chitanda's picture
Add files using upload-large-folder tool
4aa0271 verified
raw
history blame
36.2 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.5769230769230769,
"eval_steps": 500,
"global_step": 300,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 6.25e-07,
"loss": 3.1205,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 1.25e-06,
"loss": 3.2041,
"step": 2
},
{
"epoch": 0.01,
"learning_rate": 1.8750000000000003e-06,
"loss": 3.0927,
"step": 3
},
{
"epoch": 0.01,
"learning_rate": 2.5e-06,
"loss": 3.1433,
"step": 4
},
{
"epoch": 0.01,
"learning_rate": 3.125e-06,
"loss": 3.036,
"step": 5
},
{
"epoch": 0.01,
"learning_rate": 3.7500000000000005e-06,
"loss": 2.4568,
"step": 6
},
{
"epoch": 0.01,
"learning_rate": 4.3750000000000005e-06,
"loss": 1.2898,
"step": 7
},
{
"epoch": 0.02,
"learning_rate": 5e-06,
"loss": 0.97,
"step": 8
},
{
"epoch": 0.02,
"learning_rate": 5.625e-06,
"loss": 0.6386,
"step": 9
},
{
"epoch": 0.02,
"learning_rate": 6.25e-06,
"loss": 0.4549,
"step": 10
},
{
"epoch": 0.02,
"learning_rate": 6.875e-06,
"loss": 0.4718,
"step": 11
},
{
"epoch": 0.02,
"learning_rate": 7.500000000000001e-06,
"loss": 0.4106,
"step": 12
},
{
"epoch": 0.03,
"learning_rate": 8.125000000000001e-06,
"loss": 0.3442,
"step": 13
},
{
"epoch": 0.03,
"learning_rate": 8.750000000000001e-06,
"loss": 0.3178,
"step": 14
},
{
"epoch": 0.03,
"learning_rate": 9.375000000000001e-06,
"loss": 0.2579,
"step": 15
},
{
"epoch": 0.03,
"learning_rate": 1e-05,
"loss": 0.3623,
"step": 16
},
{
"epoch": 0.03,
"learning_rate": 1.0625e-05,
"loss": 0.3338,
"step": 17
},
{
"epoch": 0.03,
"learning_rate": 1.125e-05,
"loss": 0.3265,
"step": 18
},
{
"epoch": 0.04,
"learning_rate": 1.1875e-05,
"loss": 0.3804,
"step": 19
},
{
"epoch": 0.04,
"learning_rate": 1.25e-05,
"loss": 0.2543,
"step": 20
},
{
"epoch": 0.04,
"learning_rate": 1.3125e-05,
"loss": 0.3619,
"step": 21
},
{
"epoch": 0.04,
"learning_rate": 1.375e-05,
"loss": 0.3095,
"step": 22
},
{
"epoch": 0.04,
"learning_rate": 1.4375e-05,
"loss": 0.2835,
"step": 23
},
{
"epoch": 0.05,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.2666,
"step": 24
},
{
"epoch": 0.05,
"learning_rate": 1.5625e-05,
"loss": 0.3076,
"step": 25
},
{
"epoch": 0.05,
"learning_rate": 1.6250000000000002e-05,
"loss": 0.2789,
"step": 26
},
{
"epoch": 0.05,
"learning_rate": 1.6875e-05,
"loss": 0.3008,
"step": 27
},
{
"epoch": 0.05,
"learning_rate": 1.7500000000000002e-05,
"loss": 0.2998,
"step": 28
},
{
"epoch": 0.06,
"learning_rate": 1.8125e-05,
"loss": 0.272,
"step": 29
},
{
"epoch": 0.06,
"learning_rate": 1.8750000000000002e-05,
"loss": 0.371,
"step": 30
},
{
"epoch": 0.06,
"learning_rate": 1.9375e-05,
"loss": 0.3543,
"step": 31
},
{
"epoch": 0.06,
"learning_rate": 2e-05,
"loss": 0.2793,
"step": 32
},
{
"epoch": 0.06,
"learning_rate": 1.9999951432210905e-05,
"loss": 0.278,
"step": 33
},
{
"epoch": 0.07,
"learning_rate": 1.9999805729315383e-05,
"loss": 0.3078,
"step": 34
},
{
"epoch": 0.07,
"learning_rate": 1.999956289272873e-05,
"loss": 0.3137,
"step": 35
},
{
"epoch": 0.07,
"learning_rate": 1.999922292480975e-05,
"loss": 0.2804,
"step": 36
},
{
"epoch": 0.07,
"learning_rate": 1.9998785828860744e-05,
"loss": 0.2322,
"step": 37
},
{
"epoch": 0.07,
"learning_rate": 1.9998251609127465e-05,
"loss": 0.2754,
"step": 38
},
{
"epoch": 0.07,
"learning_rate": 1.999762027079909e-05,
"loss": 0.3334,
"step": 39
},
{
"epoch": 0.08,
"learning_rate": 1.9996891820008165e-05,
"loss": 0.3721,
"step": 40
},
{
"epoch": 0.08,
"learning_rate": 1.9996066263830533e-05,
"loss": 0.3324,
"step": 41
},
{
"epoch": 0.08,
"learning_rate": 1.9995143610285275e-05,
"loss": 0.2972,
"step": 42
},
{
"epoch": 0.08,
"learning_rate": 1.9994123868334655e-05,
"loss": 0.3039,
"step": 43
},
{
"epoch": 0.08,
"learning_rate": 1.9993007047883988e-05,
"loss": 0.2436,
"step": 44
},
{
"epoch": 0.09,
"learning_rate": 1.999179315978157e-05,
"loss": 0.3353,
"step": 45
},
{
"epoch": 0.09,
"learning_rate": 1.999048221581858e-05,
"loss": 0.2819,
"step": 46
},
{
"epoch": 0.09,
"learning_rate": 1.9989074228728942e-05,
"loss": 0.2769,
"step": 47
},
{
"epoch": 0.09,
"learning_rate": 1.9987569212189224e-05,
"loss": 0.253,
"step": 48
},
{
"epoch": 0.09,
"learning_rate": 1.9985967180818493e-05,
"loss": 0.2614,
"step": 49
},
{
"epoch": 0.1,
"learning_rate": 1.998426815017817e-05,
"loss": 0.3297,
"step": 50
},
{
"epoch": 0.1,
"learning_rate": 1.998247213677188e-05,
"loss": 0.3112,
"step": 51
},
{
"epoch": 0.1,
"learning_rate": 1.9980579158045322e-05,
"loss": 0.2611,
"step": 52
},
{
"epoch": 0.1,
"learning_rate": 1.9978589232386036e-05,
"loss": 0.3002,
"step": 53
},
{
"epoch": 0.1,
"learning_rate": 1.997650237912329e-05,
"loss": 0.2485,
"step": 54
},
{
"epoch": 0.11,
"learning_rate": 1.997431861852785e-05,
"loss": 0.287,
"step": 55
},
{
"epoch": 0.11,
"learning_rate": 1.9972037971811802e-05,
"loss": 0.331,
"step": 56
},
{
"epoch": 0.11,
"learning_rate": 1.996966046112834e-05,
"loss": 0.2785,
"step": 57
},
{
"epoch": 0.11,
"learning_rate": 1.996718610957155e-05,
"loss": 0.3088,
"step": 58
},
{
"epoch": 0.11,
"learning_rate": 1.9964614941176194e-05,
"loss": 0.2558,
"step": 59
},
{
"epoch": 0.12,
"learning_rate": 1.9961946980917457e-05,
"loss": 0.3732,
"step": 60
},
{
"epoch": 0.12,
"learning_rate": 1.995918225471073e-05,
"loss": 0.2935,
"step": 61
},
{
"epoch": 0.12,
"learning_rate": 1.9956320789411338e-05,
"loss": 0.2946,
"step": 62
},
{
"epoch": 0.12,
"learning_rate": 1.9953362612814294e-05,
"loss": 0.2806,
"step": 63
},
{
"epoch": 0.12,
"learning_rate": 1.9950307753654016e-05,
"loss": 0.2934,
"step": 64
},
{
"epoch": 0.12,
"learning_rate": 1.994715624160405e-05,
"loss": 0.3366,
"step": 65
},
{
"epoch": 0.13,
"learning_rate": 1.99439081072768e-05,
"loss": 0.2754,
"step": 66
},
{
"epoch": 0.13,
"learning_rate": 1.9940563382223196e-05,
"loss": 0.3244,
"step": 67
},
{
"epoch": 0.13,
"learning_rate": 1.9937122098932428e-05,
"loss": 0.3061,
"step": 68
},
{
"epoch": 0.13,
"learning_rate": 1.9933584290831593e-05,
"loss": 0.4533,
"step": 69
},
{
"epoch": 0.13,
"learning_rate": 1.9929949992285397e-05,
"loss": 1.2644,
"step": 70
},
{
"epoch": 0.14,
"learning_rate": 1.992621923859581e-05,
"loss": 0.3184,
"step": 71
},
{
"epoch": 0.14,
"learning_rate": 1.9922392066001724e-05,
"loss": 0.2933,
"step": 72
},
{
"epoch": 0.14,
"learning_rate": 1.99184685116786e-05,
"loss": 0.3077,
"step": 73
},
{
"epoch": 0.14,
"learning_rate": 1.9914448613738107e-05,
"loss": 0.3038,
"step": 74
},
{
"epoch": 0.14,
"learning_rate": 1.991033241122776e-05,
"loss": 0.3063,
"step": 75
},
{
"epoch": 0.15,
"learning_rate": 1.9906119944130527e-05,
"loss": 0.311,
"step": 76
},
{
"epoch": 0.15,
"learning_rate": 1.9901811253364458e-05,
"loss": 0.3081,
"step": 77
},
{
"epoch": 0.15,
"learning_rate": 1.9897406380782262e-05,
"loss": 0.261,
"step": 78
},
{
"epoch": 0.15,
"learning_rate": 1.989290536917093e-05,
"loss": 0.3248,
"step": 79
},
{
"epoch": 0.15,
"learning_rate": 1.9888308262251286e-05,
"loss": 0.2814,
"step": 80
},
{
"epoch": 0.16,
"learning_rate": 1.988361510467761e-05,
"loss": 0.2683,
"step": 81
},
{
"epoch": 0.16,
"learning_rate": 1.9878825942037147e-05,
"loss": 0.2755,
"step": 82
},
{
"epoch": 0.16,
"learning_rate": 1.9873940820849714e-05,
"loss": 0.3489,
"step": 83
},
{
"epoch": 0.16,
"learning_rate": 1.9868959788567213e-05,
"loss": 0.3311,
"step": 84
},
{
"epoch": 0.16,
"learning_rate": 1.9863882893573188e-05,
"loss": 0.3237,
"step": 85
},
{
"epoch": 0.17,
"learning_rate": 1.985871018518236e-05,
"loss": 0.3344,
"step": 86
},
{
"epoch": 0.17,
"learning_rate": 1.9853441713640123e-05,
"loss": 0.3195,
"step": 87
},
{
"epoch": 0.17,
"learning_rate": 1.9848077530122083e-05,
"loss": 0.3099,
"step": 88
},
{
"epoch": 0.17,
"learning_rate": 1.9842617686733546e-05,
"loss": 0.271,
"step": 89
},
{
"epoch": 0.17,
"learning_rate": 1.9837062236509013e-05,
"loss": 0.2612,
"step": 90
},
{
"epoch": 0.17,
"learning_rate": 1.983141123341168e-05,
"loss": 0.3401,
"step": 91
},
{
"epoch": 0.18,
"learning_rate": 1.9825664732332886e-05,
"loss": 0.2479,
"step": 92
},
{
"epoch": 0.18,
"learning_rate": 1.9819822789091597e-05,
"loss": 0.2462,
"step": 93
},
{
"epoch": 0.18,
"learning_rate": 1.981388546043388e-05,
"loss": 0.3064,
"step": 94
},
{
"epoch": 0.18,
"learning_rate": 1.9807852804032306e-05,
"loss": 0.3066,
"step": 95
},
{
"epoch": 0.18,
"learning_rate": 1.9801724878485438e-05,
"loss": 0.3402,
"step": 96
},
{
"epoch": 0.19,
"learning_rate": 1.979550174331724e-05,
"loss": 0.2497,
"step": 97
},
{
"epoch": 0.19,
"learning_rate": 1.9789183458976485e-05,
"loss": 0.2513,
"step": 98
},
{
"epoch": 0.19,
"learning_rate": 1.97827700868362e-05,
"loss": 0.3157,
"step": 99
},
{
"epoch": 0.19,
"learning_rate": 1.977626168919305e-05,
"loss": 0.2776,
"step": 100
},
{
"epoch": 0.19,
"learning_rate": 1.9769658329266718e-05,
"loss": 0.3394,
"step": 101
},
{
"epoch": 0.2,
"learning_rate": 1.9762960071199334e-05,
"loss": 0.2657,
"step": 102
},
{
"epoch": 0.2,
"learning_rate": 1.9756166980054812e-05,
"loss": 0.2507,
"step": 103
},
{
"epoch": 0.2,
"learning_rate": 1.9749279121818235e-05,
"loss": 0.3201,
"step": 104
},
{
"epoch": 0.2,
"learning_rate": 1.9742296563395218e-05,
"loss": 0.2509,
"step": 105
},
{
"epoch": 0.2,
"learning_rate": 1.9735219372611232e-05,
"loss": 0.3257,
"step": 106
},
{
"epoch": 0.21,
"learning_rate": 1.9728047618210995e-05,
"loss": 0.3234,
"step": 107
},
{
"epoch": 0.21,
"learning_rate": 1.9720781369857747e-05,
"loss": 0.3496,
"step": 108
},
{
"epoch": 0.21,
"learning_rate": 1.9713420698132614e-05,
"loss": 0.2802,
"step": 109
},
{
"epoch": 0.21,
"learning_rate": 1.970596567453391e-05,
"loss": 0.2662,
"step": 110
},
{
"epoch": 0.21,
"learning_rate": 1.9698416371476434e-05,
"loss": 0.2971,
"step": 111
},
{
"epoch": 0.22,
"learning_rate": 1.969077286229078e-05,
"loss": 0.271,
"step": 112
},
{
"epoch": 0.22,
"learning_rate": 1.9683035221222617e-05,
"loss": 0.3354,
"step": 113
},
{
"epoch": 0.22,
"learning_rate": 1.9675203523431964e-05,
"loss": 0.3134,
"step": 114
},
{
"epoch": 0.22,
"learning_rate": 1.9667277844992476e-05,
"loss": 0.2734,
"step": 115
},
{
"epoch": 0.22,
"learning_rate": 1.9659258262890683e-05,
"loss": 0.273,
"step": 116
},
{
"epoch": 0.23,
"learning_rate": 1.9651144855025265e-05,
"loss": 0.2422,
"step": 117
},
{
"epoch": 0.23,
"learning_rate": 1.964293770020628e-05,
"loss": 0.3048,
"step": 118
},
{
"epoch": 0.23,
"learning_rate": 1.9634636878154393e-05,
"loss": 0.3284,
"step": 119
},
{
"epoch": 0.23,
"learning_rate": 1.962624246950012e-05,
"loss": 0.289,
"step": 120
},
{
"epoch": 0.23,
"learning_rate": 1.9617754555783045e-05,
"loss": 0.2867,
"step": 121
},
{
"epoch": 0.23,
"learning_rate": 1.9609173219450998e-05,
"loss": 0.2646,
"step": 122
},
{
"epoch": 0.24,
"learning_rate": 1.960049854385929e-05,
"loss": 0.3302,
"step": 123
},
{
"epoch": 0.24,
"learning_rate": 1.9591730613269878e-05,
"loss": 0.3099,
"step": 124
},
{
"epoch": 0.24,
"learning_rate": 1.9582869512850576e-05,
"loss": 0.2953,
"step": 125
},
{
"epoch": 0.24,
"learning_rate": 1.957391532867418e-05,
"loss": 0.3058,
"step": 126
},
{
"epoch": 0.24,
"learning_rate": 1.956486814771769e-05,
"loss": 0.2787,
"step": 127
},
{
"epoch": 0.25,
"learning_rate": 1.955572805786141e-05,
"loss": 0.2357,
"step": 128
},
{
"epoch": 0.25,
"learning_rate": 1.9546495147888134e-05,
"loss": 0.3695,
"step": 129
},
{
"epoch": 0.25,
"learning_rate": 1.953716950748227e-05,
"loss": 0.2825,
"step": 130
},
{
"epoch": 0.25,
"learning_rate": 1.9527751227228964e-05,
"loss": 0.279,
"step": 131
},
{
"epoch": 0.25,
"learning_rate": 1.9518240398613226e-05,
"loss": 0.2977,
"step": 132
},
{
"epoch": 0.26,
"learning_rate": 1.9508637114019037e-05,
"loss": 0.3032,
"step": 133
},
{
"epoch": 0.26,
"learning_rate": 1.9498941466728462e-05,
"loss": 0.2568,
"step": 134
},
{
"epoch": 0.26,
"learning_rate": 1.9489153550920726e-05,
"loss": 0.3435,
"step": 135
},
{
"epoch": 0.26,
"learning_rate": 1.947927346167132e-05,
"loss": 0.2819,
"step": 136
},
{
"epoch": 0.26,
"learning_rate": 1.946930129495106e-05,
"loss": 0.2415,
"step": 137
},
{
"epoch": 0.27,
"learning_rate": 1.945923714762516e-05,
"loss": 0.3235,
"step": 138
},
{
"epoch": 0.27,
"learning_rate": 1.9449081117452304e-05,
"loss": 0.2474,
"step": 139
},
{
"epoch": 0.27,
"learning_rate": 1.9438833303083677e-05,
"loss": 0.2894,
"step": 140
},
{
"epoch": 0.27,
"learning_rate": 1.9428493804062013e-05,
"loss": 0.3165,
"step": 141
},
{
"epoch": 0.27,
"learning_rate": 1.9418062720820636e-05,
"loss": 0.2488,
"step": 142
},
{
"epoch": 0.28,
"learning_rate": 1.9407540154682473e-05,
"loss": 0.2624,
"step": 143
},
{
"epoch": 0.28,
"learning_rate": 1.9396926207859085e-05,
"loss": 0.3221,
"step": 144
},
{
"epoch": 0.28,
"learning_rate": 1.9386220983449652e-05,
"loss": 0.2695,
"step": 145
},
{
"epoch": 0.28,
"learning_rate": 1.9375424585439994e-05,
"loss": 0.2303,
"step": 146
},
{
"epoch": 0.28,
"learning_rate": 1.9364537118701542e-05,
"loss": 0.2722,
"step": 147
},
{
"epoch": 0.28,
"learning_rate": 1.935355868899034e-05,
"loss": 0.3543,
"step": 148
},
{
"epoch": 0.29,
"learning_rate": 1.9342489402945997e-05,
"loss": 0.2498,
"step": 149
},
{
"epoch": 0.29,
"learning_rate": 1.9331329368090664e-05,
"loss": 0.2593,
"step": 150
},
{
"epoch": 0.29,
"learning_rate": 1.932007869282799e-05,
"loss": 0.3057,
"step": 151
},
{
"epoch": 0.29,
"learning_rate": 1.9308737486442045e-05,
"loss": 0.2861,
"step": 152
},
{
"epoch": 0.29,
"learning_rate": 1.9297305859096305e-05,
"loss": 0.2738,
"step": 153
},
{
"epoch": 0.3,
"learning_rate": 1.9285783921832537e-05,
"loss": 0.2727,
"step": 154
},
{
"epoch": 0.3,
"learning_rate": 1.927417178656975e-05,
"loss": 0.2787,
"step": 155
},
{
"epoch": 0.3,
"learning_rate": 1.926246956610309e-05,
"loss": 0.3542,
"step": 156
},
{
"epoch": 0.3,
"learning_rate": 1.9250677374102752e-05,
"loss": 0.2699,
"step": 157
},
{
"epoch": 0.3,
"learning_rate": 1.9238795325112867e-05,
"loss": 0.3191,
"step": 158
},
{
"epoch": 0.31,
"learning_rate": 1.9226823534550418e-05,
"loss": 0.3212,
"step": 159
},
{
"epoch": 0.31,
"learning_rate": 1.921476211870408e-05,
"loss": 0.3175,
"step": 160
},
{
"epoch": 0.31,
"learning_rate": 1.9202611194733107e-05,
"loss": 0.3035,
"step": 161
},
{
"epoch": 0.31,
"learning_rate": 1.9190370880666206e-05,
"loss": 0.2824,
"step": 162
},
{
"epoch": 0.31,
"learning_rate": 1.9178041295400383e-05,
"loss": 0.3072,
"step": 163
},
{
"epoch": 0.32,
"learning_rate": 1.9165622558699763e-05,
"loss": 0.222,
"step": 164
},
{
"epoch": 0.32,
"learning_rate": 1.9153114791194475e-05,
"loss": 0.2342,
"step": 165
},
{
"epoch": 0.32,
"learning_rate": 1.9140518114379433e-05,
"loss": 0.2743,
"step": 166
},
{
"epoch": 0.32,
"learning_rate": 1.912783265061319e-05,
"loss": 0.3544,
"step": 167
},
{
"epoch": 0.32,
"learning_rate": 1.9115058523116734e-05,
"loss": 0.2701,
"step": 168
},
{
"epoch": 0.33,
"learning_rate": 1.9102195855972287e-05,
"loss": 0.2766,
"step": 169
},
{
"epoch": 0.33,
"learning_rate": 1.908924477412211e-05,
"loss": 0.2867,
"step": 170
},
{
"epoch": 0.33,
"learning_rate": 1.9076205403367287e-05,
"loss": 0.2272,
"step": 171
},
{
"epoch": 0.33,
"learning_rate": 1.9063077870366504e-05,
"loss": 0.2459,
"step": 172
},
{
"epoch": 0.33,
"learning_rate": 1.90498623026348e-05,
"loss": 0.319,
"step": 173
},
{
"epoch": 0.33,
"learning_rate": 1.903655882854237e-05,
"loss": 0.258,
"step": 174
},
{
"epoch": 0.34,
"learning_rate": 1.9023167577313267e-05,
"loss": 0.307,
"step": 175
},
{
"epoch": 0.34,
"learning_rate": 1.900968867902419e-05,
"loss": 0.3004,
"step": 176
},
{
"epoch": 0.34,
"learning_rate": 1.8996122264603202e-05,
"loss": 0.2587,
"step": 177
},
{
"epoch": 0.34,
"learning_rate": 1.898246846582844e-05,
"loss": 0.2956,
"step": 178
},
{
"epoch": 0.34,
"learning_rate": 1.8968727415326885e-05,
"loss": 0.3749,
"step": 179
},
{
"epoch": 0.35,
"learning_rate": 1.895489924657301e-05,
"loss": 0.3463,
"step": 180
},
{
"epoch": 0.35,
"learning_rate": 1.894098409388754e-05,
"loss": 0.3514,
"step": 181
},
{
"epoch": 0.35,
"learning_rate": 1.8926982092436117e-05,
"loss": 0.2839,
"step": 182
},
{
"epoch": 0.35,
"learning_rate": 1.8912893378227984e-05,
"loss": 0.2168,
"step": 183
},
{
"epoch": 0.35,
"learning_rate": 1.8898718088114688e-05,
"loss": 0.2748,
"step": 184
},
{
"epoch": 0.36,
"learning_rate": 1.8884456359788725e-05,
"loss": 0.2885,
"step": 185
},
{
"epoch": 0.36,
"learning_rate": 1.887010833178222e-05,
"loss": 0.2699,
"step": 186
},
{
"epoch": 0.36,
"learning_rate": 1.8855674143465567e-05,
"loss": 0.3133,
"step": 187
},
{
"epoch": 0.36,
"learning_rate": 1.8841153935046098e-05,
"loss": 0.2773,
"step": 188
},
{
"epoch": 0.36,
"learning_rate": 1.8826547847566692e-05,
"loss": 0.3472,
"step": 189
},
{
"epoch": 0.37,
"learning_rate": 1.8811856022904423e-05,
"loss": 0.3342,
"step": 190
},
{
"epoch": 0.37,
"learning_rate": 1.8797078603769184e-05,
"loss": 0.2864,
"step": 191
},
{
"epoch": 0.37,
"learning_rate": 1.8782215733702286e-05,
"loss": 0.3115,
"step": 192
},
{
"epoch": 0.37,
"learning_rate": 1.876726755707508e-05,
"loss": 0.2726,
"step": 193
},
{
"epoch": 0.37,
"learning_rate": 1.8752234219087538e-05,
"loss": 0.3033,
"step": 194
},
{
"epoch": 0.38,
"learning_rate": 1.8737115865766865e-05,
"loss": 0.2785,
"step": 195
},
{
"epoch": 0.38,
"learning_rate": 1.8721912643966055e-05,
"loss": 0.3222,
"step": 196
},
{
"epoch": 0.38,
"learning_rate": 1.8706624701362485e-05,
"loss": 0.2268,
"step": 197
},
{
"epoch": 0.38,
"learning_rate": 1.8691252186456465e-05,
"loss": 0.2259,
"step": 198
},
{
"epoch": 0.38,
"learning_rate": 1.8675795248569816e-05,
"loss": 0.2777,
"step": 199
},
{
"epoch": 0.38,
"learning_rate": 1.866025403784439e-05,
"loss": 0.2765,
"step": 200
},
{
"epoch": 0.39,
"learning_rate": 1.8644628705240636e-05,
"loss": 0.3088,
"step": 201
},
{
"epoch": 0.39,
"learning_rate": 1.862891940253613e-05,
"loss": 0.2766,
"step": 202
},
{
"epoch": 0.39,
"learning_rate": 1.8613126282324092e-05,
"loss": 0.2699,
"step": 203
},
{
"epoch": 0.39,
"learning_rate": 1.8597249498011906e-05,
"loss": 0.2973,
"step": 204
},
{
"epoch": 0.39,
"learning_rate": 1.858128920381963e-05,
"loss": 0.3202,
"step": 205
},
{
"epoch": 0.4,
"learning_rate": 1.8565245554778516e-05,
"loss": 0.2645,
"step": 206
},
{
"epoch": 0.4,
"learning_rate": 1.854911870672947e-05,
"loss": 0.2514,
"step": 207
},
{
"epoch": 0.4,
"learning_rate": 1.8532908816321557e-05,
"loss": 0.2983,
"step": 208
},
{
"epoch": 0.4,
"learning_rate": 1.8516616041010495e-05,
"loss": 0.2516,
"step": 209
},
{
"epoch": 0.4,
"learning_rate": 1.8500240539057093e-05,
"loss": 0.2891,
"step": 210
},
{
"epoch": 0.41,
"learning_rate": 1.848378246952574e-05,
"loss": 0.2445,
"step": 211
},
{
"epoch": 0.41,
"learning_rate": 1.8467241992282842e-05,
"loss": 0.2506,
"step": 212
},
{
"epoch": 0.41,
"learning_rate": 1.8450619267995283e-05,
"loss": 0.317,
"step": 213
},
{
"epoch": 0.41,
"learning_rate": 1.843391445812886e-05,
"loss": 0.2946,
"step": 214
},
{
"epoch": 0.41,
"learning_rate": 1.84171277249467e-05,
"loss": 0.3034,
"step": 215
},
{
"epoch": 0.42,
"learning_rate": 1.8400259231507716e-05,
"loss": 0.3061,
"step": 216
},
{
"epoch": 0.42,
"learning_rate": 1.8383309141664992e-05,
"loss": 0.2333,
"step": 217
},
{
"epoch": 0.42,
"learning_rate": 1.83662776200642e-05,
"loss": 0.318,
"step": 218
},
{
"epoch": 0.42,
"learning_rate": 1.8349164832142015e-05,
"loss": 0.2176,
"step": 219
},
{
"epoch": 0.42,
"learning_rate": 1.833197094412449e-05,
"loss": 0.2698,
"step": 220
},
{
"epoch": 0.42,
"learning_rate": 1.8314696123025456e-05,
"loss": 0.2831,
"step": 221
},
{
"epoch": 0.43,
"learning_rate": 1.8297340536644877e-05,
"loss": 0.2773,
"step": 222
},
{
"epoch": 0.43,
"learning_rate": 1.827990435356725e-05,
"loss": 0.2892,
"step": 223
},
{
"epoch": 0.43,
"learning_rate": 1.826238774315995e-05,
"loss": 0.3196,
"step": 224
},
{
"epoch": 0.43,
"learning_rate": 1.8244790875571582e-05,
"loss": 0.2989,
"step": 225
},
{
"epoch": 0.43,
"learning_rate": 1.8227113921730336e-05,
"loss": 0.2624,
"step": 226
},
{
"epoch": 0.44,
"learning_rate": 1.8209357053342325e-05,
"loss": 0.2856,
"step": 227
},
{
"epoch": 0.44,
"learning_rate": 1.819152044288992e-05,
"loss": 0.3125,
"step": 228
},
{
"epoch": 0.44,
"learning_rate": 1.8173604263630066e-05,
"loss": 0.2491,
"step": 229
},
{
"epoch": 0.44,
"learning_rate": 1.8155608689592604e-05,
"loss": 0.2719,
"step": 230
},
{
"epoch": 0.44,
"learning_rate": 1.8137533895578585e-05,
"loss": 0.2805,
"step": 231
},
{
"epoch": 0.45,
"learning_rate": 1.811938005715857e-05,
"loss": 0.2564,
"step": 232
},
{
"epoch": 0.45,
"learning_rate": 1.8101147350670905e-05,
"loss": 0.2422,
"step": 233
},
{
"epoch": 0.45,
"learning_rate": 1.8082835953220055e-05,
"loss": 0.259,
"step": 234
},
{
"epoch": 0.45,
"learning_rate": 1.806444604267483e-05,
"loss": 0.2272,
"step": 235
},
{
"epoch": 0.45,
"learning_rate": 1.8045977797666685e-05,
"loss": 0.266,
"step": 236
},
{
"epoch": 0.46,
"learning_rate": 1.8027431397587993e-05,
"loss": 0.2909,
"step": 237
},
{
"epoch": 0.46,
"learning_rate": 1.8008807022590283e-05,
"loss": 0.2731,
"step": 238
},
{
"epoch": 0.46,
"learning_rate": 1.7990104853582494e-05,
"loss": 0.2517,
"step": 239
},
{
"epoch": 0.46,
"learning_rate": 1.7971325072229227e-05,
"loss": 0.2572,
"step": 240
},
{
"epoch": 0.46,
"learning_rate": 1.7952467860948975e-05,
"loss": 0.2328,
"step": 241
},
{
"epoch": 0.47,
"learning_rate": 1.7933533402912354e-05,
"loss": 0.241,
"step": 242
},
{
"epoch": 0.47,
"learning_rate": 1.791452188204031e-05,
"loss": 0.3629,
"step": 243
},
{
"epoch": 0.47,
"learning_rate": 1.7895433483002356e-05,
"loss": 0.3116,
"step": 244
},
{
"epoch": 0.47,
"learning_rate": 1.7876268391214756e-05,
"loss": 0.2936,
"step": 245
},
{
"epoch": 0.47,
"learning_rate": 1.785702679283874e-05,
"loss": 0.2804,
"step": 246
},
{
"epoch": 0.47,
"learning_rate": 1.7837708874778683e-05,
"loss": 0.2711,
"step": 247
},
{
"epoch": 0.48,
"learning_rate": 1.78183148246803e-05,
"loss": 0.3023,
"step": 248
},
{
"epoch": 0.48,
"learning_rate": 1.7798844830928818e-05,
"loss": 0.251,
"step": 249
},
{
"epoch": 0.48,
"learning_rate": 1.777929908264715e-05,
"loss": 0.2831,
"step": 250
},
{
"epoch": 0.48,
"learning_rate": 1.775967776969405e-05,
"loss": 0.2369,
"step": 251
},
{
"epoch": 0.48,
"learning_rate": 1.7739981082662275e-05,
"loss": 0.2613,
"step": 252
},
{
"epoch": 0.49,
"learning_rate": 1.772020921287674e-05,
"loss": 0.2944,
"step": 253
},
{
"epoch": 0.49,
"learning_rate": 1.7700362352392632e-05,
"loss": 0.2577,
"step": 254
},
{
"epoch": 0.49,
"learning_rate": 1.7680440693993586e-05,
"loss": 0.277,
"step": 255
},
{
"epoch": 0.49,
"learning_rate": 1.766044443118978e-05,
"loss": 0.256,
"step": 256
},
{
"epoch": 0.49,
"learning_rate": 1.7640373758216075e-05,
"loss": 0.3175,
"step": 257
},
{
"epoch": 0.5,
"learning_rate": 1.762022887003011e-05,
"loss": 0.3031,
"step": 258
},
{
"epoch": 0.5,
"learning_rate": 1.7600009962310417e-05,
"loss": 0.262,
"step": 259
},
{
"epoch": 0.5,
"learning_rate": 1.757971723145453e-05,
"loss": 0.3123,
"step": 260
},
{
"epoch": 0.5,
"learning_rate": 1.7559350874577066e-05,
"loss": 0.2985,
"step": 261
},
{
"epoch": 0.5,
"learning_rate": 1.75389110895078e-05,
"loss": 0.2605,
"step": 262
},
{
"epoch": 0.51,
"learning_rate": 1.7518398074789776e-05,
"loss": 0.3238,
"step": 263
},
{
"epoch": 0.51,
"learning_rate": 1.7497812029677344e-05,
"loss": 0.3072,
"step": 264
},
{
"epoch": 0.51,
"learning_rate": 1.7477153154134244e-05,
"loss": 0.2944,
"step": 265
},
{
"epoch": 0.51,
"learning_rate": 1.7456421648831658e-05,
"loss": 0.2557,
"step": 266
},
{
"epoch": 0.51,
"learning_rate": 1.743561771514626e-05,
"loss": 0.2991,
"step": 267
},
{
"epoch": 0.52,
"learning_rate": 1.741474155515827e-05,
"loss": 0.3116,
"step": 268
},
{
"epoch": 0.52,
"learning_rate": 1.739379337164946e-05,
"loss": 0.2911,
"step": 269
},
{
"epoch": 0.52,
"learning_rate": 1.737277336810124e-05,
"loss": 0.3144,
"step": 270
},
{
"epoch": 0.52,
"learning_rate": 1.7351681748692622e-05,
"loss": 0.3243,
"step": 271
},
{
"epoch": 0.52,
"learning_rate": 1.7330518718298263e-05,
"loss": 0.2743,
"step": 272
},
{
"epoch": 0.53,
"learning_rate": 1.7309284482486494e-05,
"loss": 0.2532,
"step": 273
},
{
"epoch": 0.53,
"learning_rate": 1.7287979247517285e-05,
"loss": 0.2777,
"step": 274
},
{
"epoch": 0.53,
"learning_rate": 1.7266603220340273e-05,
"loss": 0.3015,
"step": 275
},
{
"epoch": 0.53,
"learning_rate": 1.7245156608592727e-05,
"loss": 0.3001,
"step": 276
},
{
"epoch": 0.53,
"learning_rate": 1.7223639620597556e-05,
"loss": 0.3408,
"step": 277
},
{
"epoch": 0.53,
"learning_rate": 1.7202052465361268e-05,
"loss": 0.2568,
"step": 278
},
{
"epoch": 0.54,
"learning_rate": 1.718039535257194e-05,
"loss": 0.3001,
"step": 279
},
{
"epoch": 0.54,
"learning_rate": 1.7158668492597186e-05,
"loss": 0.231,
"step": 280
},
{
"epoch": 0.54,
"learning_rate": 1.7136872096482123e-05,
"loss": 0.2371,
"step": 281
},
{
"epoch": 0.54,
"learning_rate": 1.7115006375947304e-05,
"loss": 0.3373,
"step": 282
},
{
"epoch": 0.54,
"learning_rate": 1.7093071543386667e-05,
"loss": 0.3651,
"step": 283
},
{
"epoch": 0.55,
"learning_rate": 1.7071067811865477e-05,
"loss": 0.307,
"step": 284
},
{
"epoch": 0.55,
"learning_rate": 1.7048995395118253e-05,
"loss": 0.3449,
"step": 285
},
{
"epoch": 0.55,
"learning_rate": 1.7026854507546694e-05,
"loss": 0.3196,
"step": 286
},
{
"epoch": 0.55,
"learning_rate": 1.7004645364217584e-05,
"loss": 0.272,
"step": 287
},
{
"epoch": 0.55,
"learning_rate": 1.698236818086073e-05,
"loss": 0.2899,
"step": 288
},
{
"epoch": 0.56,
"learning_rate": 1.6960023173866834e-05,
"loss": 0.2488,
"step": 289
},
{
"epoch": 0.56,
"learning_rate": 1.693761056028542e-05,
"loss": 0.2224,
"step": 290
},
{
"epoch": 0.56,
"learning_rate": 1.6915130557822698e-05,
"loss": 0.3136,
"step": 291
},
{
"epoch": 0.56,
"learning_rate": 1.689258338483947e-05,
"loss": 0.2671,
"step": 292
},
{
"epoch": 0.56,
"learning_rate": 1.686996926034902e-05,
"loss": 0.2923,
"step": 293
},
{
"epoch": 0.57,
"learning_rate": 1.6847288404014937e-05,
"loss": 0.3313,
"step": 294
},
{
"epoch": 0.57,
"learning_rate": 1.682454103614904e-05,
"loss": 0.2638,
"step": 295
},
{
"epoch": 0.57,
"learning_rate": 1.6801727377709195e-05,
"loss": 0.3061,
"step": 296
},
{
"epoch": 0.57,
"learning_rate": 1.67788476502972e-05,
"loss": 0.2966,
"step": 297
},
{
"epoch": 0.57,
"learning_rate": 1.6755902076156606e-05,
"loss": 0.251,
"step": 298
},
{
"epoch": 0.57,
"learning_rate": 1.6732890878170573e-05,
"loss": 0.2769,
"step": 299
},
{
"epoch": 0.58,
"learning_rate": 1.67098142798597e-05,
"loss": 0.2796,
"step": 300
}
],
"logging_steps": 1.0,
"max_steps": 1040,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 100,
"total_flos": 15645950115840.0,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}