,
"step": 399
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.4760547583833482,
"learning_rate": 5.222508966778702e-06,
"loss": 0.1935,
"step": 400
},
{
"epoch": 1.5423076923076922,
"grad_norm": 0.4166350742375338,
"learning_rate": 5.201329700547077e-06,
"loss": 0.1012,
"step": 401
},
{
"epoch": 1.546153846153846,
"grad_norm": 0.4444246935692769,
"learning_rate": 5.180146815465915e-06,
"loss": 0.115,
"step": 402
},
{
"epoch": 1.55,
"grad_norm": 0.533910509323992,
"learning_rate": 5.158960692292122e-06,
"loss": 0.1355,
"step": 403
},
{
"epoch": 1.5538461538461539,
"grad_norm": 0.43450164160623767,
"learning_rate": 5.137771711840811e-06,
"loss": 0.124,
"step": 404
},
{
"epoch": 1.5576923076923077,
"grad_norm": 0.46369894308441034,
"learning_rate": 5.116580254978447e-06,
"loss": 0.1126,
"step": 405
},
{
"epoch": 1.5615384615384615,
"grad_norm": 0.48776440784631175,
"learning_rate": 5.095386702616012e-06,
"loss": 0.1291,
"step": 406
},
{
"epoch": 1.5653846153846154,
"grad_norm": 0.4923645265888488,
"learning_rate": 5.074191435702155e-06,
"loss": 0.1211,
"step": 407
},
{
"epoch": 1.5692307692307692,
"grad_norm": 0.44345464625953795,
"learning_rate": 5.05299483521634e-06,
"loss": 0.1292,
"step": 408
},
{
"epoch": 1.573076923076923,
"grad_norm": 0.4434440705201105,
"learning_rate": 5.031797282162007e-06,
"loss": 0.1395,
"step": 409
},
{
"epoch": 1.5769230769230769,
"grad_norm": 0.45427496650319094,
"learning_rate": 5.010599157559713e-06,
"loss": 0.1455,
"step": 410
},
{
"epoch": 1.580769230769231,
"grad_norm": 0.6070725699602748,
"learning_rate": 4.98940084244029e-06,
"loss": 0.1538,
"step": 411
},
{
"epoch": 1.5846153846153848,
"grad_norm": 0.458323844100121,
"learning_rate": 4.968202717837996e-06,
"loss": 0.1299,
"step": 412
},
{
"epoch": 1.5884615384615386,
"grad_norm": 0.5391860168517009,
"learning_rate": 4.947005164783661e-06,
"loss": 0.1334,
"step": 413
},
{
"epoch": 1.5923076923076924,
"grad_norm": 0.5054626431795982,
"learning_rate": 4.925808564297847e-06,
"loss": 0.1601,
"step": 414
},
{
"epoch": 1.5961538461538463,
"grad_norm": 0.4183671734763695,
"learning_rate": 4.9046132973839895e-06,
"loss": 0.1372,
"step": 415
},
{
"epoch": 1.6,
"grad_norm": 0.4102564890956905,
"learning_rate": 4.883419745021554e-06,
"loss": 0.1707,
"step": 416
},
{
"epoch": 1.603846153846154,
"grad_norm": 0.4646782588590493,
"learning_rate": 4.862228288159191e-06,
"loss": 0.1176,
"step": 417
},
{
"epoch": 1.6076923076923078,
"grad_norm": 0.5107083280849873,
"learning_rate": 4.841039307707878e-06,
"loss": 0.1405,
"step": 418
},
{
"epoch": 1.6115384615384616,
"grad_norm": 0.4684439234140132,
"learning_rate": 4.819853184534085e-06,
"loss": 0.1343,
"step": 419
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.6079521206383508,
"learning_rate": 4.798670299452926e-06,
"loss": 0.1132,
"step": 420
},
{
"epoch": 1.6192307692307693,
"grad_norm": 0.4031654223056794,
"learning_rate": 4.7774910332213005e-06,
"loss": 0.129,
"step": 421
},
{
"epoch": 1.623076923076923,
"grad_norm": 0.5328871952949895,
"learning_rate": 4.756315766531069e-06,
"loss": 0.1359,
"step": 422
},
{
"epoch": 1.626923076923077,
"grad_norm": 0.5022257582467073,
"learning_rate": 4.735144880002199e-06,
"loss": 0.153,
"step": 423
},
{
"epoch": 1.6307692307692307,
"grad_norm": 0.4304592363289141,
"learning_rate": 4.713978754175926e-06,
"loss": 0.1439,
"step": 424
},
{
"epoch": 1.6346153846153846,
"grad_norm": 0.39302596076648394,
"learning_rate": 4.692817769507912e-06,
"loss": 0.1264,
"step": 425
},
{
"epoch": 1.6384615384615384,
"grad_norm": 0.46876631025581683,
"learning_rate": 4.671662306361409e-06,
"loss": 0.1137,
"step": 426
},
{
"epoch": 1.6423076923076922,
"grad_norm": 0.5106162680206295,
"learning_rate": 4.6505127450004216e-06,
"loss": 0.1075,
"step": 427
},
{
"epoch": 1.646153846153846,
"grad_norm": 0.45713873670960553,
"learning_rate": 4.62936946558287e-06,
"loss": 0.1157,
"step": 428
},
{
"epoch": 1.65,
"grad_norm": 0.4644840983754053,
"learning_rate": 4.608232848153757e-06,
"loss": 0.1356,
"step": 429
},
{
"epoch": 1.6538461538461537,
"grad_norm": 0.44453457980173716,
"learning_rate": 4.587103272638339e-06,
"loss": 0.1298,
"step": 430
},
{
"epoch": 1.6576923076923076,
"grad_norm": 0.5878089591324217,
"learning_rate": 4.565981118835299e-06,
"loss": 0.1428,
"step": 431
},
{
"epoch": 1.6615384615384614,
"grad_norm": 0.47131100179520885,
"learning_rate": 4.5448667664099125e-06,
"loss": 0.1474,
"step": 432
},
{
"epoch": 1.6653846153846152,
"grad_norm": 0.44845225011915896,
"learning_rate": 4.523760594887228e-06,
"loss": 0.116,
"step": 433
},
{
"epoch": 1.669230769230769,
"grad_norm": 0.5035438810092997,
"learning_rate": 4.5026629836452445e-06,
"loss": 0.2391,
"step": 434
},
{
"epoch": 1.6730769230769231,
"grad_norm": 0.42339637890957893,
"learning_rate": 4.481574311908096e-06,
"loss": 0.1198,
"step": 435
},
{
"epoch": 1.676923076923077,
"grad_norm": 0.7790337291754627,
"learning_rate": 4.460494958739223e-06,
"loss": 0.1291,
"step": 436
},
{
"epoch": 1.6807692307692308,
"grad_norm": 0.506089200738741,
"learning_rate": 4.439425303034576e-06,
"loss": 0.1776,
"step": 437
},
{
"epoch": 1.6846153846153846,
"grad_norm": 0.4507039321757444,
"learning_rate": 4.418365723515791e-06,
"loss": 0.13,
"step": 438
},
{
"epoch": 1.6884615384615385,
"grad_norm": 0.4788803675921322,
"learning_rate": 4.397316598723385e-06,
"loss": 0.1368,
"step": 439
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.4363002427021318,
"learning_rate": 4.376278307009962e-06,
"loss": 0.1525,
"step": 440
},
{
"epoch": 1.6961538461538461,
"grad_norm": 0.452040220712686,
"learning_rate": 4.355251226533396e-06,
"loss": 0.1239,
"step": 441
},
{
"epoch": 1.7,
"grad_norm": 0.4444147868578959,
"learning_rate": 4.334235735250047e-06,
"loss": 0.1348,
"step": 442
},
{
"epoch": 1.703846153846154,
"grad_norm": 0.4122642938441922,
"learning_rate": 4.313232210907959e-06,
"loss": 0.1511,
"step": 443
},
{
"epoch": 1.7076923076923078,
"grad_norm": 0.42228027702746485,
"learning_rate": 4.292241031040077e-06,
"loss": 0.1169,
"step": 444
},
{
"epoch": 1.7115384615384617,
"grad_norm": 0.5023232429315645,
"learning_rate": 4.271262572957453e-06,
"loss": 0.0999,
"step": 445
},
{
"epoch": 1.7153846153846155,
"grad_norm": 0.6395336343347494,
"learning_rate": 4.250297213742473e-06,
"loss": 0.1296,
"step": 446
},
{
"epoch": 1.7192307692307693,
"grad_norm": 0.9806901361079964,
"learning_rate": 4.229345330242067e-06,
"loss": 0.1315,
"step": 447
},
{
"epoch": 1.7230769230769232,
"grad_norm": 0.5475610390594655,
"learning_rate": 4.2084072990609505e-06,
"loss": 0.1551,
"step": 448
},
{
"epoch": 1.726923076923077,
"grad_norm": 0.4155815237477355,
"learning_rate": 4.187483496554844e-06,
"loss": 0.1337,
"step": 449
},
{
"epoch": 1.7307692307692308,
"grad_norm": 0.4785007758287145,
"learning_rate": 4.166574298823707e-06,
"loss": 0.121,
"step": 450
},
{
"epoch": 1.7346153846153847,
"grad_norm": 0.4131668662651965,
"learning_rate": 4.145680081704989e-06,
"loss": 0.1209,
"step": 451
},
{
"epoch": 1.7384615384615385,
"grad_norm": 0.42715434771819955,
"learning_rate": 4.1248012207668635e-06,
"loss": 0.1224,
"step": 452
},
{
"epoch": 1.7423076923076923,
"grad_norm": 0.5026594173929048,
"learning_rate": 4.103938091301479e-06,
"loss": 0.1265,
"step": 453
},
{
"epoch": 1.7461538461538462,
"grad_norm": 0.4459554274890392,
"learning_rate": 4.08309106831822e-06,
"loss": 0.1484,
"step": 454
},
{
"epoch": 1.75,
"grad_norm": 0.472714665955711,
"learning_rate": 4.062260526536955e-06,
"loss": 0.1504,
"step": 455
},
{
"epoch": 1.7538461538461538,
"grad_norm": 0.4988919356916683,
"learning_rate": 4.041446840381309e-06,
"loss": 0.154,
"step": 456
},
{
"epoch": 1.7576923076923077,
"grad_norm": 0.5253587239521474,
"learning_rate": 4.0206503839719335e-06,
"loss": 0.14,
"step": 457
},
{
"epoch": 1.7615384615384615,
"grad_norm": 0.4492390307989714,
"learning_rate": 3.999871531119779e-06,
"loss": 0.1136,
"step": 458
},
{
"epoch": 1.7653846153846153,
"grad_norm": 0.5023116000205898,
"learning_rate": 3.9791106553193746e-06,
"loss": 0.1212,
"step": 459
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.42454244962616716,
"learning_rate": 3.9583681297421194e-06,
"loss": 0.1459,
"step": 460
},
{
"epoch": 1.773076923076923,
"grad_norm": 0.5662100026417451,
"learning_rate": 3.937644327229572e-06,
"loss": 0.1524,
"step": 461
},
{
"epoch": 1.7769230769230768,
"grad_norm": 0.5146122535226022,
"learning_rate": 3.916939620286743e-06,
"loss": 0.1312,
"step": 462
},
{
"epoch": 1.7807692307692307,
"grad_norm": 0.4486397649285054,
"learning_rate": 3.896254381075416e-06,
"loss": 0.123,
"step": 463
},
{
"epoch": 1.7846153846153845,
"grad_norm": 0.6395888708077827,
"learning_rate": 3.875588981407433e-06,
"loss": 0.1502,
"step": 464
},
{
"epoch": 1.7884615384615383,
"grad_norm": 0.4198870967376798,
"learning_rate": 3.854943792738037e-06,
"loss": 0.1431,
"step": 465
},
{
"epoch": 1.7923076923076922,
"grad_norm": 0.606706785255943,
"learning_rate": 3.834319186159179e-06,
"loss": 0.1429,
"step": 466
},
{
"epoch": 1.796153846153846,
"grad_norm": 0.4379481592572174,
"learning_rate": 3.8137155323928526e-06,
"loss": 0.1385,
"step": 467
},
{
"epoch": 1.8,
"grad_norm": 0.6271827852645024,
"learning_rate": 3.7931332017844302e-06,
"loss": 0.1322,
"step": 468
},
{
"epoch": 1.8038461538461539,
"grad_norm": 0.4878018347723653,
"learning_rate": 3.7725725642960047e-06,
"loss": 0.1367,
"step": 469
},
{
"epoch": 1.8076923076923077,
"grad_norm": 0.46400447334347933,
"learning_rate": 3.752033989499742e-06,
"loss": 0.1187,
"step": 470
},
{
"epoch": 1.8115384615384615,
"grad_norm": 0.5050384116909469,
"learning_rate": 3.7315178465712364e-06,
"loss": 0.1363,
"step": 471
},
{
"epoch": 1.8153846153846154,
"grad_norm": 0.4426898845807984,
"learning_rate": 3.7110245042828786e-06,
"loss": 0.1275,
"step": 472
},
{
"epoch": 1.8192307692307692,
"grad_norm": 0.4434299102480052,
"learning_rate": 3.690554330997215e-06,
"loss": 0.1348,
"step": 473
},
{
"epoch": 1.823076923076923,
"grad_norm": 0.47754965677866235,
"learning_rate": 3.670107694660343e-06,
"loss": 0.1515,
"step": 474
},
{
"epoch": 1.8269230769230769,
"grad_norm": 0.43358801186896084,
"learning_rate": 3.6496849627952875e-06,
"loss": 0.1158,
"step": 475
},
{
"epoch": 1.830769230769231,
"grad_norm": 0.4226974949276786,
"learning_rate": 3.6292865024953945e-06,
"loss": 0.1367,
"step": 476
},
{
"epoch": 1.8346153846153848,
"grad_norm": 0.43661311886443205,
"learning_rate": 3.6089126804177373e-06,
"loss": 0.1344,
"step": 477
},
{
"epoch": 1.8384615384615386,
"grad_norm": 0.45333921024590756,
"learning_rate": 3.5885638627765228e-06,
"loss": 0.1427,
"step": 478
},
{
"epoch": 1.8423076923076924,
"grad_norm": 0.48402020893920095,
"learning_rate": 3.568240415336509e-06,
"loss": 0.122,
"step": 479
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.40298650859806084,
"learning_rate": 3.547942703406433e-06,
"loss": 0.1212,
"step": 480
},
{
"epoch": 1.85,
"grad_norm": 0.5042549446694097,
"learning_rate": 3.52767109183244e-06,
"loss": 0.1695,
"step": 481
},
{
"epoch": 1.853846153846154,
"grad_norm": 0.5107691624929165,
"learning_rate": 3.507425944991529e-06,
"loss": 0.1244,
"step": 482
},
{
"epoch": 1.8576923076923078,
"grad_norm": 0.49360727300936846,
"learning_rate": 3.4872076267850015e-06,
"loss": 0.1377,
"step": 483
},
{
"epoch": 1.8615384615384616,
"grad_norm": 0.4616829419196794,
"learning_rate": 3.4670165006319236e-06,
"loss": 0.1171,
"step": 484
},
{
"epoch": 1.8653846153846154,
"grad_norm": 0.41756642122985227,
"learning_rate": 3.4468529294625895e-06,
"loss": 0.1539,
"step": 485
},
{
"epoch": 1.8692307692307693,
"grad_norm": 0.4815335312917406,
"learning_rate": 3.4267172757120005e-06,
"loss": 0.1386,
"step": 486
},
{
"epoch": 1.873076923076923,
"grad_norm": 0.47899837926897626,
"learning_rate": 3.406609901313349e-06,
"loss": 0.1282,
"step": 487
},
{
"epoch": 1.876923076923077,
"grad_norm": 0.44416620378594085,
"learning_rate": 3.386531167691512e-06,
"loss": 0.1176,
"step": 488
},
{
"epoch": 1.8807692307692307,
"grad_norm": 0.4701263283677266,
"learning_rate": 3.36648143575656e-06,
"loss": 0.1398,
"step": 489
},
{
"epoch": 1.8846153846153846,
"grad_norm": 0.43707107458284244,
"learning_rate": 3.3464610658972584e-06,
"loss": 0.1521,
"step": 490
},
{
"epoch": 1.8884615384615384,
"grad_norm": 0.5420627983517349,
"learning_rate": 3.326470417974604e-06,
"loss": 0.1404,
"step": 491
},
{
"epoch": 1.8923076923076922,
"grad_norm": 0.47739629608077666,
"learning_rate": 3.3065098513153473e-06,
"loss": 0.1441,
"step": 492
},
{
"epoch": 1.896153846153846,
"grad_norm": 0.4528983270895438,
"learning_rate": 3.2865797247055354e-06,
"loss": 0.1206,
"step": 493
},
{
"epoch": 1.9,
"grad_norm": 0.45441612168206696,
"learning_rate": 3.266680396384061e-06,
"loss": 0.1355,
"step": 494
},
{
"epoch": 1.9038461538461537,
"grad_norm": 0.4781469564564768,
"learning_rate": 3.2468122240362287e-06,
"loss": 0.1556,
"step": 495
},
{
"epoch": 1.9076923076923076,
"grad_norm": 0.5608882041787969,
"learning_rate": 3.226975564787322e-06,
"loss": 0.1324,
"step": 496
},
{
"epoch": 1.9115384615384614,
"grad_norm": 0.43281591939339464,
"learning_rate": 3.2071707751961838e-06,
"loss": 0.1275,
"step": 497
},
{
"epoch": 1.9153846153846152,
"grad_norm": 0.5385345123846502,
"learning_rate": 3.187398211248811e-06,
"loss": 0.1328,
"step": 498
},
{
"epoch": 1.919230769230769,
"grad_norm": 0.4427252943912305,
"learning_rate": 3.1676582283519454e-06,
"loss": 0.1442,
"step": 499
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.5580489291247633,
"learning_rate": 3.1479511813267006e-06,
"loss": 0.1701,
"step": 500
},
{
"epoch": 1.926923076923077,
"grad_norm": 0.42738109025167625,
"learning_rate": 3.1282774244021717e-06,
"loss": 0.1487,
"step": 501
},
{
"epoch": 1.9307692307692308,
"grad_norm": 0.4197927181025155,
"learning_rate": 3.1086373112090762e-06,
"loss": 0.1401,
"step": 502
},
{
"epoch": 1.9346153846153846,
"grad_norm": 0.485578991722214,
"learning_rate": 3.089031194773392e-06,
"loss": 0.1266,
"step": 503
},
{
"epoch": 1.9384615384615385,
"grad_norm": 0.40528827387332184,
"learning_rate": 3.069459427510014e-06,
"loss": 0.1081,
"step": 504
},
{
"epoch": 1.9423076923076923,
"grad_norm": 0.2982255244423721,
"learning_rate": 3.049922361216422e-06,
"loss": 0.1225,
"step": 505
},
{
"epoch": 1.9461538461538461,
"grad_norm": 0.4494454327878644,
"learning_rate": 3.0304203470663507e-06,
"loss": 0.187,
"step": 506
},
{
"epoch": 1.95,
"grad_norm": 0.42440291119266754,
"learning_rate": 3.0109537356034856e-06,
"loss": 0.1343,
"step": 507
},
{
"epoch": 1.953846153846154,
"grad_norm": 0.48355316256761033,
"learning_rate": 2.991522876735154e-06,
"loss": 0.1632,
"step": 508
},
{
"epoch": 1.9576923076923078,
"grad_norm": 0.45841357810520994,
"learning_rate": 2.9721281197260427e-06,
"loss": 0.1434,
"step": 509
},
{
"epoch": 1.9615384615384617,
"grad_norm": 0.4903499302961153,
"learning_rate": 2.9527698131919156e-06,
"loss": 0.135,
"step": 510
},
{
"epoch": 1.9653846153846155,
"grad_norm": 0.4516756333489649,
"learning_rate": 2.9334483050933506e-06,
"loss": 0.1004,
"step": 511
},
{
"epoch": 1.9692307692307693,
"grad_norm": 0.3885518784675757,
"learning_rate": 2.91416394272948e-06,
"loss": 0.1283,
"step": 512
},
{
"epoch": 1.9730769230769232,
"grad_norm": 0.4807072989402195,
"learning_rate": 2.894917072731753e-06,
"loss": 0.1273,
"step": 513
},
{
"epoch": 1.976923076923077,
"grad_norm": 0.3298706628288234,
"learning_rate": 2.8757080410577042e-06,
"loss": 0.1256,
"step": 514
},
{
"epoch": 1.9807692307692308,
"grad_norm": 0.4358441238822011,
"learning_rate": 2.8565371929847286e-06,
"loss": 0.1548,
"step": 515
},
{
"epoch": 1.9846153846153847,
"grad_norm": 0.4838560594526077,
"learning_rate": 2.83740487310389e-06,
"loss": 0.1394,
"step": 516
},
{
"epoch": 1.9884615384615385,
"grad_norm": 0.4315845410243963,
"learning_rate": 2.81831142531371e-06,
"loss": 0.1406,
"step": 517
},
{
"epoch": 1.9923076923076923,
"grad_norm": 0.47587604267234745,
"learning_rate": 2.7992571928139984e-06,
"loss": 0.1611,
"step": 518
},
{
"epoch": 1.9961538461538462,
"grad_norm": 0.5184386458450941,
"learning_rate": 2.780242518099675e-06,
"loss": 0.1184,
"step": 519
},
{
"epoch": 2.0,
"grad_norm": 0.4684932913833738,
"learning_rate": 2.761267742954629e-06,
"loss": 0.1453,
"step": 520
},
{
"epoch": 2.003846153846154,
"grad_norm": 0.47793383729428707,
"learning_rate": 2.7423332084455543e-06,
"loss": 0.0823,
"step": 521
},
{
"epoch": 2.0076923076923077,
"grad_norm": 0.39849979531086066,
"learning_rate": 2.723439254915834e-06,
"loss": 0.0618,
"step": 522
},
{
"epoch": 2.0115384615384615,
"grad_norm": 0.33884579819308913,
"learning_rate": 2.704586221979422e-06,
"loss": 0.0707,
"step": 523
},
{
"epoch": 2.0153846153846153,
"grad_norm": 0.5981284034440648,
"learning_rate": 2.6857744485147286e-06,
"loss": 0.0602,
"step": 524
},
{
"epoch": 2.019230769230769,
"grad_norm": 0.4527716025354856,
"learning_rate": 2.667004272658541e-06,
"loss": 0.07,
"step": 525
},
{
"epoch": 2.023076923076923,
"grad_norm": 0.40671494702093786,
"learning_rate": 2.6482760317999338e-06,
"loss": 0.0699,
"step": 526
},
{
"epoch": 2.026923076923077,
"grad_norm": 0.30738206365401133,
"learning_rate": 2.629590062574221e-06,
"loss": 0.0549,
"step": 527
},
{
"epoch": 2.0307692307692307,
"grad_norm": 0.4385197938785045,
"learning_rate": 2.610946700856885e-06,
"loss": 0.0794,
"step": 528
},
{
"epoch": 2.0346153846153845,
"grad_norm": 0.4665222560596245,
"learning_rate": 2.592346281757552e-06,
"loss": 0.0617,
"step": 529
},
{
"epoch": 2.0384615384615383,
"grad_norm": 0.5058676492156543,
"learning_rate": 2.5737891396139713e-06,
"loss": 0.0796,
"step": 530
},
{
"epoch": 2.042307692307692,
"grad_norm": 0.45992569901610875,
"learning_rate": 2.5552756079859904e-06,
"loss": 0.053,
"step": 531
},
{
"epoch": 2.046153846153846,
"grad_norm": 0.5454750893677381,
"learning_rate": 2.5368060196495785e-06,
"loss": 0.0726,
"step": 532
},
{
"epoch": 2.05,
"grad_norm": 0.43951679224430495,
"learning_rate": 2.5183807065908296e-06,
"loss": 0.0523,
"step": 533
},
{
"epoch": 2.0538461538461537,
"grad_norm": 0.44358619030146745,
"learning_rate": 2.5000000000000015e-06,
"loss": 0.0694,
"step": 534
},
{
"epoch": 2.0576923076923075,
"grad_norm": 0.5521904920835358,
"learning_rate": 2.4816642302655634e-06,
"loss": 0.0658,
"step": 535
},
{
"epoch": 2.0615384615384613,
"grad_norm": 0.4735376595609427,
"learning_rate": 2.4633737269682546e-06,
"loss": 0.0774,
"step": 536
},
{
"epoch": 2.0653846153846156,
"grad_norm": 0.5086670036985256,
"learning_rate": 2.445128818875166e-06,
"loss": 0.0629,
"step": 537
},
{
"epoch": 2.0692307692307694,
"grad_norm": 0.45818088089777165,
"learning_rate": 2.4269298339338205e-06,
"loss": 0.0787,
"step": 538
},
{
"epoch": 2.0730769230769233,
"grad_norm": 0.43968654719422656,
"learning_rate": 2.408777099266291e-06,
"loss": 0.0752,
"step": 539
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.4126840760578833,
"learning_rate": 2.3906709411633073e-06,
"loss": 0.0912,
"step": 540
},
{
"epoch": 2.080769230769231,
"grad_norm": 0.4130812299037541,
"learning_rate": 2.3726116850783987e-06,
"loss": 0.0567,
"step": 541
},
{
"epoch": 2.0846153846153848,
"grad_norm": 0.4395741035888438,
"learning_rate": 2.354599655622049e-06,
"loss": 0.0674,
"step": 542
},
{
"epoch": 2.0884615384615386,
"grad_norm": 0.45086509001739355,
"learning_rate": 2.3366351765558437e-06,
"loss": 0.0756,
"step": 543
},
{
"epoch": 2.0923076923076924,
"grad_norm": 0.48282994510642635,
"learning_rate": 2.318718570786675e-06,
"loss": 0.0641,
"step": 544
},
{
"epoch": 2.0961538461538463,
"grad_norm": 0.4711906045259754,
"learning_rate": 2.3008501603609147e-06,
"loss": 0.0845,
"step": 545
},
{
"epoch": 2.1,
"grad_norm": 0.42914189401808145,
"learning_rate": 2.283030266458644e-06,
"loss": 0.0556,
"step": 546
},
{
"epoch": 2.103846153846154,
"grad_norm": 0.389827539786175,
"learning_rate": 2.265259209387867e-06,
"loss": 0.0707,
"step": 547
},
{
"epoch": 2.1076923076923078,
"grad_norm": 0.697179239425688,
"learning_rate": 2.2475373085787568e-06,
"loss": 0.0794,
"step": 548
},
{
"epoch": 2.1115384615384616,
"grad_norm": 0.4100141866606658,
"learning_rate": 2.229864882577921e-06,
"loss": 0.0604,
"step": 549
},
{
"epoch": 2.1153846153846154,
"grad_norm": 0.39522061304833667,
"learning_rate": 2.2122422490426676e-06,
"loss": 0.091,
"step": 550
},
{
"epoch": 2.1192307692307693,
"grad_norm": 0.38166594460266384,
"learning_rate": 2.194669724735296e-06,
"loss": 0.0573,
"step": 551
},
{
"epoch": 2.123076923076923,
"grad_norm": 0.4001083469139412,
"learning_rate": 2.1771476255174056e-06,
"loss": 0.0499,
"step": 552
},
{
"epoch": 2.126923076923077,
"grad_norm": 0.4626064492597376,
"learning_rate": 2.159676266344222e-06,
"loss": 0.0698,
"step": 553
},
{
"epoch": 2.1307692307692307,
"grad_norm": 0.41666309655074274,
"learning_rate": 2.1422559612589266e-06,
"loss": 0.0617,
"step": 554
},
{
"epoch": 2.1346153846153846,
"grad_norm": 0.4262870648495164,
"learning_rate": 2.124887023387017e-06,
"loss": 0.0735,
"step": 555
},
{
"epoch": 2.1384615384615384,
"grad_norm": 0.4859562467898189,
"learning_rate": 2.1075697649306838e-06,
"loss": 0.0597,
"step": 556
},
{
"epoch": 2.1423076923076922,
"grad_norm": 0.5284718879167324,
"learning_rate": 2.0903044971631854e-06,
"loss": 0.0749,
"step": 557
},
{
"epoch": 2.146153846153846,
"grad_norm": 0.4136460036845487,
"learning_rate": 2.0730915304232692e-06,
"loss": 0.0685,
"step": 558
},
{
"epoch": 2.15,
"grad_norm": 0.42256412207916466,
"learning_rate": 2.055931174109579e-06,
"loss": 0.0652,
"step": 559
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.46506993278363223,
"learning_rate": 2.0388237366751005e-06,
"loss": 0.0612,
"step": 560
},
{
"epoch": 2.1576923076923076,
"grad_norm": 0.5243666475353842,
"learning_rate": 2.02176952562162e-06,
"loss": 0.0673,
"step": 561
},
{
"epoch": 2.1615384615384614,
"grad_norm": 0.44668978137392323,
"learning_rate": 2.004768847494186e-06,
"loss": 0.0563,
"step": 562
},
{
"epoch": 2.1653846153846152,
"grad_norm": 0.45380630005482203,
"learning_rate": 1.987822007875617e-06,
"loss": 0.0835,
"step": 563
},
{
"epoch": 2.169230769230769,
"grad_norm": 0.5098253768884685,
"learning_rate": 1.9709293113809876e-06,
"loss": 0.082,
"step": 564
},
{
"epoch": 2.173076923076923,
"grad_norm": 0.455759921626837,
"learning_rate": 1.954091061652172e-06,
"loss": 0.0615,
"step": 565
},
{
"epoch": 2.1769230769230767,
"grad_norm": 0.48372996180254846,
"learning_rate": 1.9373075613523728e-06,
"loss": 0.0763,
"step": 566
},
{
"epoch": 2.1807692307692306,
"grad_norm": 0.40201657125570894,
"learning_rate": 1.920579112160685e-06,
"loss": 0.0553,
"step": 567
},
{
"epoch": 2.184615384615385,
"grad_norm": 0.4374660881776204,
"learning_rate": 1.903906014766681e-06,
"loss": 0.115,
"step": 568
},
{
"epoch": 2.1884615384615387,
"grad_norm": 0.4503874407503947,
"learning_rate": 1.8872885688649879e-06,
"loss": 0.0603,
"step": 569
},
{
"epoch": 2.1923076923076925,
"grad_norm": 0.41424779140331675,
"learning_rate": 1.8707270731499223e-06,
"loss": 0.0556,
"step": 570
},
{
"epoch": 2.1961538461538463,
"grad_norm": 0.4453033155848563,
"learning_rate": 1.854221825310103e-06,
"loss": 0.0562,
"step": 571
},
{
"epoch": 2.2,
"grad_norm": 0.45215159642326885,
"learning_rate": 1.8377731220231144e-06,
"loss": 0.0556,
"step": 572
},
{
"epoch": 2.203846153846154,
"grad_norm": 0.47518683806930967,
"learning_rate": 1.8213812589501611e-06,
"loss": 0.0619,
"step": 573
},
{
"epoch": 2.207692307692308,
"grad_norm": 0.43933463234454156,
"learning_rate": 1.8050465307307602e-06,
"loss": 0.0574,
"step": 574
},
{
"epoch": 2.2115384615384617,
"grad_norm": 0.42035870042078616,
"learning_rate": 1.78876923097745e-06,
"loss": 0.0686,
"step": 575
},
{
"epoch": 2.2153846153846155,
"grad_norm": 0.4669752341333353,
"learning_rate": 1.7725496522704998e-06,
"loss": 0.077,
"step": 576
},
{
"epoch": 2.2192307692307693,
"grad_norm": 0.42846303720639206,
"learning_rate": 1.7563880861526656e-06,
"loss": 0.0647,
"step": 577
},
{
"epoch": 2.223076923076923,
"grad_norm": 0.4285292702392478,
"learning_rate": 1.7402848231239317e-06,
"loss": 0.0597,
"step": 578
},
{
"epoch": 2.226923076923077,
"grad_norm": 0.45580885754800243,
"learning_rate": 1.7242401526363095e-06,
"loss": 0.0829,
"step": 579
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.3618185605338237,
"learning_rate": 1.70825436308862e-06,
"loss": 0.055,
"step": 580
},
{
"epoch": 2.2346153846153847,
"grad_norm": 0.47581864754355635,
"learning_rate": 1.692327741821312e-06,
"loss": 0.0732,
"step": 581
},
{
"epoch": 2.2384615384615385,
"grad_norm": 0.4619237642411215,
"learning_rate": 1.676460575111306e-06,
"loss": 0.0758,
"step": 582
},
{
"epoch": 2.2423076923076923,
"grad_norm": 0.5362765575176242,
"learning_rate": 1.6606531481668364e-06,
"loss": 0.0701,
"step": 583
},
{
"epoch": 2.246153846153846,
"grad_norm": 0.4786146132319656,
"learning_rate": 1.6449057451223354e-06,
"loss": 0.0535,
"step": 584
},
{
"epoch": 2.25,
"grad_norm": 0.4839429550887217,
"learning_rate": 1.6292186490333172e-06,
"loss": 0.0589,
"step": 585
},
{
"epoch": 2.253846153846154,
"grad_norm": 0.45077627033612566,
"learning_rate": 1.6135921418712959e-06,
"loss": 0.0627,
"step": 586
},
{
"epoch": 2.2576923076923077,
"grad_norm": 0.5741588985926117,
"learning_rate": 1.5980265045187139e-06,
"loss": 0.1126,
"step": 587
},
{
"epoch": 2.2615384615384615,
"grad_norm": 0.4730374054996576,
"learning_rate": 1.5825220167638945e-06,
"loss": 0.084,
"step": 588
},
{
"epoch": 2.2653846153846153,
"grad_norm": 0.507165197959675,
"learning_rate": 1.567078957296016e-06,
"loss": 0.062,
"step": 589
},
{
"epoch": 2.269230769230769,
"grad_norm": 0.4214352448201897,
"learning_rate": 1.5516976037000941e-06,
"loss": 0.0563,
"step": 590
},
{
"epoch": 2.273076923076923,
"grad_norm": 0.4056826036266477,
"learning_rate": 1.5363782324520033e-06,
"loss": 0.0715,
"step": 591
},
{
"epoch": 2.276923076923077,
"grad_norm": 0.42984780162111796,
"learning_rate": 1.5211211189134955e-06,
"loss": 0.0586,
"step": 592
},
{
"epoch": 2.2807692307692307,
"grad_norm": 0.406564554228904,
"learning_rate": 1.5059265373272574e-06,
"loss": 0.067,
"step": 593
},
{
"epoch": 2.2846153846153845,
"grad_norm": 0.3711180987151303,
"learning_rate": 1.4907947608119866e-06,
"loss": 0.0613,
"step": 594
},
{
"epoch": 2.2884615384615383,
"grad_norm": 0.418168980244583,
"learning_rate": 1.475726061357463e-06,
"loss": 0.065,
"step": 595
},
{
"epoch": 2.292307692307692,
"grad_norm": 0.5190985953199084,
"learning_rate": 1.4607207098196851e-06,
"loss": 0.0892,
"step": 596
},
{
"epoch": 2.296153846153846,
"grad_norm": 0.4816524197622573,
"learning_rate": 1.4457789759159813e-06,
"loss": 0.0749,
"step": 597
},
{
"epoch": 2.3,
"grad_norm": 0.38794064261472316,
"learning_rate": 1.430901128220174e-06,
"loss": 0.0684,
"step": 598
},
{
"epoch": 2.3038461538461537,
"grad_norm": 0.6120321930974998,
"learning_rate": 1.4160874341577447e-06,
"loss": 0.0628,
"step": 599
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.47542545440481365,
"learning_rate": 1.4013381600010278e-06,
"loss": 0.0664,
"step": 600
},
{
"epoch": 2.3115384615384613,
"grad_norm": 0.3701693597529931,
"learning_rate": 1.3866535708644335e-06,
"loss": 0.0673,
"step": 601
},
{
"epoch": 2.315384615384615,
"grad_norm": 0.4854777269446637,
"learning_rate": 1.3720339306996666e-06,
"loss": 0.0759,
"step": 602
},
{
"epoch": 2.319230769230769,
"grad_norm": 0.4316692129118829,
"learning_rate": 1.3574795022910014e-06,
"loss": 0.0839,
"step": 603
},
{
"epoch": 2.3230769230769233,
"grad_norm": 0.44038784142415865,
"learning_rate": 1.3429905472505344e-06,
"loss": 0.0581,
"step": 604
},
{
"epoch": 2.326923076923077,
"grad_norm": 0.38299630437239945,
"learning_rate": 1.3285673260135073e-06,
"loss": 0.0632,
"step": 605
},
{
"epoch": 2.330769230769231,
"grad_norm": 0.41522982556143445,
"learning_rate": 1.314210097833607e-06,
"loss": 0.06,
"step": 606
},
{
"epoch": 2.3346153846153848,
"grad_norm": 0.6027057153935131,
"learning_rate": 1.2999191207783129e-06,
"loss": 0.0696,
"step": 607
},
{
"epoch": 2.3384615384615386,
"grad_norm": 0.4818374982320264,
"learning_rate": 1.2856946517242608e-06,
"loss": 0.0699,
"step": 608
},
{
"epoch": 2.3423076923076924,
"grad_norm": 0.5497023747464141,
"learning_rate": 1.2715369463526173e-06,
"loss": 0.0701,
"step": 609
},
{
"epoch": 2.3461538461538463,
"grad_norm": 0.4606936503659749,
"learning_rate": 1.257446259144494e-06,
"loss": 0.0597,
"step": 610
},
{
"epoch": 2.35,
"grad_norm": 0.7733361652350942,
"learning_rate": 1.2434228433763657e-06,
"loss": 0.083,
"step": 611
},
{
"epoch": 2.353846153846154,
"grad_norm": 0.4181960857278969,
"learning_rate": 1.2294669511155193e-06,
"loss": 0.0618,
"step": 612
},
{
"epoch": 2.3576923076923078,
"grad_norm": 0.39776114984038474,
"learning_rate": 1.215578833215526e-06,
"loss": 0.0625,
"step": 613
},
{
"epoch": 2.3615384615384616,
"grad_norm": 0.4880050014301517,
"learning_rate": 1.201758739311728e-06,
"loss": 0.0822,
"step": 614
},
{
"epoch": 2.3653846153846154,
"grad_norm": 0.44536411778117985,
"learning_rate": 1.1880069178167586e-06,
"loss": 0.0777,
"step": 615
},
{
"epoch": 2.3692307692307693,
"grad_norm": 0.39395356198291914,
"learning_rate": 1.1743236159160654e-06,
"loss": 0.0723,
"step": 616
},
{
"epoch": 2.373076923076923,
"grad_norm": 0.7269529693254339,
"learning_rate": 1.1607090795634802e-06,
"loss": 0.0671,
"step": 617
},
{
"epoch": 2.376923076923077,
"grad_norm": 0.4631259071739605,
"learning_rate": 1.1471635534767877e-06,
"loss": 0.1831,
"step": 618
},
{
"epoch": 2.3807692307692307,
"grad_norm": 0.44325180615966636,
"learning_rate": 1.133687281133331e-06,
"loss": 0.0791,
"step": 619
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.5025359478656346,
"learning_rate": 1.1202805047656406e-06,
"loss": 0.0707,
"step": 620
},
{
"epoch": 2.3884615384615384,
"grad_norm": 0.4057501287460486,
"learning_rate": 1.1069434653570633e-06,
"loss": 0.0557,
"step": 621
},
{
"epoch": 2.3923076923076922,
"grad_norm": 0.41886340424455265,
"learning_rate": 1.0936764026374547e-06,
"loss": 0.0708,
"step": 622
},
{
"epoch": 2.396153846153846,
"grad_norm": 0.43469512581455305,
"learning_rate": 1.0804795550788473e-06,
"loss": 0.0672,
"step": 623
},
{
"epoch": 2.4,
"grad_norm": 0.4704489982815487,
"learning_rate": 1.0673531598911824e-06,
"loss": 0.0645,
"step": 624
},
{
"epoch": 2.4038461538461537,
"grad_norm": 0.398996906821157,
"learning_rate": 1.0542974530180327e-06,
"loss": 0.0605,
"step": 625
},
{
"epoch": 2.4076923076923076,
"grad_norm": 0.43819142825978763,
"learning_rate": 1.0413126691323667e-06,
"loss": 0.0837,
"step": 626
},
{
"epoch": 2.4115384615384614,
"grad_norm": 0.4060412921814134,
"learning_rate": 1.0283990416323336e-06,
"loss": 0.0635,
"step": 627
},
{
"epoch": 2.4153846153846152,
"grad_norm": 0.46829338641849055,
"learning_rate": 1.0155568026370637e-06,
"loss": 0.0632,
"step": 628
},
{
"epoch": 2.419230769230769,
"grad_norm": 0.5373321128668832,
"learning_rate": 1.0027861829824953e-06,
"loss": 0.0646,
"step": 629
},
{
"epoch": 2.423076923076923,
"grad_norm": 0.4401341491958945,
"learning_rate": 9.900874122172294e-07,
"loss": 0.0592,
"step": 630
},
{
"epoch": 2.4269230769230767,
"grad_norm": 0.45094184478719057,
"learning_rate": 9.774607185984004e-07,
"loss": 0.0699,
"step": 631
},
{
"epoch": 2.430769230769231,
"grad_norm": 0.5196273936486302,
"learning_rate": 9.649063290875771e-07,
"loss": 0.0798,
"step": 632
},
{
"epoch": 2.434615384615385,
"grad_norm": 0.4687383020027772,
"learning_rate": 9.524244693466773e-07,
"loss": 0.0605,
"step": 633
},
{
"epoch": 2.4384615384615387,
"grad_norm": 0.5146146347990106,
"learning_rate": 9.400153637339182e-07,
"loss": 0.0842,
"step": 634
},
{
"epoch": 2.4423076923076925,
"grad_norm": 0.40260626082139755,
"learning_rate": 9.276792352997782e-07,
"loss": 0.0556,
"step": 635
},
{
"epoch": 2.4461538461538463,
"grad_norm": 0.4481826256236517,
"learning_rate": 9.154163057829879e-07,
"loss": 0.0761,
"step": 636
},
{
"epoch": 2.45,
"grad_norm": 0.41944155435055575,
"learning_rate": 9.032267956065516e-07,
"loss": 0.0676,
"step": 637
},
{
"epoch": 2.453846153846154,
"grad_norm": 0.4383860163400684,
"learning_rate": 8.911109238737748e-07,
"loss": 0.0529,
"step": 638
},
{
"epoch": 2.457692307692308,
"grad_norm": 0.4645613732345803,
"learning_rate": 8.790689083643328e-07,
"loss": 0.0574,
"step": 639
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.35603715523267637,
"learning_rate": 8.671009655303531e-07,
"loss": 0.0518,
"step": 640
},
{
"epoch": 2.4653846153846155,
"grad_norm": 0.484911791946085,
"learning_rate": 8.552073104925296e-07,
"loss": 0.0756,
"step": 641
},
{
"epoch": 2.4692307692307693,
"grad_norm": 0.4448835245403103,
"learning_rate": 8.433881570362484e-07,
"loss": 0.0584,
"step": 642
},
{
"epoch": 2.473076923076923,
"grad_norm": 0.5020188985002484,
"learning_rate": 8.316437176077491e-07,
"loss": 0.0616,
"step": 643
},
{
"epoch": 2.476923076923077,
"grad_norm": 0.5200167683308845,
"learning_rate": 8.199742033103091e-07,
"loss": 0.062,
"step": 644
},
{
"epoch": 2.480769230769231,
"grad_norm": 0.4573430233163218,
"learning_rate": 8.083798239004408e-07,
"loss": 0.0712,
"step": 645
},
{
"epoch": 2.4846153846153847,
"grad_norm": 0.49496718950274476,
"learning_rate": 7.968607877841333e-07,
"loss": 0.0964,
"step": 646
},
{
"epoch": 2.4884615384615385,
"grad_norm": 0.4215958009957212,
"learning_rate": 7.854173020130906e-07,
"loss": 0.063,
"step": 647
},
{
"epoch": 2.4923076923076923,
"grad_norm": 0.5127374763344157,
"learning_rate": 7.740495722810271e-07,
"loss": 0.0683,
"step": 648
},
{
"epoch": 2.496153846153846,
"grad_norm": 0.42358687784643495,
"learning_rate": 7.627578029199562e-07,
"loss": 0.0552,
"step": 649
},
{
"epoch": 2.5,
"grad_norm": 0.5225793858462355,
"learning_rate": 7.515421968965242e-07,
"loss": 0.0492,
"step": 650
},
{
"epoch": 2.503846153846154,
"grad_norm": 0.4416738566056911,
"learning_rate": 7.404029558083653e-07,
"loss": 0.0701,
"step": 651
},
{
"epoch": 2.5076923076923077,
"grad_norm": 0.5069647152670649,
"learning_rate": 7.293402798804667e-07,
"loss": 0.0638,
"step": 652
},
{
"epoch": 2.5115384615384615,
"grad_norm": 0.4687887382968979,
"learning_rate": 7.183543679615834e-07,
"loss": 0.1321,
"step": 653
},
{
"epoch": 2.5153846153846153,
"grad_norm": 0.44347745384195286,
"learning_rate": 7.074454175206524e-07,
"loss": 0.0741,
"step": 654
},
{
"epoch": 2.519230769230769,
"grad_norm": 0.5113531173675347,
"learning_rate": 6.966136246432492e-07,
"loss": 0.0782,
"step": 655
},
{
"epoch": 2.523076923076923,
"grad_norm": 0.4407843471989148,
"learning_rate": 6.858591840280627e-07,
"loss": 0.073,
"step": 656
},
{
"epoch": 2.526923076923077,
"grad_norm": 0.4020139795894034,
"learning_rate": 6.751822889833926e-07,
"loss": 0.0573,
"step": 657
},
{
"epoch": 2.5307692307692307,
"grad_norm": 0.4465340811083204,
"learning_rate": 6.645831314236817e-07,
"loss": 0.0543,
"step": 658
},
{
"epoch": 2.5346153846153845,
"grad_norm": 0.5171561487439342,
"learning_rate": 6.540619018660555e-07,
"loss": 0.0543,
"step": 659
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.46741646213340754,
"learning_rate": 6.436187894269086e-07,
"loss": 0.0656,
"step": 660
},
{
"epoch": 2.542307692307692,
"grad_norm": 0.4793423571464268,
"learning_rate": 6.332539818184985e-07,
"loss": 0.056,
"step": 661
},
{
"epoch": 2.546153846153846,
"grad_norm": 0.48347475686488794,
"learning_rate": 6.229676653455719e-07,
"loss": 0.0744,
"step": 662
},
{
"epoch": 2.55,
"grad_norm": 0.5315080233456347,
"learning_rate": 6.127600249020216e-07,
"loss": 0.0623,
"step": 663
},
{
"epoch": 2.5538461538461537,
"grad_norm": 0.4851407436009277,
"learning_rate": 6.026312439675553e-07,
"loss": 0.0758,
"step": 664
},
{
"epoch": 2.5576923076923075,
"grad_norm": 0.4360177713384227,
"learning_rate": 5.925815046044026e-07,
"loss": 0.0663,
"step": 665
},
{
"epoch": 2.5615384615384613,
"grad_norm": 0.41429518201674725,
"learning_rate": 5.826109874540409e-07,
"loss": 0.069,
"step": 666
},
{
"epoch": 2.565384615384615,
"grad_norm": 0.42237116339680864,
"learning_rate": 5.727198717339511e-07,
"loss": 0.0517,
"step": 667
},
{
"epoch": 2.569230769230769,
"grad_norm": 0.42524552678029826,
"learning_rate": 5.6290833523439e-07,
"loss": 0.0686,
"step": 668
},
{
"epoch": 2.573076923076923,
"grad_norm": 0.47443992757608855,
"learning_rate": 5.531765543152002e-07,
"loss": 0.0519,
"step": 669
},
{
"epoch": 2.5769230769230766,
"grad_norm": 0.39671120076980904,
"learning_rate": 5.435247039026398e-07,
"loss": 0.0553,
"step": 670
},
{
"epoch": 2.580769230769231,
"grad_norm": 0.4818384831578317,
"learning_rate": 5.33952957486234e-07,
"loss": 0.0558,
"step": 671
},
{
"epoch": 2.5846153846153848,
"grad_norm": 0.4157290446719997,
"learning_rate": 5.244614871156612e-07,
"loss": 0.0578,
"step": 672
},
{
"epoch": 2.5884615384615386,
"grad_norm": 0.5219405243168577,
"learning_rate": 5.150504633976572e-07,
"loss": 0.0753,
"step": 673
},
{
"epoch": 2.5923076923076924,
"grad_norm": 0.45344166037267974,
"learning_rate": 5.057200554929509e-07,
"loss": 0.0699,
"step": 674
},
{
"epoch": 2.5961538461538463,
"grad_norm": 0.6406295342563705,
"learning_rate": 4.964704311132224e-07,
"loss": 0.0566,
"step": 675
},
{
"epoch": 2.6,
"grad_norm": 0.4453060496510882,
"learning_rate": 4.873017565180871e-07,
"loss": 0.0677,
"step": 676
},
{
"epoch": 2.603846153846154,
"grad_norm": 0.3908293562665863,
"learning_rate": 4.782141965121129e-07,
"loss": 0.0702,
"step": 677
},
{
"epoch": 2.6076923076923078,
"grad_norm": 0.44756978753413024,
"learning_rate": 4.6920791444184934e-07,
"loss": 0.0713,
"step": 678
},
{
"epoch": 2.6115384615384616,
"grad_norm": 0.4403934315922283,
"learning_rate": 4.602830721928997e-07,
"loss": 0.0675,
"step": 679
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.4432327973787358,
"learning_rate": 4.5143983018700485e-07,
"loss": 0.0792,
"step": 680
},
{
"epoch": 2.6192307692307693,
"grad_norm": 0.3992384101232743,
"learning_rate": 4.4267834737916295e-07,
"loss": 0.0596,
"step": 681
},
{
"epoch": 2.623076923076923,
"grad_norm": 0.48012513021919445,
"learning_rate": 4.33998781254773e-07,
"loss": 0.0663,
"step": 682
},
{
"epoch": 2.626923076923077,
"grad_norm": 0.4085728658283165,
"learning_rate": 4.2540128782679934e-07,
"loss": 0.0624,
"step": 683
},
{
"epoch": 2.6307692307692307,
"grad_norm": 0.5052469302836516,
"learning_rate": 4.1688602163297564e-07,
"loss": 0.0558,
"step": 684
},
{
"epoch": 2.6346153846153846,
"grad_norm": 0.45030495174256574,
"learning_rate": 4.0845313573301736e-07,
"loss": 0.0706,
"step": 685
},
{
"epoch": 2.6384615384615384,
"grad_norm": 0.42957054356358493,
"learning_rate": 4.001027817058789e-07,
"loss": 0.0626,
"step": 686
},
{
"epoch": 2.6423076923076922,
"grad_norm": 0.4677109023171871,
"learning_rate": 3.9183510964702463e-07,
"loss": 0.0597,
"step": 687
},
{
"epoch": 2.646153846153846,
"grad_norm": 0.3845143600282186,
"learning_rate": 3.836502681657289e-07,
"loss": 0.052,
"step": 688
},
{
"epoch": 2.65,
"grad_norm": 0.44759299621353943,
"learning_rate": 3.755484043824131e-07,
"loss": 0.0554,
"step": 689
},
{
"epoch": 2.6538461538461537,
"grad_norm": 0.3549672762549962,
"learning_rate": 3.675296639259912e-07,
"loss": 0.0616,
"step": 690
},
{
"epoch": 2.6576923076923076,
"grad_norm": 0.43040437533263826,
"learning_rate": 3.595941909312595e-07,
"loss": 0.0654,
"step": 691
},
{
"epoch": 2.6615384615384614,
"grad_norm": 0.4476015301026374,
"learning_rate": 3.517421280363004e-07,
"loss": 0.0658,
"step": 692
},
{
"epoch": 2.6653846153846152,
"grad_norm": 0.39032699169276863,
"learning_rate": 3.439736163799251e-07,
"loss": 0.0533,
"step": 693
},
{
"epoch": 2.669230769230769,
"grad_norm": 0.4743895966377355,
"learning_rate": 3.362887955991301e-07,
"loss": 0.0618,
"step": 694
},
{
"epoch": 2.6730769230769234,
"grad_norm": 0.3380419073317233,
"learning_rate": 3.2868780382658895e-07,
"loss": 0.0755,
"step": 695
},
{
"epoch": 2.676923076923077,
"grad_norm": 0.3926027380137787,
"learning_rate": 3.2117077768817395e-07,
"loss": 0.0623,
"step": 696
},
{
"epoch": 2.680769230769231,
"grad_norm": 0.4248277803074578,
"learning_rate": 3.1373785230049356e-07,
"loss": 0.0597,
"step": 697
},
{
"epoch": 2.684615384615385,
"grad_norm": 0.4179306975033615,
"learning_rate": 3.0638916126846885e-07,
"loss": 0.0712,
"step": 698
},
{
"epoch": 2.6884615384615387,
"grad_norm": 0.4565736182349653,
"learning_rate": 2.991248366829291e-07,
"loss": 0.0605,
"step": 699
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.6855299765624544,
"learning_rate": 2.91945009118238e-07,
"loss": 0.0742,
"step": 700
},
{
"epoch": 2.6961538461538463,
"grad_norm": 0.37160938906793123,
"learning_rate": 2.848498076299483e-07,
"loss": 0.0692,
"step": 701
},
{
"epoch": 2.7,
"grad_norm": 0.49033072228973207,
"learning_rate": 2.7783935975247867e-07,
"loss": 0.0655,
"step": 702
},
{
"epoch": 2.703846153846154,
"grad_norm": 0.4184867250481062,
"learning_rate": 2.7091379149682683e-07,
"loss": 0.0584,
"step": 703
},
{
"epoch": 2.707692307692308,
"grad_norm": 0.41556422702895396,
"learning_rate": 2.6407322734829763e-07,
"loss": 0.0623,
"step": 704
},
{
"epoch": 2.7115384615384617,
"grad_norm": 0.49068641747489905,
"learning_rate": 2.573177902642726e-07,
"loss": 0.0553,
"step": 705
},
{
"epoch": 2.7153846153846155,
"grad_norm": 0.4558179430348964,
"learning_rate": 2.506476016719922e-07,
"loss": 0.0592,
"step": 706
},
{
"epoch": 2.7192307692307693,
"grad_norm": 0.43059547858891134,
"learning_rate": 2.440627814663804e-07,
"loss": 0.0641,
"step": 707
},
{
"epoch": 2.723076923076923,
"grad_norm": 0.44001521637392654,
"learning_rate": 2.3756344800788421e-07,
"loss": 0.0692,
"step": 708
},
{
"epoch": 2.726923076923077,
"grad_norm": 0.7249126375352565,
"learning_rate": 2.3114971812034981e-07,
"loss": 0.0936,
"step": 709
},
{
"epoch": 2.730769230769231,
"grad_norm": 0.42217580546626365,
"learning_rate": 2.2482170708892083e-07,
"loss": 0.0644,
"step": 710
},
{
"epoch": 2.7346153846153847,
"grad_norm": 0.44899366484062386,
"learning_rate": 2.1857952865796616e-07,
"loss": 0.0645,
"step": 711
},
{
"epoch": 2.7384615384615385,
"grad_norm": 0.4778318063037461,
"learning_rate": 2.124232950290367e-07,
"loss": 0.0481,
"step": 712
},
{
"epoch": 2.7423076923076923,
"grad_norm": 0.3802842672305909,
"learning_rate": 2.0635311685884675e-07,
"loss": 0.046,
"step": 713
},
{
"epoch": 2.746153846153846,
"grad_norm": 0.4282326635637059,
"learning_rate": 2.0036910325728521e-07,
"loss": 0.0808,
"step": 714
},
{
"epoch": 2.75,
"grad_norm": 0.4197752917116193,
"learning_rate": 1.9447136178545766e-07,
"loss": 0.0602,
"step": 715
},
{
"epoch": 2.753846153846154,
"grad_norm": 0.46675738251287097,
"learning_rate": 1.8865999845374794e-07,
"loss": 0.0435,
"step": 716
},
{
"epoch": 2.7576923076923077,
"grad_norm": 0.3699556658556475,
"learning_rate": 1.8293511771991624e-07,
"loss": 0.0783,
"step": 717
},
{
"epoch": 2.7615384615384615,
"grad_norm": 0.4960098078583109,
"learning_rate": 1.7729682248721848e-07,
"loss": 0.0669,
"step": 718
},
{
"epoch": 2.7653846153846153,
"grad_norm": 0.4299855405719279,
"learning_rate": 1.7174521410256162e-07,
"loss": 0.0551,
"step": 719
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.4394884965679465,
"learning_rate": 1.6628039235467686e-07,
"loss": 0.085,
"step": 720
},
{
"epoch": 2.773076923076923,
"grad_norm": 0.48458397058811936,
"learning_rate": 1.6090245547232707e-07,
"loss": 0.0778,
"step": 721
},
{
"epoch": 2.776923076923077,
"grad_norm": 0.42182816117515204,
"learning_rate": 1.5561150012254446e-07,
"loss": 0.0703,
"step": 722
},
{
"epoch": 2.7807692307692307,
"grad_norm": 0.4649102401838792,
"learning_rate": 1.5040762140888843e-07,
"loss": 0.0671,
"step": 723
},
{
"epoch": 2.7846153846153845,
"grad_norm": 0.4160887198965109,
"learning_rate": 1.4529091286973994e-07,
"loss": 0.0498,
"step": 724
},
{
"epoch": 2.7884615384615383,
"grad_norm": 0.4189956495776838,
"learning_rate": 1.402614664766172e-07,
"loss": 0.0627,
"step": 725
},
{
"epoch": 2.792307692307692,
"grad_norm": 0.319352198814085,
"learning_rate": 1.353193726325247e-07,
"loss": 0.0614,
"step": 726
},
{
"epoch": 2.796153846153846,
"grad_norm": 0.5202838709185891,
"learning_rate": 1.3046472017032685e-07,
"loss": 0.0695,
"step": 727
},
{
"epoch": 2.8,
"grad_norm": 0.37560385129244095,
"learning_rate": 1.2569759635115086e-07,
"loss": 0.0701,
"step": 728
},
{
"epoch": 2.8038461538461537,
"grad_norm": 0.5334475881929764,
"learning_rate": 1.210180868628219e-07,
"loss": 0.0596,
"step": 729
},
{
"epoch": 2.8076923076923075,
"grad_norm": 0.5125886019042502,
"learning_rate": 1.1642627581831767e-07,
"loss": 0.0685,
"step": 730
},
{
"epoch": 2.8115384615384613,
"grad_norm": 0.3700563230017658,
"learning_rate": 1.1192224575425848e-07,
"loss": 0.061,
"step": 731
},
{
"epoch": 2.815384615384615,
"grad_norm": 0.41940435231775963,
"learning_rate": 1.0750607762942622e-07,
"loss": 0.0658,
"step": 732
},
{
"epoch": 2.819230769230769,
"grad_norm": 0.38454225029708794,
"learning_rate": 1.0317785082330555e-07,
"loss": 0.0461,
"step": 733
},
{
"epoch": 2.823076923076923,
"grad_norm": 0.42509450235055063,
"learning_rate": 9.89376431346606e-08,
"loss": 0.0603,
"step": 734
},
{
"epoch": 2.8269230769230766,
"grad_norm": 0.42317891385504064,
"learning_rate": 9.478553078013042e-08,
"loss": 0.0568,
"step": 735
},
{
"epoch": 2.830769230769231,
"grad_norm": 0.5106931410391424,
"learning_rate": 9.072158839286748e-08,
"loss": 0.0613,
"step": 736
},
{
"epoch": 2.8346153846153848,
"grad_norm": 0.49463080380279223,
"learning_rate": 8.674588902118919e-08,
"loss": 0.2092,
"step": 737
},
{
"epoch": 2.8384615384615386,
"grad_norm": 0.4336433854162459,
"learning_rate": 8.285850412726837e-08,
"loss": 0.0621,
"step": 738
},
{
"epoch": 2.8423076923076924,
"grad_norm": 0.37514200049691104,
"learning_rate": 7.905950358584768e-08,
"loss": 0.0528,
"step": 739
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.3730901811552213,
"learning_rate": 7.534895568298395e-08,
"loss": 0.0516,
"step": 740
},
{
"epoch": 2.85,
"grad_norm": 0.41267926524997883,
"learning_rate": 7.172692711482022e-08,
"loss": 0.0711,
"step": 741
},
{
"epoch": 2.853846153846154,
"grad_norm": 0.4578685790427763,
"learning_rate": 6.819348298638839e-08,
"loss": 0.0623,
"step": 742
},
{
"epoch": 2.8576923076923078,
"grad_norm": 0.7029418818640065,
"learning_rate": 6.474868681043578e-08,
"loss": 0.0622,
"step": 743
},
{
"epoch": 2.8615384615384616,
"grad_norm": 0.48710233473711223,
"learning_rate": 6.13926005062876e-08,
"loss": 0.0584,
"step": 744
},
{
"epoch": 2.8653846153846154,
"grad_norm": 0.41317067982425054,
"learning_rate": 5.8125284398730666e-08,
"loss": 0.0626,
"step": 745
},
{
"epoch": 2.8692307692307693,
"grad_norm": 0.5093697151137649,
"learning_rate": 5.4946797216931524e-08,
"loss": 0.0662,
"step": 746
},
{
"epoch": 2.873076923076923,
"grad_norm": 0.40757130466275254,
"learning_rate": 5.185719609337836e-08,
"loss": 0.0631,
"step": 747
},
{
"epoch": 2.876923076923077,
"grad_norm": 0.4321713509379994,
"learning_rate": 4.885653656285627e-08,
"loss": 0.0629,
"step": 748
},
{
"epoch": 2.8807692307692307,
"grad_norm": 0.573596300985358,
"learning_rate": 4.5944872561448084e-08,
"loss": 0.0704,
"step": 749
},
{
"epoch": 2.8846153846153846,
"grad_norm": 0.41883721735875484,
"learning_rate": 4.3122256425563444e-08,
"loss": 0.0613,
"step": 750
},
{
"epoch": 2.8884615384615384,
"grad_norm": 0.3988000631350868,
"learning_rate": 4.038873889100237e-08,
"loss": 0.0684,
"step": 751
},
{
"epoch": 2.8923076923076922,
"grad_norm": 0.5421953778564682,
"learning_rate": 3.77443690920376e-08,
"loss": 0.0586,
"step": 752
},
{
"epoch": 2.896153846153846,
"grad_norm": 0.4426853492862627,
"learning_rate": 3.518919456053649e-08,
"loss": 0.061,
"step": 753
},
{
"epoch": 2.9,
"grad_norm": 0.4259685039160188,
"learning_rate": 3.2723261225102164e-08,
"loss": 0.0573,
"step": 754
},
{
"epoch": 2.9038461538461537,
"grad_norm": 0.41944883995141774,
"learning_rate": 3.034661341025258e-08,
"loss": 0.0599,
"step": 755
},
{
"epoch": 2.9076923076923076,
"grad_norm": 0.4996659647744855,
"learning_rate": 2.8059293835620006e-08,
"loss": 0.0615,
"step": 756
},
{
"epoch": 2.9115384615384614,
"grad_norm": 0.40948510302310037,
"learning_rate": 2.5861343615184997e-08,
"loss": 0.0609,
"step": 757
},
{
"epoch": 2.9153846153846152,
"grad_norm": 0.4588178161394445,
"learning_rate": 2.3752802256536423e-08,
"loss": 0.0689,
"step": 758
},
{
"epoch": 2.919230769230769,
"grad_norm": 0.5331401297450866,
"learning_rate": 2.173370766016314e-08,
"loss": 0.0832,
"step": 759
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.4254634822660441,
"learning_rate": 1.98040961187701e-08,
"loss": 0.0573,
"step": 760
},
{
"epoch": 2.926923076923077,
"grad_norm": 0.4566833195505137,
"learning_rate": 1.7964002316628316e-08,
"loss": 0.0736,
"step": 761
},
{
"epoch": 2.930769230769231,
"grad_norm": 0.45595057625055585,
"learning_rate": 1.6213459328950355e-08,
"loss": 0.0907,
"step": 762
},
{
"epoch": 2.934615384615385,
"grad_norm": 0.8678040878022159,
"learning_rate": 1.4552498621295264e-08,
"loss": 0.0744,
"step": 763
},
{
"epoch": 2.9384615384615387,
"grad_norm": 0.4152937290618546,
"learning_rate": 1.2981150049004021e-08,
"loss": 0.0569,
"step": 764
},
{
"epoch": 2.9423076923076925,
"grad_norm": 0.478455053506472,
"learning_rate": 1.1499441856663296e-08,
"loss": 0.0783,
"step": 765
},
{
"epoch": 2.9461538461538463,
"grad_norm": 0.4632254792093425,
"learning_rate": 1.0107400677596413e-08,
"loss": 0.0532,
"step": 766
},
{
"epoch": 2.95,
"grad_norm": 0.6097330309027501,
"learning_rate": 8.805051533384846e-09,
"loss": 0.0656,
"step": 767
},
{
"epoch": 2.953846153846154,
"grad_norm": 0.5036451954014907,
"learning_rate": 7.59241783341913e-09,
"loss": 0.0734,
"step": 768
},
{
"epoch": 2.957692307692308,
"grad_norm": 0.5606890776908487,
"learning_rate": 6.469521374477539e-09,
"loss": 0.0662,
"step": 769
},
{
"epoch": 2.9615384615384617,
"grad_norm": 0.4611275355710929,
"learning_rate": 5.436382340335833e-09,
"loss": 0.0667,
"step": 770
},
{
"epoch": 2.9653846153846155,
"grad_norm": 0.45309916819777246,
"learning_rate": 4.493019301401447e-09,
"loss": 0.0546,
"step": 771
},
{
"epoch": 2.9692307692307693,
"grad_norm": 0.4490484721832489,
"learning_rate": 3.6394492143820847e-09,
"loss": 0.071,
"step": 772
},
{
"epoch": 2.973076923076923,
"grad_norm": 0.5100845117140136,
"learning_rate": 2.875687421980966e-09,
"loss": 0.0703,
"step": 773
},
{
"epoch": 2.976923076923077,
"grad_norm": 0.39733724968330264,
"learning_rate": 2.201747652618713e-09,
"loss": 0.0675,
"step": 774
},
{
"epoch": 2.980769230769231,
"grad_norm": 0.5321838753795445,
"learning_rate": 1.6176420201902132e-09,
"loss": 0.0628,
"step": 775
},
{
"epoch": 2.9846153846153847,
"grad_norm": 0.3692650504722248,
"learning_rate": 1.1233810238425735e-09,
"loss": 0.0505,
"step": 776
},
{
"epoch": 2.9884615384615385,
"grad_norm": 0.43167989777734067,
"learning_rate": 7.189735477913795e-10,
"loss": 0.0733,
"step": 777
},
{
"epoch": 2.9923076923076923,
"grad_norm": 0.46822506342648357,
"learning_rate": 4.0442686115582665e-10,
"loss": 0.0636,
"step": 778
},
{
"epoch": 2.996153846153846,
"grad_norm": 0.388302049103191,
"learning_rate": 1.797466178327101e-10,
"loss": 0.0652,
"step": 779
},
{
"epoch": 3.0,
"grad_norm": 0.39178542173282355,
"learning_rate": 4.4936856390398465e-11,
"loss": 0.0631,
"step": 780
},
{
"epoch": 3.0,
"step": 780,
"total_flos": 4.4809696107823104e+17,
"train_loss": 0.15228378044871183,
"train_runtime": 27691.5469,
"train_samples_per_second": 0.901,
"train_steps_per_second": 0.028
}
],
"logging_steps": 1,
"max_steps": 780,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 4.4809696107823104e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}