End of training
Browse files- trainer_state.json +664 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,664 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 100.0,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 10600,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.94,
|
| 13 |
+
"learning_rate": 1.3773584905660377e-06,
|
| 14 |
+
"loss": 127.5473,
|
| 15 |
+
"step": 100
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 1.89,
|
| 19 |
+
"learning_rate": 3.2641509433962264e-06,
|
| 20 |
+
"loss": 93.1935,
|
| 21 |
+
"step": 200
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 2.83,
|
| 25 |
+
"learning_rate": 5.150943396226416e-06,
|
| 26 |
+
"loss": 46.8948,
|
| 27 |
+
"step": 300
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 3.77,
|
| 31 |
+
"learning_rate": 7.037735849056604e-06,
|
| 32 |
+
"loss": 29.0448,
|
| 33 |
+
"step": 400
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 4.72,
|
| 37 |
+
"learning_rate": 8.924528301886793e-06,
|
| 38 |
+
"loss": 21.3695,
|
| 39 |
+
"step": 500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 5.66,
|
| 43 |
+
"learning_rate": 1.0811320754716984e-05,
|
| 44 |
+
"loss": 16.0309,
|
| 45 |
+
"step": 600
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 6.6,
|
| 49 |
+
"learning_rate": 1.2698113207547172e-05,
|
| 50 |
+
"loss": 11.634,
|
| 51 |
+
"step": 700
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 7.55,
|
| 55 |
+
"learning_rate": 1.458490566037736e-05,
|
| 56 |
+
"loss": 8.507,
|
| 57 |
+
"step": 800
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 8.49,
|
| 61 |
+
"learning_rate": 1.6471698113207547e-05,
|
| 62 |
+
"loss": 6.5544,
|
| 63 |
+
"step": 900
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 9.43,
|
| 67 |
+
"learning_rate": 1.8358490566037736e-05,
|
| 68 |
+
"loss": 5.4801,
|
| 69 |
+
"step": 1000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 10.38,
|
| 73 |
+
"learning_rate": 1.9999908365494266e-05,
|
| 74 |
+
"loss": 4.8409,
|
| 75 |
+
"step": 1100
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 11.32,
|
| 79 |
+
"learning_rate": 1.999307723208382e-05,
|
| 80 |
+
"loss": 4.5821,
|
| 81 |
+
"step": 1200
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 12.26,
|
| 85 |
+
"learning_rate": 1.9975410248472147e-05,
|
| 86 |
+
"loss": 4.4581,
|
| 87 |
+
"step": 1300
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 13.21,
|
| 91 |
+
"learning_rate": 1.9946926571599945e-05,
|
| 92 |
+
"loss": 4.4058,
|
| 93 |
+
"step": 1400
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 14.15,
|
| 97 |
+
"learning_rate": 1.9907657087334364e-05,
|
| 98 |
+
"loss": 4.3872,
|
| 99 |
+
"step": 1500
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 15.09,
|
| 103 |
+
"learning_rate": 1.9857644376978347e-05,
|
| 104 |
+
"loss": 4.3754,
|
| 105 |
+
"step": 1600
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 16.04,
|
| 109 |
+
"learning_rate": 1.9796942671098212e-05,
|
| 110 |
+
"loss": 4.3666,
|
| 111 |
+
"step": 1700
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 16.98,
|
| 115 |
+
"learning_rate": 1.9725617790719523e-05,
|
| 116 |
+
"loss": 4.3582,
|
| 117 |
+
"step": 1800
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 17.92,
|
| 121 |
+
"learning_rate": 1.9643747075954974e-05,
|
| 122 |
+
"loss": 4.3503,
|
| 123 |
+
"step": 1900
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 18.87,
|
| 127 |
+
"learning_rate": 1.9551419302141784e-05,
|
| 128 |
+
"loss": 4.3376,
|
| 129 |
+
"step": 2000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 19.81,
|
| 133 |
+
"learning_rate": 1.9448734583579404e-05,
|
| 134 |
+
"loss": 4.3143,
|
| 135 |
+
"step": 2100
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 20.75,
|
| 139 |
+
"learning_rate": 1.9335804264972018e-05,
|
| 140 |
+
"loss": 4.3017,
|
| 141 |
+
"step": 2200
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 21.7,
|
| 145 |
+
"learning_rate": 1.9212750800693506e-05,
|
| 146 |
+
"loss": 4.2686,
|
| 147 |
+
"step": 2300
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 22.64,
|
| 151 |
+
"learning_rate": 1.90797076220058e-05,
|
| 152 |
+
"loss": 4.2361,
|
| 153 |
+
"step": 2400
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 23.58,
|
| 157 |
+
"learning_rate": 1.8936818992374606e-05,
|
| 158 |
+
"loss": 4.1892,
|
| 159 |
+
"step": 2500
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 24.53,
|
| 163 |
+
"learning_rate": 1.8784239851039375e-05,
|
| 164 |
+
"loss": 4.0793,
|
| 165 |
+
"step": 2600
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 25.47,
|
| 169 |
+
"learning_rate": 1.8622135645007168e-05,
|
| 170 |
+
"loss": 3.8419,
|
| 171 |
+
"step": 2700
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 26.42,
|
| 175 |
+
"learning_rate": 1.845068214965255e-05,
|
| 176 |
+
"loss": 3.507,
|
| 177 |
+
"step": 2800
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 27.36,
|
| 181 |
+
"learning_rate": 1.8270065278118092e-05,
|
| 182 |
+
"loss": 3.0923,
|
| 183 |
+
"step": 2900
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 28.3,
|
| 187 |
+
"learning_rate": 1.8080480879722103e-05,
|
| 188 |
+
"loss": 2.5866,
|
| 189 |
+
"step": 3000
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 29.25,
|
| 193 |
+
"learning_rate": 1.7882134527592244e-05,
|
| 194 |
+
"loss": 2.2309,
|
| 195 |
+
"step": 3100
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 30.19,
|
| 199 |
+
"learning_rate": 1.7677351806066863e-05,
|
| 200 |
+
"loss": 1.9554,
|
| 201 |
+
"step": 3200
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 31.13,
|
| 205 |
+
"learning_rate": 1.7466600884087986e-05,
|
| 206 |
+
"loss": 1.7519,
|
| 207 |
+
"step": 3300
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 32.08,
|
| 211 |
+
"learning_rate": 1.7243535137637227e-05,
|
| 212 |
+
"loss": 1.5826,
|
| 213 |
+
"step": 3400
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 33.02,
|
| 217 |
+
"learning_rate": 1.7012614967583424e-05,
|
| 218 |
+
"loss": 1.4243,
|
| 219 |
+
"step": 3500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 33.96,
|
| 223 |
+
"learning_rate": 1.6774090768906323e-05,
|
| 224 |
+
"loss": 1.3115,
|
| 225 |
+
"step": 3600
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 34.91,
|
| 229 |
+
"learning_rate": 1.6528221181905217e-05,
|
| 230 |
+
"loss": 1.2043,
|
| 231 |
+
"step": 3700
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 35.85,
|
| 235 |
+
"learning_rate": 1.6275272811746063e-05,
|
| 236 |
+
"loss": 1.131,
|
| 237 |
+
"step": 3800
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 36.79,
|
| 241 |
+
"learning_rate": 1.6018150231520486e-05,
|
| 242 |
+
"loss": 1.0536,
|
| 243 |
+
"step": 3900
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 37.74,
|
| 247 |
+
"learning_rate": 1.575193832697346e-05,
|
| 248 |
+
"loss": 0.9722,
|
| 249 |
+
"step": 4000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 38.68,
|
| 253 |
+
"learning_rate": 1.5479489390466186e-05,
|
| 254 |
+
"loss": 0.932,
|
| 255 |
+
"step": 4100
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 39.62,
|
| 259 |
+
"learning_rate": 1.520109884810164e-05,
|
| 260 |
+
"loss": 0.8723,
|
| 261 |
+
"step": 4200
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 40.57,
|
| 265 |
+
"learning_rate": 1.4917068568678037e-05,
|
| 266 |
+
"loss": 0.8349,
|
| 267 |
+
"step": 4300
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 41.51,
|
| 271 |
+
"learning_rate": 1.4627706536361717e-05,
|
| 272 |
+
"loss": 0.8096,
|
| 273 |
+
"step": 4400
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 42.45,
|
| 277 |
+
"learning_rate": 1.4333326516728909e-05,
|
| 278 |
+
"loss": 0.8033,
|
| 279 |
+
"step": 4500
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 43.4,
|
| 283 |
+
"learning_rate": 1.4034247716538545e-05,
|
| 284 |
+
"loss": 0.757,
|
| 285 |
+
"step": 4600
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 44.34,
|
| 289 |
+
"learning_rate": 1.373079443760497e-05,
|
| 290 |
+
"loss": 0.7242,
|
| 291 |
+
"step": 4700
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 45.28,
|
| 295 |
+
"learning_rate": 1.3423295725145908e-05,
|
| 296 |
+
"loss": 0.6985,
|
| 297 |
+
"step": 4800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 46.23,
|
| 301 |
+
"learning_rate": 1.3115214387653711e-05,
|
| 302 |
+
"loss": 0.6804,
|
| 303 |
+
"step": 4900
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 47.17,
|
| 307 |
+
"learning_rate": 1.2800661191392916e-05,
|
| 308 |
+
"loss": 0.6658,
|
| 309 |
+
"step": 5000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 48.11,
|
| 313 |
+
"learning_rate": 1.2483071138273168e-05,
|
| 314 |
+
"loss": 0.6392,
|
| 315 |
+
"step": 5100
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 49.06,
|
| 319 |
+
"learning_rate": 1.2162788602520851e-05,
|
| 320 |
+
"loss": 0.6237,
|
| 321 |
+
"step": 5200
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 50.0,
|
| 325 |
+
"learning_rate": 1.1840160877917335e-05,
|
| 326 |
+
"loss": 0.5931,
|
| 327 |
+
"step": 5300
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 50.94,
|
| 331 |
+
"learning_rate": 1.1515537801215944e-05,
|
| 332 |
+
"loss": 0.5826,
|
| 333 |
+
"step": 5400
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 51.89,
|
| 337 |
+
"learning_rate": 1.1189271372801474e-05,
|
| 338 |
+
"loss": 0.5833,
|
| 339 |
+
"step": 5500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 52.83,
|
| 343 |
+
"learning_rate": 1.0861715375003623e-05,
|
| 344 |
+
"loss": 0.5603,
|
| 345 |
+
"step": 5600
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 53.77,
|
| 349 |
+
"learning_rate": 1.0533224988478176e-05,
|
| 350 |
+
"loss": 0.564,
|
| 351 |
+
"step": 5700
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 54.72,
|
| 355 |
+
"learning_rate": 1.0204156407071964e-05,
|
| 356 |
+
"loss": 0.5516,
|
| 357 |
+
"step": 5800
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 55.66,
|
| 361 |
+
"learning_rate": 9.874866451589151e-06,
|
| 362 |
+
"loss": 0.5528,
|
| 363 |
+
"step": 5900
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 56.6,
|
| 367 |
+
"learning_rate": 9.549001881675188e-06,
|
| 368 |
+
"loss": 0.5396,
|
| 369 |
+
"step": 6000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 57.55,
|
| 373 |
+
"learning_rate": 9.22033352215851e-06,
|
| 374 |
+
"loss": 0.5292,
|
| 375 |
+
"step": 6100
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 58.49,
|
| 379 |
+
"learning_rate": 8.892510582822682e-06,
|
| 380 |
+
"loss": 0.5093,
|
| 381 |
+
"step": 6200
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 59.43,
|
| 385 |
+
"learning_rate": 8.565888533777731e-06,
|
| 386 |
+
"loss": 0.5153,
|
| 387 |
+
"step": 6300
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 60.38,
|
| 391 |
+
"learning_rate": 8.240821542965487e-06,
|
| 392 |
+
"loss": 0.4972,
|
| 393 |
+
"step": 6400
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 61.32,
|
| 397 |
+
"learning_rate": 7.917662092122658e-06,
|
| 398 |
+
"loss": 0.4761,
|
| 399 |
+
"step": 6500
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 62.26,
|
| 403 |
+
"learning_rate": 7.5967605945723e-06,
|
| 404 |
+
"loss": 0.4784,
|
| 405 |
+
"step": 6600
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 63.21,
|
| 409 |
+
"learning_rate": 7.278465015258145e-06,
|
| 410 |
+
"loss": 0.4546,
|
| 411 |
+
"step": 6700
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 64.15,
|
| 415 |
+
"learning_rate": 6.963120493433808e-06,
|
| 416 |
+
"loss": 0.4692,
|
| 417 |
+
"step": 6800
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 65.09,
|
| 421 |
+
"learning_rate": 6.65106896841597e-06,
|
| 422 |
+
"loss": 0.4569,
|
| 423 |
+
"step": 6900
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 66.04,
|
| 427 |
+
"learning_rate": 6.342648808807368e-06,
|
| 428 |
+
"loss": 0.465,
|
| 429 |
+
"step": 7000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 66.98,
|
| 433 |
+
"learning_rate": 6.038194445591676e-06,
|
| 434 |
+
"loss": 0.4592,
|
| 435 |
+
"step": 7100
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 67.92,
|
| 439 |
+
"learning_rate": 5.738036009498027e-06,
|
| 440 |
+
"loss": 0.4518,
|
| 441 |
+
"step": 7200
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 68.87,
|
| 445 |
+
"learning_rate": 5.442498973028489e-06,
|
| 446 |
+
"loss": 0.4557,
|
| 447 |
+
"step": 7300
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 69.81,
|
| 451 |
+
"learning_rate": 5.151903797536631e-06,
|
| 452 |
+
"loss": 0.4392,
|
| 453 |
+
"step": 7400
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 70.75,
|
| 457 |
+
"learning_rate": 4.866565585739835e-06,
|
| 458 |
+
"loss": 0.4348,
|
| 459 |
+
"step": 7500
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 71.7,
|
| 463 |
+
"learning_rate": 4.586793740042181e-06,
|
| 464 |
+
"loss": 0.4322,
|
| 465 |
+
"step": 7600
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 72.64,
|
| 469 |
+
"learning_rate": 4.312891627038364e-06,
|
| 470 |
+
"loss": 0.4285,
|
| 471 |
+
"step": 7700
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 73.58,
|
| 475 |
+
"learning_rate": 4.047802118060723e-06,
|
| 476 |
+
"loss": 0.4296,
|
| 477 |
+
"step": 7800
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 74.53,
|
| 481 |
+
"learning_rate": 3.7864578043142407e-06,
|
| 482 |
+
"loss": 0.4309,
|
| 483 |
+
"step": 7900
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 75.47,
|
| 487 |
+
"learning_rate": 3.5318510560731366e-06,
|
| 488 |
+
"loss": 0.4341,
|
| 489 |
+
"step": 8000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 76.42,
|
| 493 |
+
"learning_rate": 3.284257952519031e-06,
|
| 494 |
+
"loss": 0.4252,
|
| 495 |
+
"step": 8100
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 77.36,
|
| 499 |
+
"learning_rate": 3.0439469676883693e-06,
|
| 500 |
+
"loss": 0.4232,
|
| 501 |
+
"step": 8200
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 78.3,
|
| 505 |
+
"learning_rate": 2.811178679356429e-06,
|
| 506 |
+
"loss": 0.42,
|
| 507 |
+
"step": 8300
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 79.25,
|
| 511 |
+
"learning_rate": 2.586205486483554e-06,
|
| 512 |
+
"loss": 0.4056,
|
| 513 |
+
"step": 8400
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 80.19,
|
| 517 |
+
"learning_rate": 2.369271335529936e-06,
|
| 518 |
+
"loss": 0.41,
|
| 519 |
+
"step": 8500
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 81.13,
|
| 523 |
+
"learning_rate": 2.160611455935786e-06,
|
| 524 |
+
"loss": 0.419,
|
| 525 |
+
"step": 8600
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 82.08,
|
| 529 |
+
"learning_rate": 1.9604521050536506e-06,
|
| 530 |
+
"loss": 0.4078,
|
| 531 |
+
"step": 8700
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 83.02,
|
| 535 |
+
"learning_rate": 1.7690103228094845e-06,
|
| 536 |
+
"loss": 0.4194,
|
| 537 |
+
"step": 8800
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 83.96,
|
| 541 |
+
"learning_rate": 1.5864936963585164e-06,
|
| 542 |
+
"loss": 0.3995,
|
| 543 |
+
"step": 8900
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 84.91,
|
| 547 |
+
"learning_rate": 1.413100134991111e-06,
|
| 548 |
+
"loss": 0.4128,
|
| 549 |
+
"step": 9000
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 85.85,
|
| 553 |
+
"learning_rate": 1.2490176555326306e-06,
|
| 554 |
+
"loss": 0.4075,
|
| 555 |
+
"step": 9100
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 86.79,
|
| 559 |
+
"learning_rate": 1.0944241784701025e-06,
|
| 560 |
+
"loss": 0.4075,
|
| 561 |
+
"step": 9200
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 87.74,
|
| 565 |
+
"learning_rate": 9.49487335026692e-07,
|
| 566 |
+
"loss": 0.4012,
|
| 567 |
+
"step": 9300
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 88.68,
|
| 571 |
+
"learning_rate": 8.143642853931688e-07,
|
| 572 |
+
"loss": 0.4066,
|
| 573 |
+
"step": 9400
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 89.62,
|
| 577 |
+
"learning_rate": 6.892015483135095e-07,
|
| 578 |
+
"loss": 0.401,
|
| 579 |
+
"step": 9500
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 90.57,
|
| 583 |
+
"learning_rate": 5.741348422094128e-07,
|
| 584 |
+
"loss": 0.3933,
|
| 585 |
+
"step": 9600
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 91.51,
|
| 589 |
+
"learning_rate": 4.69288938015946e-07,
|
| 590 |
+
"loss": 0.4163,
|
| 591 |
+
"step": 9700
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 92.45,
|
| 595 |
+
"learning_rate": 3.747775238879847e-07,
|
| 596 |
+
"loss": 0.3918,
|
| 597 |
+
"step": 9800
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 93.4,
|
| 601 |
+
"learning_rate": 2.907030819240908e-07,
|
| 602 |
+
"loss": 0.409,
|
| 603 |
+
"step": 9900
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 94.34,
|
| 607 |
+
"learning_rate": 2.1783985444653433e-07,
|
| 608 |
+
"loss": 0.3987,
|
| 609 |
+
"step": 10000
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 95.28,
|
| 613 |
+
"learning_rate": 1.5479499961417067e-07,
|
| 614 |
+
"loss": 0.395,
|
| 615 |
+
"step": 10100
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 96.23,
|
| 619 |
+
"learning_rate": 1.024256518460387e-07,
|
| 620 |
+
"loss": 0.4019,
|
| 621 |
+
"step": 10200
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 97.17,
|
| 625 |
+
"learning_rate": 6.078859709449215e-08,
|
| 626 |
+
"loss": 0.3952,
|
| 627 |
+
"step": 10300
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 98.11,
|
| 631 |
+
"learning_rate": 2.992898390362875e-08,
|
| 632 |
+
"loss": 0.4062,
|
| 633 |
+
"step": 10400
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 99.06,
|
| 637 |
+
"learning_rate": 9.88027445312123e-09,
|
| 638 |
+
"loss": 0.3926,
|
| 639 |
+
"step": 10500
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 100.0,
|
| 643 |
+
"learning_rate": 7.027047924512698e-10,
|
| 644 |
+
"loss": 0.4021,
|
| 645 |
+
"step": 10600
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 100.0,
|
| 649 |
+
"step": 10600,
|
| 650 |
+
"total_flos": 6.516030907718042e+16,
|
| 651 |
+
"train_loss": 4.689974609231049,
|
| 652 |
+
"train_runtime": 6096.2204,
|
| 653 |
+
"train_samples_per_second": 55.51,
|
| 654 |
+
"train_steps_per_second": 1.739
|
| 655 |
+
}
|
| 656 |
+
],
|
| 657 |
+
"logging_steps": 100,
|
| 658 |
+
"max_steps": 10600,
|
| 659 |
+
"num_train_epochs": 100,
|
| 660 |
+
"save_steps": 100,
|
| 661 |
+
"total_flos": 6.516030907718042e+16,
|
| 662 |
+
"trial_name": null,
|
| 663 |
+
"trial_params": null
|
| 664 |
+
}
|