fast_agent_sw / trainer_state.json
yuchenlin's picture
add checkpoint 0405 500
31be1f7
{
"best_metric": 0.11370635777711868,
"best_model_checkpoint": "model_ckpts/flan_large_0405/checkpoint-500",
"epoch": 1.1960549910340705,
"global_step": 500,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 0.0,
"loss": 2.5078,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 0.0001,
"loss": 1.6978,
"step": 2
},
{
"epoch": 0.01,
"learning_rate": 0.0001,
"loss": 1.7317,
"step": 3
},
{
"epoch": 0.01,
"learning_rate": 0.0001,
"loss": 1.7302,
"step": 4
},
{
"epoch": 0.01,
"learning_rate": 0.0001,
"loss": 1.6091,
"step": 5
},
{
"epoch": 0.01,
"learning_rate": 0.0001,
"loss": 1.5813,
"step": 6
},
{
"epoch": 0.02,
"learning_rate": 0.0001,
"loss": 1.2533,
"step": 7
},
{
"epoch": 0.02,
"learning_rate": 0.0001,
"loss": 1.1624,
"step": 8
},
{
"epoch": 0.02,
"learning_rate": 0.0001,
"loss": 1.3477,
"step": 9
},
{
"epoch": 0.02,
"learning_rate": 0.0001,
"loss": 1.0195,
"step": 10
},
{
"epoch": 0.03,
"learning_rate": 0.0001,
"loss": 1.208,
"step": 11
},
{
"epoch": 0.03,
"learning_rate": 0.0001,
"loss": 1.0493,
"step": 12
},
{
"epoch": 0.03,
"learning_rate": 0.0001,
"loss": 0.95,
"step": 13
},
{
"epoch": 0.03,
"learning_rate": 0.0001,
"loss": 1.0308,
"step": 14
},
{
"epoch": 0.04,
"learning_rate": 0.0001,
"loss": 0.9485,
"step": 15
},
{
"epoch": 0.04,
"learning_rate": 0.0001,
"loss": 0.897,
"step": 16
},
{
"epoch": 0.04,
"learning_rate": 0.0001,
"loss": 0.7709,
"step": 17
},
{
"epoch": 0.04,
"learning_rate": 0.0001,
"loss": 0.7307,
"step": 18
},
{
"epoch": 0.05,
"learning_rate": 0.0001,
"loss": 0.6785,
"step": 19
},
{
"epoch": 0.05,
"learning_rate": 0.0001,
"loss": 0.7227,
"step": 20
},
{
"epoch": 0.05,
"learning_rate": 0.0001,
"loss": 0.8107,
"step": 21
},
{
"epoch": 0.05,
"learning_rate": 0.0001,
"loss": 0.5814,
"step": 22
},
{
"epoch": 0.05,
"learning_rate": 0.0001,
"loss": 0.5578,
"step": 23
},
{
"epoch": 0.06,
"learning_rate": 0.0001,
"loss": 0.5636,
"step": 24
},
{
"epoch": 0.06,
"learning_rate": 0.0001,
"loss": 0.5974,
"step": 25
},
{
"epoch": 0.06,
"learning_rate": 0.0001,
"loss": 0.6145,
"step": 26
},
{
"epoch": 0.06,
"learning_rate": 0.0001,
"loss": 0.5777,
"step": 27
},
{
"epoch": 0.07,
"learning_rate": 0.0001,
"loss": 0.5687,
"step": 28
},
{
"epoch": 0.07,
"learning_rate": 0.0001,
"loss": 0.4308,
"step": 29
},
{
"epoch": 0.07,
"learning_rate": 0.0001,
"loss": 0.4899,
"step": 30
},
{
"epoch": 0.07,
"learning_rate": 0.0001,
"loss": 0.5817,
"step": 31
},
{
"epoch": 0.08,
"learning_rate": 0.0001,
"loss": 0.5653,
"step": 32
},
{
"epoch": 0.08,
"learning_rate": 0.0001,
"loss": 0.4182,
"step": 33
},
{
"epoch": 0.08,
"learning_rate": 0.0001,
"loss": 0.5268,
"step": 34
},
{
"epoch": 0.08,
"learning_rate": 0.0001,
"loss": 0.474,
"step": 35
},
{
"epoch": 0.09,
"learning_rate": 0.0001,
"loss": 0.4036,
"step": 36
},
{
"epoch": 0.09,
"learning_rate": 0.0001,
"loss": 0.4179,
"step": 37
},
{
"epoch": 0.09,
"learning_rate": 0.0001,
"loss": 0.3609,
"step": 38
},
{
"epoch": 0.09,
"learning_rate": 0.0001,
"loss": 0.3455,
"step": 39
},
{
"epoch": 0.1,
"learning_rate": 0.0001,
"loss": 0.4168,
"step": 40
},
{
"epoch": 0.1,
"learning_rate": 0.0001,
"loss": 0.3247,
"step": 41
},
{
"epoch": 0.1,
"learning_rate": 0.0001,
"loss": 0.4622,
"step": 42
},
{
"epoch": 0.1,
"learning_rate": 0.0001,
"loss": 0.2171,
"step": 43
},
{
"epoch": 0.11,
"learning_rate": 0.0001,
"loss": 0.2581,
"step": 44
},
{
"epoch": 0.11,
"learning_rate": 0.0001,
"loss": 0.4231,
"step": 45
},
{
"epoch": 0.11,
"learning_rate": 0.0001,
"loss": 0.33,
"step": 46
},
{
"epoch": 0.11,
"learning_rate": 0.0001,
"loss": 0.2593,
"step": 47
},
{
"epoch": 0.11,
"learning_rate": 0.0001,
"loss": 0.4014,
"step": 48
},
{
"epoch": 0.12,
"learning_rate": 0.0001,
"loss": 0.2516,
"step": 49
},
{
"epoch": 0.12,
"learning_rate": 0.0001,
"loss": 0.3418,
"step": 50
},
{
"epoch": 0.12,
"learning_rate": 0.0001,
"loss": 0.4186,
"step": 51
},
{
"epoch": 0.12,
"learning_rate": 0.0001,
"loss": 0.2576,
"step": 52
},
{
"epoch": 0.13,
"learning_rate": 0.0001,
"loss": 0.305,
"step": 53
},
{
"epoch": 0.13,
"learning_rate": 0.0001,
"loss": 0.3837,
"step": 54
},
{
"epoch": 0.13,
"learning_rate": 0.0001,
"loss": 0.3304,
"step": 55
},
{
"epoch": 0.13,
"learning_rate": 0.0001,
"loss": 0.3303,
"step": 56
},
{
"epoch": 0.14,
"learning_rate": 0.0001,
"loss": 0.372,
"step": 57
},
{
"epoch": 0.14,
"learning_rate": 0.0001,
"loss": 0.3307,
"step": 58
},
{
"epoch": 0.14,
"learning_rate": 0.0001,
"loss": 0.2297,
"step": 59
},
{
"epoch": 0.14,
"learning_rate": 0.0001,
"loss": 0.3559,
"step": 60
},
{
"epoch": 0.15,
"learning_rate": 0.0001,
"loss": 0.3179,
"step": 61
},
{
"epoch": 0.15,
"learning_rate": 0.0001,
"loss": 0.2752,
"step": 62
},
{
"epoch": 0.15,
"learning_rate": 0.0001,
"loss": 0.2005,
"step": 63
},
{
"epoch": 0.15,
"learning_rate": 0.0001,
"loss": 0.2872,
"step": 64
},
{
"epoch": 0.16,
"learning_rate": 0.0001,
"loss": 0.276,
"step": 65
},
{
"epoch": 0.16,
"learning_rate": 0.0001,
"loss": 0.3224,
"step": 66
},
{
"epoch": 0.16,
"learning_rate": 0.0001,
"loss": 0.322,
"step": 67
},
{
"epoch": 0.16,
"learning_rate": 0.0001,
"loss": 0.2767,
"step": 68
},
{
"epoch": 0.16,
"learning_rate": 0.0001,
"loss": 0.2195,
"step": 69
},
{
"epoch": 0.17,
"learning_rate": 0.0001,
"loss": 0.304,
"step": 70
},
{
"epoch": 0.17,
"learning_rate": 0.0001,
"loss": 0.2738,
"step": 71
},
{
"epoch": 0.17,
"learning_rate": 0.0001,
"loss": 0.2279,
"step": 72
},
{
"epoch": 0.17,
"learning_rate": 0.0001,
"loss": 0.3606,
"step": 73
},
{
"epoch": 0.18,
"learning_rate": 0.0001,
"loss": 0.3302,
"step": 74
},
{
"epoch": 0.18,
"learning_rate": 0.0001,
"loss": 0.2201,
"step": 75
},
{
"epoch": 0.18,
"learning_rate": 0.0001,
"loss": 0.326,
"step": 76
},
{
"epoch": 0.18,
"learning_rate": 0.0001,
"loss": 0.1913,
"step": 77
},
{
"epoch": 0.19,
"learning_rate": 0.0001,
"loss": 0.2184,
"step": 78
},
{
"epoch": 0.19,
"learning_rate": 0.0001,
"loss": 0.3153,
"step": 79
},
{
"epoch": 0.19,
"learning_rate": 0.0001,
"loss": 0.2675,
"step": 80
},
{
"epoch": 0.19,
"learning_rate": 0.0001,
"loss": 0.3211,
"step": 81
},
{
"epoch": 0.2,
"learning_rate": 0.0001,
"loss": 0.2165,
"step": 82
},
{
"epoch": 0.2,
"learning_rate": 0.0001,
"loss": 0.2695,
"step": 83
},
{
"epoch": 0.2,
"learning_rate": 0.0001,
"loss": 0.2831,
"step": 84
},
{
"epoch": 0.2,
"learning_rate": 0.0001,
"loss": 0.2391,
"step": 85
},
{
"epoch": 0.21,
"learning_rate": 0.0001,
"loss": 0.3091,
"step": 86
},
{
"epoch": 0.21,
"learning_rate": 0.0001,
"loss": 0.318,
"step": 87
},
{
"epoch": 0.21,
"learning_rate": 0.0001,
"loss": 0.2103,
"step": 88
},
{
"epoch": 0.21,
"learning_rate": 0.0001,
"loss": 0.2863,
"step": 89
},
{
"epoch": 0.22,
"learning_rate": 0.0001,
"loss": 0.2196,
"step": 90
},
{
"epoch": 0.22,
"learning_rate": 0.0001,
"loss": 0.2062,
"step": 91
},
{
"epoch": 0.22,
"learning_rate": 0.0001,
"loss": 0.2386,
"step": 92
},
{
"epoch": 0.22,
"learning_rate": 0.0001,
"loss": 0.2202,
"step": 93
},
{
"epoch": 0.22,
"learning_rate": 0.0001,
"loss": 0.3076,
"step": 94
},
{
"epoch": 0.23,
"learning_rate": 0.0001,
"loss": 0.1863,
"step": 95
},
{
"epoch": 0.23,
"learning_rate": 0.0001,
"loss": 0.1582,
"step": 96
},
{
"epoch": 0.23,
"learning_rate": 0.0001,
"loss": 0.2404,
"step": 97
},
{
"epoch": 0.23,
"learning_rate": 0.0001,
"loss": 0.1352,
"step": 98
},
{
"epoch": 0.24,
"learning_rate": 0.0001,
"loss": 0.163,
"step": 99
},
{
"epoch": 0.24,
"learning_rate": 0.0001,
"loss": 0.2484,
"step": 100
},
{
"epoch": 0.24,
"eval_loss": 0.2323194146156311,
"eval_runtime": 745.3304,
"eval_samples_per_second": 38.313,
"eval_steps_per_second": 0.301,
"step": 100
},
{
"epoch": 0.24,
"learning_rate": 0.0001,
"loss": 0.2678,
"step": 101
},
{
"epoch": 0.24,
"learning_rate": 0.0001,
"loss": 0.236,
"step": 102
},
{
"epoch": 0.25,
"learning_rate": 0.0001,
"loss": 0.1675,
"step": 103
},
{
"epoch": 0.25,
"learning_rate": 0.0001,
"loss": 0.248,
"step": 104
},
{
"epoch": 0.25,
"learning_rate": 0.0001,
"loss": 0.2074,
"step": 105
},
{
"epoch": 0.25,
"learning_rate": 0.0001,
"loss": 0.2126,
"step": 106
},
{
"epoch": 0.26,
"learning_rate": 0.0001,
"loss": 0.1987,
"step": 107
},
{
"epoch": 0.26,
"learning_rate": 0.0001,
"loss": 0.251,
"step": 108
},
{
"epoch": 0.26,
"learning_rate": 0.0001,
"loss": 0.2269,
"step": 109
},
{
"epoch": 0.26,
"learning_rate": 0.0001,
"loss": 0.2492,
"step": 110
},
{
"epoch": 0.27,
"learning_rate": 0.0001,
"loss": 0.1903,
"step": 111
},
{
"epoch": 0.27,
"learning_rate": 0.0001,
"loss": 0.1727,
"step": 112
},
{
"epoch": 0.27,
"learning_rate": 0.0001,
"loss": 0.1838,
"step": 113
},
{
"epoch": 0.27,
"learning_rate": 0.0001,
"loss": 0.1745,
"step": 114
},
{
"epoch": 0.27,
"learning_rate": 0.0001,
"loss": 0.2179,
"step": 115
},
{
"epoch": 0.28,
"learning_rate": 0.0001,
"loss": 0.2167,
"step": 116
},
{
"epoch": 0.28,
"learning_rate": 0.0001,
"loss": 0.245,
"step": 117
},
{
"epoch": 0.28,
"learning_rate": 0.0001,
"loss": 0.2076,
"step": 118
},
{
"epoch": 0.28,
"learning_rate": 0.0001,
"loss": 0.2191,
"step": 119
},
{
"epoch": 0.29,
"learning_rate": 0.0001,
"loss": 0.1282,
"step": 120
},
{
"epoch": 0.29,
"learning_rate": 0.0001,
"loss": 0.1745,
"step": 121
},
{
"epoch": 0.29,
"learning_rate": 0.0001,
"loss": 0.2184,
"step": 122
},
{
"epoch": 0.29,
"learning_rate": 0.0001,
"loss": 0.1918,
"step": 123
},
{
"epoch": 0.3,
"learning_rate": 0.0001,
"loss": 0.1963,
"step": 124
},
{
"epoch": 0.3,
"learning_rate": 0.0001,
"loss": 0.1903,
"step": 125
},
{
"epoch": 0.3,
"learning_rate": 0.0001,
"loss": 0.1922,
"step": 126
},
{
"epoch": 0.3,
"learning_rate": 0.0001,
"loss": 0.2419,
"step": 127
},
{
"epoch": 0.31,
"learning_rate": 0.0001,
"loss": 0.1929,
"step": 128
},
{
"epoch": 0.31,
"learning_rate": 0.0001,
"loss": 0.2104,
"step": 129
},
{
"epoch": 0.31,
"learning_rate": 0.0001,
"loss": 0.225,
"step": 130
},
{
"epoch": 0.31,
"learning_rate": 0.0001,
"loss": 0.1813,
"step": 131
},
{
"epoch": 0.32,
"learning_rate": 0.0001,
"loss": 0.1714,
"step": 132
},
{
"epoch": 0.32,
"learning_rate": 0.0001,
"loss": 0.1767,
"step": 133
},
{
"epoch": 0.32,
"learning_rate": 0.0001,
"loss": 0.2196,
"step": 134
},
{
"epoch": 0.32,
"learning_rate": 0.0001,
"loss": 0.2148,
"step": 135
},
{
"epoch": 0.33,
"learning_rate": 0.0001,
"loss": 0.1339,
"step": 136
},
{
"epoch": 0.33,
"learning_rate": 0.0001,
"loss": 0.1522,
"step": 137
},
{
"epoch": 0.33,
"learning_rate": 0.0001,
"loss": 0.1982,
"step": 138
},
{
"epoch": 0.33,
"learning_rate": 0.0001,
"loss": 0.1814,
"step": 139
},
{
"epoch": 0.33,
"learning_rate": 0.0001,
"loss": 0.1864,
"step": 140
},
{
"epoch": 0.34,
"learning_rate": 0.0001,
"loss": 0.2606,
"step": 141
},
{
"epoch": 0.34,
"learning_rate": 0.0001,
"loss": 0.168,
"step": 142
},
{
"epoch": 0.34,
"learning_rate": 0.0001,
"loss": 0.2088,
"step": 143
},
{
"epoch": 0.34,
"learning_rate": 0.0001,
"loss": 0.1994,
"step": 144
},
{
"epoch": 0.35,
"learning_rate": 0.0001,
"loss": 0.1828,
"step": 145
},
{
"epoch": 0.35,
"learning_rate": 0.0001,
"loss": 0.1221,
"step": 146
},
{
"epoch": 0.35,
"learning_rate": 0.0001,
"loss": 0.1517,
"step": 147
},
{
"epoch": 0.35,
"learning_rate": 0.0001,
"loss": 0.1943,
"step": 148
},
{
"epoch": 0.36,
"learning_rate": 0.0001,
"loss": 0.188,
"step": 149
},
{
"epoch": 0.36,
"learning_rate": 0.0001,
"loss": 0.172,
"step": 150
},
{
"epoch": 0.36,
"learning_rate": 0.0001,
"loss": 0.1469,
"step": 151
},
{
"epoch": 0.36,
"learning_rate": 0.0001,
"loss": 0.1994,
"step": 152
},
{
"epoch": 0.37,
"learning_rate": 0.0001,
"loss": 0.1611,
"step": 153
},
{
"epoch": 0.37,
"learning_rate": 0.0001,
"loss": 0.1402,
"step": 154
},
{
"epoch": 0.37,
"learning_rate": 0.0001,
"loss": 0.2158,
"step": 155
},
{
"epoch": 0.37,
"learning_rate": 0.0001,
"loss": 0.1654,
"step": 156
},
{
"epoch": 0.38,
"learning_rate": 0.0001,
"loss": 0.2554,
"step": 157
},
{
"epoch": 0.38,
"learning_rate": 0.0001,
"loss": 0.1627,
"step": 158
},
{
"epoch": 0.38,
"learning_rate": 0.0001,
"loss": 0.1986,
"step": 159
},
{
"epoch": 0.38,
"learning_rate": 0.0001,
"loss": 0.2143,
"step": 160
},
{
"epoch": 0.38,
"learning_rate": 0.0001,
"loss": 0.1283,
"step": 161
},
{
"epoch": 0.39,
"learning_rate": 0.0001,
"loss": 0.1617,
"step": 162
},
{
"epoch": 0.39,
"learning_rate": 0.0001,
"loss": 0.1441,
"step": 163
},
{
"epoch": 0.39,
"learning_rate": 0.0001,
"loss": 0.1734,
"step": 164
},
{
"epoch": 0.39,
"learning_rate": 0.0001,
"loss": 0.1542,
"step": 165
},
{
"epoch": 0.4,
"learning_rate": 0.0001,
"loss": 0.2069,
"step": 166
},
{
"epoch": 0.4,
"learning_rate": 0.0001,
"loss": 0.0998,
"step": 167
},
{
"epoch": 0.4,
"learning_rate": 0.0001,
"loss": 0.146,
"step": 168
},
{
"epoch": 0.4,
"learning_rate": 0.0001,
"loss": 0.154,
"step": 169
},
{
"epoch": 0.41,
"learning_rate": 0.0001,
"loss": 0.1645,
"step": 170
},
{
"epoch": 0.41,
"learning_rate": 0.0001,
"loss": 0.1778,
"step": 171
},
{
"epoch": 0.41,
"learning_rate": 0.0001,
"loss": 0.1429,
"step": 172
},
{
"epoch": 0.41,
"learning_rate": 0.0001,
"loss": 0.1241,
"step": 173
},
{
"epoch": 0.42,
"learning_rate": 0.0001,
"loss": 0.1148,
"step": 174
},
{
"epoch": 0.42,
"learning_rate": 0.0001,
"loss": 0.1938,
"step": 175
},
{
"epoch": 0.42,
"learning_rate": 0.0001,
"loss": 0.1278,
"step": 176
},
{
"epoch": 0.42,
"learning_rate": 0.0001,
"loss": 0.1434,
"step": 177
},
{
"epoch": 0.43,
"learning_rate": 0.0001,
"loss": 0.162,
"step": 178
},
{
"epoch": 0.43,
"learning_rate": 0.0001,
"loss": 0.1742,
"step": 179
},
{
"epoch": 0.43,
"learning_rate": 0.0001,
"loss": 0.0997,
"step": 180
},
{
"epoch": 0.43,
"learning_rate": 0.0001,
"loss": 0.1884,
"step": 181
},
{
"epoch": 0.44,
"learning_rate": 0.0001,
"loss": 0.1964,
"step": 182
},
{
"epoch": 0.44,
"learning_rate": 0.0001,
"loss": 0.1243,
"step": 183
},
{
"epoch": 0.44,
"learning_rate": 0.0001,
"loss": 0.144,
"step": 184
},
{
"epoch": 0.44,
"learning_rate": 0.0001,
"loss": 0.1383,
"step": 185
},
{
"epoch": 0.44,
"learning_rate": 0.0001,
"loss": 0.1293,
"step": 186
},
{
"epoch": 0.45,
"learning_rate": 0.0001,
"loss": 0.1764,
"step": 187
},
{
"epoch": 0.45,
"learning_rate": 0.0001,
"loss": 0.2355,
"step": 188
},
{
"epoch": 0.45,
"learning_rate": 0.0001,
"loss": 0.1255,
"step": 189
},
{
"epoch": 0.45,
"learning_rate": 0.0001,
"loss": 0.109,
"step": 190
},
{
"epoch": 0.46,
"learning_rate": 0.0001,
"loss": 0.1034,
"step": 191
},
{
"epoch": 0.46,
"learning_rate": 0.0001,
"loss": 0.1215,
"step": 192
},
{
"epoch": 0.46,
"learning_rate": 0.0001,
"loss": 0.1503,
"step": 193
},
{
"epoch": 0.46,
"learning_rate": 0.0001,
"loss": 0.1129,
"step": 194
},
{
"epoch": 0.47,
"learning_rate": 0.0001,
"loss": 0.148,
"step": 195
},
{
"epoch": 0.47,
"learning_rate": 0.0001,
"loss": 0.0934,
"step": 196
},
{
"epoch": 0.47,
"learning_rate": 0.0001,
"loss": 0.1609,
"step": 197
},
{
"epoch": 0.47,
"learning_rate": 0.0001,
"loss": 0.1081,
"step": 198
},
{
"epoch": 0.48,
"learning_rate": 0.0001,
"loss": 0.1551,
"step": 199
},
{
"epoch": 0.48,
"learning_rate": 0.0001,
"loss": 0.1065,
"step": 200
},
{
"epoch": 0.48,
"eval_loss": 0.1597822606563568,
"eval_runtime": 742.3097,
"eval_samples_per_second": 38.469,
"eval_steps_per_second": 0.302,
"step": 200
},
{
"epoch": 0.48,
"learning_rate": 0.0001,
"loss": 0.1052,
"step": 201
},
{
"epoch": 0.48,
"learning_rate": 0.0001,
"loss": 0.1293,
"step": 202
},
{
"epoch": 0.49,
"learning_rate": 0.0001,
"loss": 0.1495,
"step": 203
},
{
"epoch": 0.49,
"learning_rate": 0.0001,
"loss": 0.1517,
"step": 204
},
{
"epoch": 0.49,
"learning_rate": 0.0001,
"loss": 0.1119,
"step": 205
},
{
"epoch": 0.49,
"learning_rate": 0.0001,
"loss": 0.0861,
"step": 206
},
{
"epoch": 0.49,
"learning_rate": 0.0001,
"loss": 0.1849,
"step": 207
},
{
"epoch": 0.5,
"learning_rate": 0.0001,
"loss": 0.1094,
"step": 208
},
{
"epoch": 0.5,
"learning_rate": 0.0001,
"loss": 0.1109,
"step": 209
},
{
"epoch": 0.5,
"learning_rate": 0.0001,
"loss": 0.1624,
"step": 210
},
{
"epoch": 0.5,
"learning_rate": 0.0001,
"loss": 0.116,
"step": 211
},
{
"epoch": 0.51,
"learning_rate": 0.0001,
"loss": 0.1119,
"step": 212
},
{
"epoch": 0.51,
"learning_rate": 0.0001,
"loss": 0.1807,
"step": 213
},
{
"epoch": 0.51,
"learning_rate": 0.0001,
"loss": 0.1207,
"step": 214
},
{
"epoch": 0.51,
"learning_rate": 0.0001,
"loss": 0.1183,
"step": 215
},
{
"epoch": 0.52,
"learning_rate": 0.0001,
"loss": 0.1325,
"step": 216
},
{
"epoch": 0.52,
"learning_rate": 0.0001,
"loss": 0.171,
"step": 217
},
{
"epoch": 0.52,
"learning_rate": 0.0001,
"loss": 0.1784,
"step": 218
},
{
"epoch": 0.52,
"learning_rate": 0.0001,
"loss": 0.1501,
"step": 219
},
{
"epoch": 0.53,
"learning_rate": 0.0001,
"loss": 0.1461,
"step": 220
},
{
"epoch": 0.53,
"learning_rate": 0.0001,
"loss": 0.157,
"step": 221
},
{
"epoch": 0.53,
"learning_rate": 0.0001,
"loss": 0.1206,
"step": 222
},
{
"epoch": 0.53,
"learning_rate": 0.0001,
"loss": 0.1425,
"step": 223
},
{
"epoch": 0.54,
"learning_rate": 0.0001,
"loss": 0.1121,
"step": 224
},
{
"epoch": 0.54,
"learning_rate": 0.0001,
"loss": 0.1284,
"step": 225
},
{
"epoch": 0.54,
"learning_rate": 0.0001,
"loss": 0.1298,
"step": 226
},
{
"epoch": 0.54,
"learning_rate": 0.0001,
"loss": 0.1237,
"step": 227
},
{
"epoch": 0.55,
"learning_rate": 0.0001,
"loss": 0.1248,
"step": 228
},
{
"epoch": 0.55,
"learning_rate": 0.0001,
"loss": 0.1291,
"step": 229
},
{
"epoch": 0.55,
"learning_rate": 0.0001,
"loss": 0.1222,
"step": 230
},
{
"epoch": 0.55,
"learning_rate": 0.0001,
"loss": 0.1968,
"step": 231
},
{
"epoch": 0.55,
"learning_rate": 0.0001,
"loss": 0.0949,
"step": 232
},
{
"epoch": 0.56,
"learning_rate": 0.0001,
"loss": 0.1565,
"step": 233
},
{
"epoch": 0.56,
"learning_rate": 0.0001,
"loss": 0.1129,
"step": 234
},
{
"epoch": 0.56,
"learning_rate": 0.0001,
"loss": 0.1257,
"step": 235
},
{
"epoch": 0.56,
"learning_rate": 0.0001,
"loss": 0.1656,
"step": 236
},
{
"epoch": 0.57,
"learning_rate": 0.0001,
"loss": 0.102,
"step": 237
},
{
"epoch": 0.57,
"learning_rate": 0.0001,
"loss": 0.1087,
"step": 238
},
{
"epoch": 0.57,
"learning_rate": 0.0001,
"loss": 0.099,
"step": 239
},
{
"epoch": 0.57,
"learning_rate": 0.0001,
"loss": 0.1121,
"step": 240
},
{
"epoch": 0.58,
"learning_rate": 0.0001,
"loss": 0.0874,
"step": 241
},
{
"epoch": 0.58,
"learning_rate": 0.0001,
"loss": 0.1101,
"step": 242
},
{
"epoch": 0.58,
"learning_rate": 0.0001,
"loss": 0.184,
"step": 243
},
{
"epoch": 0.58,
"learning_rate": 0.0001,
"loss": 0.0976,
"step": 244
},
{
"epoch": 0.59,
"learning_rate": 0.0001,
"loss": 0.1441,
"step": 245
},
{
"epoch": 0.59,
"learning_rate": 0.0001,
"loss": 0.1351,
"step": 246
},
{
"epoch": 0.59,
"learning_rate": 0.0001,
"loss": 0.1064,
"step": 247
},
{
"epoch": 0.59,
"learning_rate": 0.0001,
"loss": 0.1187,
"step": 248
},
{
"epoch": 0.6,
"learning_rate": 0.0001,
"loss": 0.1293,
"step": 249
},
{
"epoch": 0.6,
"learning_rate": 0.0001,
"loss": 0.1029,
"step": 250
},
{
"epoch": 0.6,
"learning_rate": 0.0001,
"loss": 0.1587,
"step": 251
},
{
"epoch": 0.6,
"learning_rate": 0.0001,
"loss": 0.0953,
"step": 252
},
{
"epoch": 0.6,
"learning_rate": 0.0001,
"loss": 0.1559,
"step": 253
},
{
"epoch": 0.61,
"learning_rate": 0.0001,
"loss": 0.1039,
"step": 254
},
{
"epoch": 0.61,
"learning_rate": 0.0001,
"loss": 0.0976,
"step": 255
},
{
"epoch": 0.61,
"learning_rate": 0.0001,
"loss": 0.1049,
"step": 256
},
{
"epoch": 0.61,
"learning_rate": 0.0001,
"loss": 0.1667,
"step": 257
},
{
"epoch": 0.62,
"learning_rate": 0.0001,
"loss": 0.0903,
"step": 258
},
{
"epoch": 0.62,
"learning_rate": 0.0001,
"loss": 0.1276,
"step": 259
},
{
"epoch": 0.62,
"learning_rate": 0.0001,
"loss": 0.2343,
"step": 260
},
{
"epoch": 0.62,
"learning_rate": 0.0001,
"loss": 0.1194,
"step": 261
},
{
"epoch": 0.63,
"learning_rate": 0.0001,
"loss": 0.0777,
"step": 262
},
{
"epoch": 0.63,
"learning_rate": 0.0001,
"loss": 0.0963,
"step": 263
},
{
"epoch": 0.63,
"learning_rate": 0.0001,
"loss": 0.1069,
"step": 264
},
{
"epoch": 0.63,
"learning_rate": 0.0001,
"loss": 0.0844,
"step": 265
},
{
"epoch": 0.64,
"learning_rate": 0.0001,
"loss": 0.1499,
"step": 266
},
{
"epoch": 0.64,
"learning_rate": 0.0001,
"loss": 0.1296,
"step": 267
},
{
"epoch": 0.64,
"learning_rate": 0.0001,
"loss": 0.1067,
"step": 268
},
{
"epoch": 0.64,
"learning_rate": 0.0001,
"loss": 0.0639,
"step": 269
},
{
"epoch": 0.65,
"learning_rate": 0.0001,
"loss": 0.1154,
"step": 270
},
{
"epoch": 0.65,
"learning_rate": 0.0001,
"loss": 0.1174,
"step": 271
},
{
"epoch": 0.65,
"learning_rate": 0.0001,
"loss": 0.1039,
"step": 272
},
{
"epoch": 0.65,
"learning_rate": 0.0001,
"loss": 0.1385,
"step": 273
},
{
"epoch": 0.66,
"learning_rate": 0.0001,
"loss": 0.0796,
"step": 274
},
{
"epoch": 0.66,
"learning_rate": 0.0001,
"loss": 0.1025,
"step": 275
},
{
"epoch": 0.66,
"learning_rate": 0.0001,
"loss": 0.0897,
"step": 276
},
{
"epoch": 0.66,
"learning_rate": 0.0001,
"loss": 0.1147,
"step": 277
},
{
"epoch": 0.66,
"learning_rate": 0.0001,
"loss": 0.1517,
"step": 278
},
{
"epoch": 0.67,
"learning_rate": 0.0001,
"loss": 0.1051,
"step": 279
},
{
"epoch": 0.67,
"learning_rate": 0.0001,
"loss": 0.126,
"step": 280
},
{
"epoch": 0.67,
"learning_rate": 0.0001,
"loss": 0.1054,
"step": 281
},
{
"epoch": 0.67,
"learning_rate": 0.0001,
"loss": 0.1048,
"step": 282
},
{
"epoch": 0.68,
"learning_rate": 0.0001,
"loss": 0.0903,
"step": 283
},
{
"epoch": 0.68,
"learning_rate": 0.0001,
"loss": 0.091,
"step": 284
},
{
"epoch": 0.68,
"learning_rate": 0.0001,
"loss": 0.0979,
"step": 285
},
{
"epoch": 0.68,
"learning_rate": 0.0001,
"loss": 0.123,
"step": 286
},
{
"epoch": 0.69,
"learning_rate": 0.0001,
"loss": 0.1121,
"step": 287
},
{
"epoch": 0.69,
"learning_rate": 0.0001,
"loss": 0.0994,
"step": 288
},
{
"epoch": 0.69,
"learning_rate": 0.0001,
"loss": 0.1557,
"step": 289
},
{
"epoch": 0.69,
"learning_rate": 0.0001,
"loss": 0.1326,
"step": 290
},
{
"epoch": 0.7,
"learning_rate": 0.0001,
"loss": 0.1365,
"step": 291
},
{
"epoch": 0.7,
"learning_rate": 0.0001,
"loss": 0.1168,
"step": 292
},
{
"epoch": 0.7,
"learning_rate": 0.0001,
"loss": 0.1322,
"step": 293
},
{
"epoch": 0.7,
"learning_rate": 0.0001,
"loss": 0.0986,
"step": 294
},
{
"epoch": 0.71,
"learning_rate": 0.0001,
"loss": 0.1147,
"step": 295
},
{
"epoch": 0.71,
"learning_rate": 0.0001,
"loss": 0.0901,
"step": 296
},
{
"epoch": 0.71,
"learning_rate": 0.0001,
"loss": 0.1243,
"step": 297
},
{
"epoch": 0.71,
"learning_rate": 0.0001,
"loss": 0.1681,
"step": 298
},
{
"epoch": 0.71,
"learning_rate": 0.0001,
"loss": 0.1462,
"step": 299
},
{
"epoch": 0.72,
"learning_rate": 0.0001,
"loss": 0.1437,
"step": 300
},
{
"epoch": 0.72,
"eval_loss": 0.14470969140529633,
"eval_runtime": 743.1207,
"eval_samples_per_second": 38.427,
"eval_steps_per_second": 0.301,
"step": 300
},
{
"epoch": 0.72,
"learning_rate": 0.0001,
"loss": 0.1221,
"step": 301
},
{
"epoch": 0.72,
"learning_rate": 0.0001,
"loss": 0.1407,
"step": 302
},
{
"epoch": 0.72,
"learning_rate": 0.0001,
"loss": 0.1183,
"step": 303
},
{
"epoch": 0.73,
"learning_rate": 0.0001,
"loss": 0.1697,
"step": 304
},
{
"epoch": 0.73,
"learning_rate": 0.0001,
"loss": 0.1082,
"step": 305
},
{
"epoch": 0.73,
"learning_rate": 0.0001,
"loss": 0.1427,
"step": 306
},
{
"epoch": 0.73,
"learning_rate": 0.0001,
"loss": 0.121,
"step": 307
},
{
"epoch": 0.74,
"learning_rate": 0.0001,
"loss": 0.0992,
"step": 308
},
{
"epoch": 0.74,
"learning_rate": 0.0001,
"loss": 0.0809,
"step": 309
},
{
"epoch": 0.74,
"learning_rate": 0.0001,
"loss": 0.1033,
"step": 310
},
{
"epoch": 0.74,
"learning_rate": 0.0001,
"loss": 0.1494,
"step": 311
},
{
"epoch": 0.75,
"learning_rate": 0.0001,
"loss": 0.1527,
"step": 312
},
{
"epoch": 0.75,
"learning_rate": 0.0001,
"loss": 0.076,
"step": 313
},
{
"epoch": 0.75,
"learning_rate": 0.0001,
"loss": 0.1454,
"step": 314
},
{
"epoch": 0.75,
"learning_rate": 0.0001,
"loss": 0.1334,
"step": 315
},
{
"epoch": 0.76,
"learning_rate": 0.0001,
"loss": 0.105,
"step": 316
},
{
"epoch": 0.76,
"learning_rate": 0.0001,
"loss": 0.1335,
"step": 317
},
{
"epoch": 0.76,
"learning_rate": 0.0001,
"loss": 0.068,
"step": 318
},
{
"epoch": 0.76,
"learning_rate": 0.0001,
"loss": 0.1387,
"step": 319
},
{
"epoch": 0.77,
"learning_rate": 0.0001,
"loss": 0.0852,
"step": 320
},
{
"epoch": 0.77,
"learning_rate": 0.0001,
"loss": 0.1014,
"step": 321
},
{
"epoch": 0.77,
"learning_rate": 0.0001,
"loss": 0.0915,
"step": 322
},
{
"epoch": 0.77,
"learning_rate": 0.0001,
"loss": 0.1422,
"step": 323
},
{
"epoch": 0.77,
"learning_rate": 0.0001,
"loss": 0.1514,
"step": 324
},
{
"epoch": 0.78,
"learning_rate": 0.0001,
"loss": 0.0835,
"step": 325
},
{
"epoch": 0.78,
"learning_rate": 0.0001,
"loss": 0.1049,
"step": 326
},
{
"epoch": 0.78,
"learning_rate": 0.0001,
"loss": 0.0932,
"step": 327
},
{
"epoch": 0.78,
"learning_rate": 0.0001,
"loss": 0.0956,
"step": 328
},
{
"epoch": 0.79,
"learning_rate": 0.0001,
"loss": 0.0723,
"step": 329
},
{
"epoch": 0.79,
"learning_rate": 0.0001,
"loss": 0.0935,
"step": 330
},
{
"epoch": 0.79,
"learning_rate": 0.0001,
"loss": 0.0611,
"step": 331
},
{
"epoch": 0.79,
"learning_rate": 0.0001,
"loss": 0.0507,
"step": 332
},
{
"epoch": 0.8,
"learning_rate": 0.0001,
"loss": 0.1117,
"step": 333
},
{
"epoch": 0.8,
"learning_rate": 0.0001,
"loss": 0.1091,
"step": 334
},
{
"epoch": 0.8,
"learning_rate": 0.0001,
"loss": 0.1065,
"step": 335
},
{
"epoch": 0.8,
"learning_rate": 0.0001,
"loss": 0.0682,
"step": 336
},
{
"epoch": 0.81,
"learning_rate": 0.0001,
"loss": 0.0918,
"step": 337
},
{
"epoch": 0.81,
"learning_rate": 0.0001,
"loss": 0.0936,
"step": 338
},
{
"epoch": 0.81,
"learning_rate": 0.0001,
"loss": 0.0987,
"step": 339
},
{
"epoch": 0.81,
"learning_rate": 0.0001,
"loss": 0.1106,
"step": 340
},
{
"epoch": 0.82,
"learning_rate": 0.0001,
"loss": 0.1227,
"step": 341
},
{
"epoch": 0.82,
"learning_rate": 0.0001,
"loss": 0.0749,
"step": 342
},
{
"epoch": 0.82,
"learning_rate": 0.0001,
"loss": 0.0789,
"step": 343
},
{
"epoch": 0.82,
"learning_rate": 0.0001,
"loss": 0.0669,
"step": 344
},
{
"epoch": 0.82,
"learning_rate": 0.0001,
"loss": 0.1007,
"step": 345
},
{
"epoch": 0.83,
"learning_rate": 0.0001,
"loss": 0.1062,
"step": 346
},
{
"epoch": 0.83,
"learning_rate": 0.0001,
"loss": 0.0954,
"step": 347
},
{
"epoch": 0.83,
"learning_rate": 0.0001,
"loss": 0.0738,
"step": 348
},
{
"epoch": 0.83,
"learning_rate": 0.0001,
"loss": 0.0871,
"step": 349
},
{
"epoch": 0.84,
"learning_rate": 0.0001,
"loss": 0.1021,
"step": 350
},
{
"epoch": 0.84,
"learning_rate": 0.0001,
"loss": 0.0832,
"step": 351
},
{
"epoch": 0.84,
"learning_rate": 0.0001,
"loss": 0.0735,
"step": 352
},
{
"epoch": 0.84,
"learning_rate": 0.0001,
"loss": 0.0896,
"step": 353
},
{
"epoch": 0.85,
"learning_rate": 0.0001,
"loss": 0.0902,
"step": 354
},
{
"epoch": 0.85,
"learning_rate": 0.0001,
"loss": 0.1407,
"step": 355
},
{
"epoch": 0.85,
"learning_rate": 0.0001,
"loss": 0.1257,
"step": 356
},
{
"epoch": 0.85,
"learning_rate": 0.0001,
"loss": 0.116,
"step": 357
},
{
"epoch": 0.86,
"learning_rate": 0.0001,
"loss": 0.0881,
"step": 358
},
{
"epoch": 0.86,
"learning_rate": 0.0001,
"loss": 0.0791,
"step": 359
},
{
"epoch": 0.86,
"learning_rate": 0.0001,
"loss": 0.0869,
"step": 360
},
{
"epoch": 0.86,
"learning_rate": 0.0001,
"loss": 0.0979,
"step": 361
},
{
"epoch": 0.87,
"learning_rate": 0.0001,
"loss": 0.0692,
"step": 362
},
{
"epoch": 0.87,
"learning_rate": 0.0001,
"loss": 0.0904,
"step": 363
},
{
"epoch": 0.87,
"learning_rate": 0.0001,
"loss": 0.1184,
"step": 364
},
{
"epoch": 0.87,
"learning_rate": 0.0001,
"loss": 0.1342,
"step": 365
},
{
"epoch": 0.88,
"learning_rate": 0.0001,
"loss": 0.0731,
"step": 366
},
{
"epoch": 0.88,
"learning_rate": 0.0001,
"loss": 0.0932,
"step": 367
},
{
"epoch": 0.88,
"learning_rate": 0.0001,
"loss": 0.1749,
"step": 368
},
{
"epoch": 0.88,
"learning_rate": 0.0001,
"loss": 0.0575,
"step": 369
},
{
"epoch": 0.88,
"learning_rate": 0.0001,
"loss": 0.1017,
"step": 370
},
{
"epoch": 0.89,
"learning_rate": 0.0001,
"loss": 0.1101,
"step": 371
},
{
"epoch": 0.89,
"learning_rate": 0.0001,
"loss": 0.0676,
"step": 372
},
{
"epoch": 0.89,
"learning_rate": 0.0001,
"loss": 0.1105,
"step": 373
},
{
"epoch": 0.89,
"learning_rate": 0.0001,
"loss": 0.1682,
"step": 374
},
{
"epoch": 0.9,
"learning_rate": 0.0001,
"loss": 0.1013,
"step": 375
},
{
"epoch": 0.9,
"learning_rate": 0.0001,
"loss": 0.0659,
"step": 376
},
{
"epoch": 0.9,
"learning_rate": 0.0001,
"loss": 0.1023,
"step": 377
},
{
"epoch": 0.9,
"learning_rate": 0.0001,
"loss": 0.1295,
"step": 378
},
{
"epoch": 0.91,
"learning_rate": 0.0001,
"loss": 0.0614,
"step": 379
},
{
"epoch": 0.91,
"learning_rate": 0.0001,
"loss": 0.0714,
"step": 380
},
{
"epoch": 0.91,
"learning_rate": 0.0001,
"loss": 0.1129,
"step": 381
},
{
"epoch": 0.91,
"learning_rate": 0.0001,
"loss": 0.0878,
"step": 382
},
{
"epoch": 0.92,
"learning_rate": 0.0001,
"loss": 0.067,
"step": 383
},
{
"epoch": 0.92,
"learning_rate": 0.0001,
"loss": 0.1109,
"step": 384
},
{
"epoch": 0.92,
"learning_rate": 0.0001,
"loss": 0.0599,
"step": 385
},
{
"epoch": 0.92,
"learning_rate": 0.0001,
"loss": 0.0956,
"step": 386
},
{
"epoch": 0.93,
"learning_rate": 0.0001,
"loss": 0.0856,
"step": 387
},
{
"epoch": 0.93,
"learning_rate": 0.0001,
"loss": 0.1055,
"step": 388
},
{
"epoch": 0.93,
"learning_rate": 0.0001,
"loss": 0.0676,
"step": 389
},
{
"epoch": 0.93,
"learning_rate": 0.0001,
"loss": 0.0693,
"step": 390
},
{
"epoch": 0.93,
"learning_rate": 0.0001,
"loss": 0.1189,
"step": 391
},
{
"epoch": 0.94,
"learning_rate": 0.0001,
"loss": 0.0918,
"step": 392
},
{
"epoch": 0.94,
"learning_rate": 0.0001,
"loss": 0.1094,
"step": 393
},
{
"epoch": 0.94,
"learning_rate": 0.0001,
"loss": 0.0821,
"step": 394
},
{
"epoch": 0.94,
"learning_rate": 0.0001,
"loss": 0.0706,
"step": 395
},
{
"epoch": 0.95,
"learning_rate": 0.0001,
"loss": 0.0654,
"step": 396
},
{
"epoch": 0.95,
"learning_rate": 0.0001,
"loss": 0.0792,
"step": 397
},
{
"epoch": 0.95,
"learning_rate": 0.0001,
"loss": 0.0672,
"step": 398
},
{
"epoch": 0.95,
"learning_rate": 0.0001,
"loss": 0.0914,
"step": 399
},
{
"epoch": 0.96,
"learning_rate": 0.0001,
"loss": 0.1309,
"step": 400
},
{
"epoch": 0.96,
"eval_loss": 0.1290283352136612,
"eval_runtime": 742.5155,
"eval_samples_per_second": 38.458,
"eval_steps_per_second": 0.302,
"step": 400
},
{
"epoch": 0.96,
"learning_rate": 0.0001,
"loss": 0.0941,
"step": 401
},
{
"epoch": 0.96,
"learning_rate": 0.0001,
"loss": 0.1102,
"step": 402
},
{
"epoch": 0.96,
"learning_rate": 0.0001,
"loss": 0.0829,
"step": 403
},
{
"epoch": 0.97,
"learning_rate": 0.0001,
"loss": 0.1196,
"step": 404
},
{
"epoch": 0.97,
"learning_rate": 0.0001,
"loss": 0.1193,
"step": 405
},
{
"epoch": 0.97,
"learning_rate": 0.0001,
"loss": 0.0769,
"step": 406
},
{
"epoch": 0.97,
"learning_rate": 0.0001,
"loss": 0.1003,
"step": 407
},
{
"epoch": 0.98,
"learning_rate": 0.0001,
"loss": 0.0811,
"step": 408
},
{
"epoch": 0.98,
"learning_rate": 0.0001,
"loss": 0.0569,
"step": 409
},
{
"epoch": 0.98,
"learning_rate": 0.0001,
"loss": 0.0563,
"step": 410
},
{
"epoch": 0.98,
"learning_rate": 0.0001,
"loss": 0.0679,
"step": 411
},
{
"epoch": 0.99,
"learning_rate": 0.0001,
"loss": 0.1029,
"step": 412
},
{
"epoch": 0.99,
"learning_rate": 0.0001,
"loss": 0.0881,
"step": 413
},
{
"epoch": 0.99,
"learning_rate": 0.0001,
"loss": 0.0984,
"step": 414
},
{
"epoch": 0.99,
"learning_rate": 0.0001,
"loss": 0.1467,
"step": 415
},
{
"epoch": 0.99,
"learning_rate": 0.0001,
"loss": 0.1257,
"step": 416
},
{
"epoch": 1.0,
"learning_rate": 0.0001,
"loss": 0.0938,
"step": 417
},
{
"epoch": 1.0,
"learning_rate": 0.0001,
"loss": 0.0712,
"step": 418
},
{
"epoch": 1.0,
"learning_rate": 0.0001,
"loss": 0.1414,
"step": 419
},
{
"epoch": 1.0,
"learning_rate": 0.0001,
"loss": 0.0995,
"step": 420
},
{
"epoch": 1.01,
"learning_rate": 0.0001,
"loss": 0.0589,
"step": 421
},
{
"epoch": 1.01,
"learning_rate": 0.0001,
"loss": 0.0746,
"step": 422
},
{
"epoch": 1.01,
"learning_rate": 0.0001,
"loss": 0.1177,
"step": 423
},
{
"epoch": 1.01,
"learning_rate": 0.0001,
"loss": 0.0941,
"step": 424
},
{
"epoch": 1.02,
"learning_rate": 0.0001,
"loss": 0.0541,
"step": 425
},
{
"epoch": 1.02,
"learning_rate": 0.0001,
"loss": 0.1259,
"step": 426
},
{
"epoch": 1.02,
"learning_rate": 0.0001,
"loss": 0.0741,
"step": 427
},
{
"epoch": 1.02,
"learning_rate": 0.0001,
"loss": 0.0885,
"step": 428
},
{
"epoch": 1.03,
"learning_rate": 0.0001,
"loss": 0.0881,
"step": 429
},
{
"epoch": 1.03,
"learning_rate": 0.0001,
"loss": 0.0925,
"step": 430
},
{
"epoch": 1.03,
"learning_rate": 0.0001,
"loss": 0.1111,
"step": 431
},
{
"epoch": 1.03,
"learning_rate": 0.0001,
"loss": 0.0652,
"step": 432
},
{
"epoch": 1.04,
"learning_rate": 0.0001,
"loss": 0.0618,
"step": 433
},
{
"epoch": 1.04,
"learning_rate": 0.0001,
"loss": 0.0876,
"step": 434
},
{
"epoch": 1.04,
"learning_rate": 0.0001,
"loss": 0.0636,
"step": 435
},
{
"epoch": 1.04,
"learning_rate": 0.0001,
"loss": 0.0783,
"step": 436
},
{
"epoch": 1.05,
"learning_rate": 0.0001,
"loss": 0.0568,
"step": 437
},
{
"epoch": 1.05,
"learning_rate": 0.0001,
"loss": 0.0551,
"step": 438
},
{
"epoch": 1.05,
"learning_rate": 0.0001,
"loss": 0.0632,
"step": 439
},
{
"epoch": 1.05,
"learning_rate": 0.0001,
"loss": 0.0737,
"step": 440
},
{
"epoch": 1.05,
"learning_rate": 0.0001,
"loss": 0.0662,
"step": 441
},
{
"epoch": 1.06,
"learning_rate": 0.0001,
"loss": 0.0901,
"step": 442
},
{
"epoch": 1.06,
"learning_rate": 0.0001,
"loss": 0.088,
"step": 443
},
{
"epoch": 1.06,
"learning_rate": 0.0001,
"loss": 0.0818,
"step": 444
},
{
"epoch": 1.06,
"learning_rate": 0.0001,
"loss": 0.064,
"step": 445
},
{
"epoch": 1.07,
"learning_rate": 0.0001,
"loss": 0.0681,
"step": 446
},
{
"epoch": 1.07,
"learning_rate": 0.0001,
"loss": 0.1519,
"step": 447
},
{
"epoch": 1.07,
"learning_rate": 0.0001,
"loss": 0.1011,
"step": 448
},
{
"epoch": 1.07,
"learning_rate": 0.0001,
"loss": 0.132,
"step": 449
},
{
"epoch": 1.08,
"learning_rate": 0.0001,
"loss": 0.0719,
"step": 450
},
{
"epoch": 1.08,
"learning_rate": 0.0001,
"loss": 0.0498,
"step": 451
},
{
"epoch": 1.08,
"learning_rate": 0.0001,
"loss": 0.0629,
"step": 452
},
{
"epoch": 1.08,
"learning_rate": 0.0001,
"loss": 0.0712,
"step": 453
},
{
"epoch": 1.09,
"learning_rate": 0.0001,
"loss": 0.0664,
"step": 454
},
{
"epoch": 1.09,
"learning_rate": 0.0001,
"loss": 0.0803,
"step": 455
},
{
"epoch": 1.09,
"learning_rate": 0.0001,
"loss": 0.0827,
"step": 456
},
{
"epoch": 1.09,
"learning_rate": 0.0001,
"loss": 0.0693,
"step": 457
},
{
"epoch": 1.1,
"learning_rate": 0.0001,
"loss": 0.0871,
"step": 458
},
{
"epoch": 1.1,
"learning_rate": 0.0001,
"loss": 0.0834,
"step": 459
},
{
"epoch": 1.1,
"learning_rate": 0.0001,
"loss": 0.0655,
"step": 460
},
{
"epoch": 1.1,
"learning_rate": 0.0001,
"loss": 0.0767,
"step": 461
},
{
"epoch": 1.11,
"learning_rate": 0.0001,
"loss": 0.0839,
"step": 462
},
{
"epoch": 1.11,
"learning_rate": 0.0001,
"loss": 0.1142,
"step": 463
},
{
"epoch": 1.11,
"learning_rate": 0.0001,
"loss": 0.1267,
"step": 464
},
{
"epoch": 1.11,
"learning_rate": 0.0001,
"loss": 0.0499,
"step": 465
},
{
"epoch": 1.11,
"learning_rate": 0.0001,
"loss": 0.0503,
"step": 466
},
{
"epoch": 1.12,
"learning_rate": 0.0001,
"loss": 0.0508,
"step": 467
},
{
"epoch": 1.12,
"learning_rate": 0.0001,
"loss": 0.0976,
"step": 468
},
{
"epoch": 1.12,
"learning_rate": 0.0001,
"loss": 0.0805,
"step": 469
},
{
"epoch": 1.12,
"learning_rate": 0.0001,
"loss": 0.0685,
"step": 470
},
{
"epoch": 1.13,
"learning_rate": 0.0001,
"loss": 0.0665,
"step": 471
},
{
"epoch": 1.13,
"learning_rate": 0.0001,
"loss": 0.069,
"step": 472
},
{
"epoch": 1.13,
"learning_rate": 0.0001,
"loss": 0.0647,
"step": 473
},
{
"epoch": 1.13,
"learning_rate": 0.0001,
"loss": 0.067,
"step": 474
},
{
"epoch": 1.14,
"learning_rate": 0.0001,
"loss": 0.0806,
"step": 475
},
{
"epoch": 1.14,
"learning_rate": 0.0001,
"loss": 0.1115,
"step": 476
},
{
"epoch": 1.14,
"learning_rate": 0.0001,
"loss": 0.0682,
"step": 477
},
{
"epoch": 1.14,
"learning_rate": 0.0001,
"loss": 0.0937,
"step": 478
},
{
"epoch": 1.15,
"learning_rate": 0.0001,
"loss": 0.0413,
"step": 479
},
{
"epoch": 1.15,
"learning_rate": 0.0001,
"loss": 0.0706,
"step": 480
},
{
"epoch": 1.15,
"learning_rate": 0.0001,
"loss": 0.0569,
"step": 481
},
{
"epoch": 1.15,
"learning_rate": 0.0001,
"loss": 0.1024,
"step": 482
},
{
"epoch": 1.16,
"learning_rate": 0.0001,
"loss": 0.0681,
"step": 483
},
{
"epoch": 1.16,
"learning_rate": 0.0001,
"loss": 0.0706,
"step": 484
},
{
"epoch": 1.16,
"learning_rate": 0.0001,
"loss": 0.0826,
"step": 485
},
{
"epoch": 1.16,
"learning_rate": 0.0001,
"loss": 0.0764,
"step": 486
},
{
"epoch": 1.16,
"learning_rate": 0.0001,
"loss": 0.092,
"step": 487
},
{
"epoch": 1.17,
"learning_rate": 0.0001,
"loss": 0.0537,
"step": 488
},
{
"epoch": 1.17,
"learning_rate": 0.0001,
"loss": 0.0624,
"step": 489
},
{
"epoch": 1.17,
"learning_rate": 0.0001,
"loss": 0.0668,
"step": 490
},
{
"epoch": 1.17,
"learning_rate": 0.0001,
"loss": 0.0895,
"step": 491
},
{
"epoch": 1.18,
"learning_rate": 0.0001,
"loss": 0.116,
"step": 492
},
{
"epoch": 1.18,
"learning_rate": 0.0001,
"loss": 0.0633,
"step": 493
},
{
"epoch": 1.18,
"learning_rate": 0.0001,
"loss": 0.0773,
"step": 494
},
{
"epoch": 1.18,
"learning_rate": 0.0001,
"loss": 0.0839,
"step": 495
},
{
"epoch": 1.19,
"learning_rate": 0.0001,
"loss": 0.0499,
"step": 496
},
{
"epoch": 1.19,
"learning_rate": 0.0001,
"loss": 0.0703,
"step": 497
},
{
"epoch": 1.19,
"learning_rate": 0.0001,
"loss": 0.0499,
"step": 498
},
{
"epoch": 1.19,
"learning_rate": 0.0001,
"loss": 0.0507,
"step": 499
},
{
"epoch": 1.2,
"learning_rate": 0.0001,
"loss": 0.0817,
"step": 500
},
{
"epoch": 1.2,
"eval_loss": 0.11370635777711868,
"eval_runtime": 744.2675,
"eval_samples_per_second": 38.368,
"eval_steps_per_second": 0.301,
"step": 500
}
],
"max_steps": 1254,
"num_train_epochs": 3,
"total_flos": 2.2140294273590886e+17,
"trial_name": null,
"trial_params": null
}