code-llama160m / trainer_state.json
eqhylxx's picture
code teacher forward
0929911
raw
history blame
19.3 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.9936102236421727,
"eval_steps": 500,
"global_step": 156,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.000000000000001e-06,
"loss": 4.0414,
"step": 1
},
{
"epoch": 0.03,
"learning_rate": 8.000000000000001e-06,
"loss": 4.0692,
"step": 2
},
{
"epoch": 0.04,
"learning_rate": 1.2e-05,
"loss": 3.9926,
"step": 3
},
{
"epoch": 0.05,
"learning_rate": 1.6000000000000003e-05,
"loss": 3.8362,
"step": 4
},
{
"epoch": 0.06,
"learning_rate": 2e-05,
"loss": 3.6034,
"step": 5
},
{
"epoch": 0.08,
"learning_rate": 1.999783578606323e-05,
"loss": 3.3434,
"step": 6
},
{
"epoch": 0.09,
"learning_rate": 1.9991344081017312e-05,
"loss": 3.3582,
"step": 7
},
{
"epoch": 0.1,
"learning_rate": 1.9980527694749952e-05,
"loss": 3.3779,
"step": 8
},
{
"epoch": 0.12,
"learning_rate": 1.996539130905593e-05,
"loss": 3.1884,
"step": 9
},
{
"epoch": 0.13,
"learning_rate": 1.9945941475610623e-05,
"loss": 3.0909,
"step": 10
},
{
"epoch": 0.14,
"learning_rate": 1.9922186613134152e-05,
"loss": 2.9483,
"step": 11
},
{
"epoch": 0.15,
"learning_rate": 1.9894137003747404e-05,
"loss": 2.9103,
"step": 12
},
{
"epoch": 0.17,
"learning_rate": 1.986180478852149e-05,
"loss": 3.073,
"step": 13
},
{
"epoch": 0.18,
"learning_rate": 1.9825203962222573e-05,
"loss": 2.8503,
"step": 14
},
{
"epoch": 0.19,
"learning_rate": 1.9784350367254322e-05,
"loss": 2.6766,
"step": 15
},
{
"epoch": 0.2,
"learning_rate": 1.9739261686800662e-05,
"loss": 2.5458,
"step": 16
},
{
"epoch": 0.22,
"learning_rate": 1.968995743717171e-05,
"loss": 2.7816,
"step": 17
},
{
"epoch": 0.23,
"learning_rate": 1.963645895935632e-05,
"loss": 2.6675,
"step": 18
},
{
"epoch": 0.24,
"learning_rate": 1.9578789409784727e-05,
"loss": 2.6929,
"step": 19
},
{
"epoch": 0.26,
"learning_rate": 1.951697375030553e-05,
"loss": 2.5884,
"step": 20
},
{
"epoch": 0.27,
"learning_rate": 1.9451038737381078e-05,
"loss": 2.5874,
"step": 21
},
{
"epoch": 0.28,
"learning_rate": 1.9381012910506146e-05,
"loss": 2.3095,
"step": 22
},
{
"epoch": 0.29,
"learning_rate": 1.930692657985482e-05,
"loss": 2.4514,
"step": 23
},
{
"epoch": 0.31,
"learning_rate": 1.9228811813160972e-05,
"loss": 2.3887,
"step": 24
},
{
"epoch": 0.32,
"learning_rate": 1.9146702421837952e-05,
"loss": 2.2314,
"step": 25
},
{
"epoch": 0.33,
"learning_rate": 1.906063394634356e-05,
"loss": 2.4158,
"step": 26
},
{
"epoch": 0.35,
"learning_rate": 1.8970643640796642e-05,
"loss": 2.4192,
"step": 27
},
{
"epoch": 0.36,
"learning_rate": 1.887677045685188e-05,
"loss": 2.3371,
"step": 28
},
{
"epoch": 0.37,
"learning_rate": 1.877905502683987e-05,
"loss": 2.2763,
"step": 29
},
{
"epoch": 0.38,
"learning_rate": 1.8677539646179706e-05,
"loss": 2.2117,
"step": 30
},
{
"epoch": 0.4,
"learning_rate": 1.8572268255071718e-05,
"loss": 2.2059,
"step": 31
},
{
"epoch": 0.41,
"learning_rate": 1.8463286419478256e-05,
"loss": 2.0932,
"step": 32
},
{
"epoch": 0.42,
"learning_rate": 1.8350641311400813e-05,
"loss": 2.2732,
"step": 33
},
{
"epoch": 0.43,
"learning_rate": 1.8234381688461943e-05,
"loss": 2.2529,
"step": 34
},
{
"epoch": 0.45,
"learning_rate": 1.8114557872800906e-05,
"loss": 2.107,
"step": 35
},
{
"epoch": 0.46,
"learning_rate": 1.799122172929206e-05,
"loss": 2.2048,
"step": 36
},
{
"epoch": 0.47,
"learning_rate": 1.7864426643095537e-05,
"loss": 2.0147,
"step": 37
},
{
"epoch": 0.49,
"learning_rate": 1.773422749654988e-05,
"loss": 1.9337,
"step": 38
},
{
"epoch": 0.5,
"learning_rate": 1.7600680645416583e-05,
"loss": 2.0663,
"step": 39
},
{
"epoch": 0.51,
"learning_rate": 1.746384389448694e-05,
"loss": 1.9993,
"step": 40
},
{
"epoch": 0.52,
"learning_rate": 1.7323776472561625e-05,
"loss": 2.139,
"step": 41
},
{
"epoch": 0.54,
"learning_rate": 1.7180539006813973e-05,
"loss": 2.0415,
"step": 42
},
{
"epoch": 0.55,
"learning_rate": 1.7034193496547903e-05,
"loss": 2.076,
"step": 43
},
{
"epoch": 0.56,
"learning_rate": 1.6884803286362e-05,
"loss": 1.9077,
"step": 44
},
{
"epoch": 0.58,
"learning_rate": 1.6732433038731245e-05,
"loss": 1.9536,
"step": 45
},
{
"epoch": 0.59,
"learning_rate": 1.657714870601833e-05,
"loss": 2.0756,
"step": 46
},
{
"epoch": 0.6,
"learning_rate": 1.641901750192666e-05,
"loss": 1.8334,
"step": 47
},
{
"epoch": 0.61,
"learning_rate": 1.6258107872407376e-05,
"loss": 1.9679,
"step": 48
},
{
"epoch": 0.63,
"learning_rate": 1.609448946603304e-05,
"loss": 2.1307,
"step": 49
},
{
"epoch": 0.64,
"learning_rate": 1.592823310385073e-05,
"loss": 2.0794,
"step": 50
},
{
"epoch": 0.65,
"learning_rate": 1.5759410748727663e-05,
"loss": 1.8599,
"step": 51
},
{
"epoch": 0.66,
"learning_rate": 1.5588095474202597e-05,
"loss": 1.9418,
"step": 52
},
{
"epoch": 0.68,
"learning_rate": 1.5414361432856475e-05,
"loss": 1.9263,
"step": 53
},
{
"epoch": 0.69,
"learning_rate": 1.5238283824216015e-05,
"loss": 1.8794,
"step": 54
},
{
"epoch": 0.7,
"learning_rate": 1.5059938862204126e-05,
"loss": 1.9508,
"step": 55
},
{
"epoch": 0.72,
"learning_rate": 1.4879403742151283e-05,
"loss": 2.0843,
"step": 56
},
{
"epoch": 0.73,
"learning_rate": 1.469675660738206e-05,
"loss": 2.1207,
"step": 57
},
{
"epoch": 0.74,
"learning_rate": 1.4512076515391375e-05,
"loss": 2.0181,
"step": 58
},
{
"epoch": 0.75,
"learning_rate": 1.4325443403625012e-05,
"loss": 1.9578,
"step": 59
},
{
"epoch": 0.77,
"learning_rate": 1.4136938054879284e-05,
"loss": 1.9789,
"step": 60
},
{
"epoch": 0.78,
"learning_rate": 1.3946642062334765e-05,
"loss": 1.9475,
"step": 61
},
{
"epoch": 0.79,
"learning_rate": 1.3754637794239303e-05,
"loss": 1.9798,
"step": 62
},
{
"epoch": 0.81,
"learning_rate": 1.356100835825547e-05,
"loss": 2.0374,
"step": 63
},
{
"epoch": 0.82,
"learning_rate": 1.3365837565488065e-05,
"loss": 1.8257,
"step": 64
},
{
"epoch": 0.83,
"learning_rate": 1.316920989420703e-05,
"loss": 1.7065,
"step": 65
},
{
"epoch": 0.84,
"learning_rate": 1.2971210453281675e-05,
"loss": 2.0069,
"step": 66
},
{
"epoch": 0.86,
"learning_rate": 1.2771924945341906e-05,
"loss": 1.8164,
"step": 67
},
{
"epoch": 0.87,
"learning_rate": 1.257143962968246e-05,
"loss": 1.8483,
"step": 68
},
{
"epoch": 0.88,
"learning_rate": 1.236984128492619e-05,
"loss": 1.9107,
"step": 69
},
{
"epoch": 0.89,
"learning_rate": 1.2167217171462566e-05,
"loss": 1.8914,
"step": 70
},
{
"epoch": 0.91,
"learning_rate": 1.1963654993677645e-05,
"loss": 1.7124,
"step": 71
},
{
"epoch": 0.92,
"learning_rate": 1.1759242861991855e-05,
"loss": 1.8255,
"step": 72
},
{
"epoch": 0.93,
"learning_rate": 1.155406925472205e-05,
"loss": 1.7687,
"step": 73
},
{
"epoch": 0.95,
"learning_rate": 1.1348222979784289e-05,
"loss": 1.6878,
"step": 74
},
{
"epoch": 0.96,
"learning_rate": 1.1141793136253987e-05,
"loss": 1.6335,
"step": 75
},
{
"epoch": 0.97,
"learning_rate": 1.09348690758e-05,
"loss": 1.8839,
"step": 76
},
{
"epoch": 0.98,
"learning_rate": 1.072754036400944e-05,
"loss": 1.7922,
"step": 77
},
{
"epoch": 1.0,
"learning_rate": 1.0519896741619803e-05,
"loss": 1.8622,
"step": 78
},
{
"epoch": 1.01,
"learning_rate": 1.0312028085675393e-05,
"loss": 1.9027,
"step": 79
},
{
"epoch": 1.02,
"learning_rate": 1.0104024370624644e-05,
"loss": 1.8657,
"step": 80
},
{
"epoch": 1.04,
"learning_rate": 9.89597562937536e-06,
"loss": 1.8238,
"step": 81
},
{
"epoch": 1.05,
"learning_rate": 9.687971914324607e-06,
"loss": 1.9649,
"step": 82
},
{
"epoch": 1.06,
"learning_rate": 9.480103258380198e-06,
"loss": 1.8795,
"step": 83
},
{
"epoch": 1.07,
"learning_rate": 9.272459635990563e-06,
"loss": 1.8047,
"step": 84
},
{
"epoch": 1.09,
"learning_rate": 9.065130924199998e-06,
"loss": 1.8717,
"step": 85
},
{
"epoch": 1.1,
"learning_rate": 8.858206863746018e-06,
"loss": 1.9954,
"step": 86
},
{
"epoch": 1.11,
"learning_rate": 8.651777020215713e-06,
"loss": 1.7742,
"step": 87
},
{
"epoch": 1.12,
"learning_rate": 8.445930745277953e-06,
"loss": 1.747,
"step": 88
},
{
"epoch": 1.14,
"learning_rate": 8.240757138008149e-06,
"loss": 1.7671,
"step": 89
},
{
"epoch": 1.15,
"learning_rate": 8.036345006322358e-06,
"loss": 1.7731,
"step": 90
},
{
"epoch": 1.16,
"learning_rate": 7.832782828537437e-06,
"loss": 1.9667,
"step": 91
},
{
"epoch": 1.18,
"learning_rate": 7.630158715073813e-06,
"loss": 1.8442,
"step": 92
},
{
"epoch": 1.19,
"learning_rate": 7.428560370317542e-06,
"loss": 1.7114,
"step": 93
},
{
"epoch": 1.2,
"learning_rate": 7.228075054658096e-06,
"loss": 1.7463,
"step": 94
},
{
"epoch": 1.21,
"learning_rate": 7.028789546718327e-06,
"loss": 1.7918,
"step": 95
},
{
"epoch": 1.23,
"learning_rate": 6.8307901057929735e-06,
"loss": 1.7758,
"step": 96
},
{
"epoch": 1.24,
"learning_rate": 6.634162434511939e-06,
"loss": 1.9351,
"step": 97
},
{
"epoch": 1.25,
"learning_rate": 6.438991641744531e-06,
"loss": 1.9534,
"step": 98
},
{
"epoch": 1.27,
"learning_rate": 6.245362205760703e-06,
"loss": 1.7821,
"step": 99
},
{
"epoch": 1.28,
"learning_rate": 6.053357937665237e-06,
"loss": 1.6681,
"step": 100
},
{
"epoch": 1.29,
"learning_rate": 5.863061945120719e-06,
"loss": 1.6398,
"step": 101
},
{
"epoch": 1.3,
"learning_rate": 5.674556596374993e-06,
"loss": 1.8436,
"step": 102
},
{
"epoch": 1.32,
"learning_rate": 5.487923484608629e-06,
"loss": 1.648,
"step": 103
},
{
"epoch": 1.33,
"learning_rate": 5.3032433926179395e-06,
"loss": 1.734,
"step": 104
},
{
"epoch": 1.34,
"learning_rate": 5.120596257848716e-06,
"loss": 1.8794,
"step": 105
},
{
"epoch": 1.35,
"learning_rate": 4.940061137795876e-06,
"loss": 1.7761,
"step": 106
},
{
"epoch": 1.37,
"learning_rate": 4.7617161757839895e-06,
"loss": 1.716,
"step": 107
},
{
"epoch": 1.38,
"learning_rate": 4.5856385671435285e-06,
"loss": 1.7799,
"step": 108
},
{
"epoch": 1.39,
"learning_rate": 4.411904525797408e-06,
"loss": 1.6644,
"step": 109
},
{
"epoch": 1.41,
"learning_rate": 4.240589251272342e-06,
"loss": 1.6787,
"step": 110
},
{
"epoch": 1.42,
"learning_rate": 4.0717668961492725e-06,
"loss": 1.7178,
"step": 111
},
{
"epoch": 1.43,
"learning_rate": 3.905510533966959e-06,
"loss": 1.6878,
"step": 112
},
{
"epoch": 1.44,
"learning_rate": 3.7418921275926245e-06,
"loss": 1.5657,
"step": 113
},
{
"epoch": 1.46,
"learning_rate": 3.5809824980733445e-06,
"loss": 1.7181,
"step": 114
},
{
"epoch": 1.47,
"learning_rate": 3.422851293981676e-06,
"loss": 1.6137,
"step": 115
},
{
"epoch": 1.48,
"learning_rate": 3.2675669612687565e-06,
"loss": 1.5988,
"step": 116
},
{
"epoch": 1.5,
"learning_rate": 3.115196713638e-06,
"loss": 1.6402,
"step": 117
},
{
"epoch": 1.51,
"learning_rate": 2.965806503452098e-06,
"loss": 1.5697,
"step": 118
},
{
"epoch": 1.52,
"learning_rate": 2.819460993186032e-06,
"loss": 1.7401,
"step": 119
},
{
"epoch": 1.53,
"learning_rate": 2.6762235274383775e-06,
"loss": 1.6331,
"step": 120
},
{
"epoch": 1.55,
"learning_rate": 2.5361561055130625e-06,
"loss": 1.6747,
"step": 121
},
{
"epoch": 1.56,
"learning_rate": 2.3993193545834182e-06,
"loss": 1.6595,
"step": 122
},
{
"epoch": 1.57,
"learning_rate": 2.265772503450122e-06,
"loss": 1.5915,
"step": 123
},
{
"epoch": 1.58,
"learning_rate": 2.1355733569044633e-06,
"loss": 1.7466,
"step": 124
},
{
"epoch": 1.6,
"learning_rate": 2.008778270707944e-06,
"loss": 1.5603,
"step": 125
},
{
"epoch": 1.61,
"learning_rate": 1.8854421271990964e-06,
"loss": 1.6223,
"step": 126
},
{
"epoch": 1.62,
"learning_rate": 1.7656183115380577e-06,
"loss": 1.7658,
"step": 127
},
{
"epoch": 1.64,
"learning_rate": 1.6493586885991908e-06,
"loss": 1.7695,
"step": 128
},
{
"epoch": 1.65,
"learning_rate": 1.536713580521746e-06,
"loss": 1.5982,
"step": 129
},
{
"epoch": 1.66,
"learning_rate": 1.4277317449282834e-06,
"loss": 1.6277,
"step": 130
},
{
"epoch": 1.67,
"learning_rate": 1.3224603538202929e-06,
"loss": 1.741,
"step": 131
},
{
"epoch": 1.69,
"learning_rate": 1.220944973160133e-06,
"loss": 1.5835,
"step": 132
},
{
"epoch": 1.7,
"learning_rate": 1.1232295431481222e-06,
"loss": 1.6105,
"step": 133
},
{
"epoch": 1.71,
"learning_rate": 1.0293563592033595e-06,
"loss": 1.769,
"step": 134
},
{
"epoch": 1.73,
"learning_rate": 9.393660536564408e-07,
"loss": 1.9176,
"step": 135
},
{
"epoch": 1.74,
"learning_rate": 8.532975781620511e-07,
"loss": 1.7732,
"step": 136
},
{
"epoch": 1.75,
"learning_rate": 7.711881868390292e-07,
"loss": 1.7261,
"step": 137
},
{
"epoch": 1.76,
"learning_rate": 6.930734201451817e-07,
"loss": 1.6634,
"step": 138
},
{
"epoch": 1.78,
"learning_rate": 6.189870894938587e-07,
"loss": 1.8281,
"step": 139
},
{
"epoch": 1.79,
"learning_rate": 5.489612626189245e-07,
"loss": 1.6842,
"step": 140
},
{
"epoch": 1.8,
"learning_rate": 4.830262496944693e-07,
"loss": 1.8311,
"step": 141
},
{
"epoch": 1.81,
"learning_rate": 4.21210590215273e-07,
"loss": 1.7023,
"step": 142
},
{
"epoch": 1.83,
"learning_rate": 3.635410406436857e-07,
"loss": 1.5255,
"step": 143
},
{
"epoch": 1.84,
"learning_rate": 3.100425628282899e-07,
"loss": 1.7267,
"step": 144
},
{
"epoch": 1.85,
"learning_rate": 2.607383131993424e-07,
"loss": 1.6569,
"step": 145
},
{
"epoch": 1.87,
"learning_rate": 2.1564963274568028e-07,
"loss": 1.6669,
"step": 146
},
{
"epoch": 1.88,
"learning_rate": 1.7479603777742937e-07,
"loss": 1.6977,
"step": 147
},
{
"epoch": 1.89,
"learning_rate": 1.3819521147851122e-07,
"loss": 1.8195,
"step": 148
},
{
"epoch": 1.9,
"learning_rate": 1.0586299625259699e-07,
"loss": 1.5223,
"step": 149
},
{
"epoch": 1.92,
"learning_rate": 7.781338686584928e-08,
"loss": 1.6346,
"step": 150
},
{
"epoch": 1.93,
"learning_rate": 5.405852438937764e-08,
"loss": 1.6406,
"step": 151
},
{
"epoch": 1.94,
"learning_rate": 3.460869094407127e-08,
"loss": 1.5628,
"step": 152
},
{
"epoch": 1.96,
"learning_rate": 1.947230525005006e-08,
"loss": 1.441,
"step": 153
},
{
"epoch": 1.97,
"learning_rate": 8.655918982689582e-09,
"loss": 1.6986,
"step": 154
},
{
"epoch": 1.98,
"learning_rate": 2.164213936770576e-09,
"loss": 1.7365,
"step": 155
},
{
"epoch": 1.99,
"learning_rate": 0.0,
"loss": 1.6223,
"step": 156
},
{
"epoch": 1.99,
"step": 156,
"total_flos": 8448650564861952.0,
"train_loss": 1.2015542892309337,
"train_runtime": 1225.7143,
"train_samples_per_second": 16.317,
"train_steps_per_second": 0.127
}
],
"logging_steps": 1.0,
"max_steps": 156,
"num_train_epochs": 2,
"save_steps": 50,
"total_flos": 8448650564861952.0,
"trial_name": null,
"trial_params": null
}