Commit
•
57aa631
1
Parent(s):
34bfd20
End of training
Browse files- all_results.json +8 -0
- train_results.json +8 -0
- train_results.txt +6 -0
- trainer_state.json +437 -0
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"train_loss": 2.009976102938465,
|
| 4 |
+
"train_runtime": 3338.0012,
|
| 5 |
+
"train_samples": 42367,
|
| 6 |
+
"train_samples_per_second": 63.462,
|
| 7 |
+
"train_steps_per_second": 1.983
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"train_loss": 2.009976102938465,
|
| 4 |
+
"train_runtime": 3338.0012,
|
| 5 |
+
"train_samples": 42367,
|
| 6 |
+
"train_samples_per_second": 63.462,
|
| 7 |
+
"train_steps_per_second": 1.983
|
| 8 |
+
}
|
train_results.txt
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
epoch = 5.0
|
| 2 |
+
train_loss = 2.009976102938465
|
| 3 |
+
train_runtime = 3338.0012
|
| 4 |
+
train_samples = 42367
|
| 5 |
+
train_samples_per_second = 63.462
|
| 6 |
+
train_steps_per_second = 1.983
|
trainer_state.json
ADDED
|
@@ -0,0 +1,437 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"global_step": 6620,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.08,
|
| 12 |
+
"learning_rate": 6.042296072507553e-06,
|
| 13 |
+
"loss": 6.2033,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.15,
|
| 18 |
+
"learning_rate": 1.2084592145015106e-05,
|
| 19 |
+
"loss": 2.9531,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.23,
|
| 24 |
+
"learning_rate": 1.812688821752266e-05,
|
| 25 |
+
"loss": 2.6921,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.3,
|
| 30 |
+
"learning_rate": 2.4169184290030212e-05,
|
| 31 |
+
"loss": 2.5516,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.38,
|
| 36 |
+
"learning_rate": 3.0211480362537768e-05,
|
| 37 |
+
"loss": 2.4571,
|
| 38 |
+
"step": 500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.45,
|
| 42 |
+
"learning_rate": 3.625377643504532e-05,
|
| 43 |
+
"loss": 2.4143,
|
| 44 |
+
"step": 600
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.53,
|
| 48 |
+
"learning_rate": 3.974488083249413e-05,
|
| 49 |
+
"loss": 2.3907,
|
| 50 |
+
"step": 700
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.6,
|
| 54 |
+
"learning_rate": 3.907351460221551e-05,
|
| 55 |
+
"loss": 2.3511,
|
| 56 |
+
"step": 800
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.68,
|
| 60 |
+
"learning_rate": 3.840214837193689e-05,
|
| 61 |
+
"loss": 2.318,
|
| 62 |
+
"step": 900
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.76,
|
| 66 |
+
"learning_rate": 3.773078214165828e-05,
|
| 67 |
+
"loss": 2.3014,
|
| 68 |
+
"step": 1000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.83,
|
| 72 |
+
"learning_rate": 3.705941591137966e-05,
|
| 73 |
+
"loss": 2.2913,
|
| 74 |
+
"step": 1100
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.91,
|
| 78 |
+
"learning_rate": 3.638804968110104e-05,
|
| 79 |
+
"loss": 2.2729,
|
| 80 |
+
"step": 1200
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.98,
|
| 84 |
+
"learning_rate": 3.571668345082243e-05,
|
| 85 |
+
"loss": 2.2123,
|
| 86 |
+
"step": 1300
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.06,
|
| 90 |
+
"learning_rate": 3.504531722054381e-05,
|
| 91 |
+
"loss": 2.1213,
|
| 92 |
+
"step": 1400
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 1.13,
|
| 96 |
+
"learning_rate": 3.437395099026519e-05,
|
| 97 |
+
"loss": 2.0869,
|
| 98 |
+
"step": 1500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.21,
|
| 102 |
+
"learning_rate": 3.370258475998657e-05,
|
| 103 |
+
"loss": 2.0782,
|
| 104 |
+
"step": 1600
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 1.28,
|
| 108 |
+
"learning_rate": 3.303121852970796e-05,
|
| 109 |
+
"loss": 2.0852,
|
| 110 |
+
"step": 1700
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 1.36,
|
| 114 |
+
"learning_rate": 3.235985229942934e-05,
|
| 115 |
+
"loss": 2.0695,
|
| 116 |
+
"step": 1800
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.44,
|
| 120 |
+
"learning_rate": 3.168848606915072e-05,
|
| 121 |
+
"loss": 2.0743,
|
| 122 |
+
"step": 1900
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.51,
|
| 126 |
+
"learning_rate": 3.101711983887211e-05,
|
| 127 |
+
"loss": 2.0884,
|
| 128 |
+
"step": 2000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.59,
|
| 132 |
+
"learning_rate": 3.0345753608593493e-05,
|
| 133 |
+
"loss": 2.059,
|
| 134 |
+
"step": 2100
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 1.66,
|
| 138 |
+
"learning_rate": 2.967438737831487e-05,
|
| 139 |
+
"loss": 2.0654,
|
| 140 |
+
"step": 2200
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.74,
|
| 144 |
+
"learning_rate": 2.9003021148036254e-05,
|
| 145 |
+
"loss": 2.0302,
|
| 146 |
+
"step": 2300
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 1.81,
|
| 150 |
+
"learning_rate": 2.833165491775764e-05,
|
| 151 |
+
"loss": 2.062,
|
| 152 |
+
"step": 2400
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 1.89,
|
| 156 |
+
"learning_rate": 2.7660288687479022e-05,
|
| 157 |
+
"loss": 2.0167,
|
| 158 |
+
"step": 2500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.96,
|
| 162 |
+
"learning_rate": 2.6988922457200404e-05,
|
| 163 |
+
"loss": 2.0536,
|
| 164 |
+
"step": 2600
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 2.04,
|
| 168 |
+
"learning_rate": 2.631755622692179e-05,
|
| 169 |
+
"loss": 1.9644,
|
| 170 |
+
"step": 2700
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 2.11,
|
| 174 |
+
"learning_rate": 2.5646189996643172e-05,
|
| 175 |
+
"loss": 1.8799,
|
| 176 |
+
"step": 2800
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 2.19,
|
| 180 |
+
"learning_rate": 2.4974823766364555e-05,
|
| 181 |
+
"loss": 1.8726,
|
| 182 |
+
"step": 2900
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 2.27,
|
| 186 |
+
"learning_rate": 2.4303457536085934e-05,
|
| 187 |
+
"loss": 1.8819,
|
| 188 |
+
"step": 3000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 2.34,
|
| 192 |
+
"learning_rate": 2.363209130580732e-05,
|
| 193 |
+
"loss": 1.8726,
|
| 194 |
+
"step": 3100
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 2.42,
|
| 198 |
+
"learning_rate": 2.2960725075528702e-05,
|
| 199 |
+
"loss": 1.8584,
|
| 200 |
+
"step": 3200
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 2.49,
|
| 204 |
+
"learning_rate": 2.2289358845250084e-05,
|
| 205 |
+
"loss": 1.8824,
|
| 206 |
+
"step": 3300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 2.57,
|
| 210 |
+
"learning_rate": 2.161799261497147e-05,
|
| 211 |
+
"loss": 1.8721,
|
| 212 |
+
"step": 3400
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 2.64,
|
| 216 |
+
"learning_rate": 2.0946626384692852e-05,
|
| 217 |
+
"loss": 1.8858,
|
| 218 |
+
"step": 3500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 2.72,
|
| 222 |
+
"learning_rate": 2.0275260154414234e-05,
|
| 223 |
+
"loss": 1.8946,
|
| 224 |
+
"step": 3600
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 2.79,
|
| 228 |
+
"learning_rate": 1.9603893924135617e-05,
|
| 229 |
+
"loss": 1.8755,
|
| 230 |
+
"step": 3700
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 2.87,
|
| 234 |
+
"learning_rate": 1.8932527693857003e-05,
|
| 235 |
+
"loss": 1.8765,
|
| 236 |
+
"step": 3800
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 2.95,
|
| 240 |
+
"learning_rate": 1.826116146357838e-05,
|
| 241 |
+
"loss": 1.8682,
|
| 242 |
+
"step": 3900
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 3.02,
|
| 246 |
+
"learning_rate": 1.7589795233299767e-05,
|
| 247 |
+
"loss": 1.8225,
|
| 248 |
+
"step": 4000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 3.1,
|
| 252 |
+
"learning_rate": 1.691842900302115e-05,
|
| 253 |
+
"loss": 1.7437,
|
| 254 |
+
"step": 4100
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 3.17,
|
| 258 |
+
"learning_rate": 1.6247062772742532e-05,
|
| 259 |
+
"loss": 1.7122,
|
| 260 |
+
"step": 4200
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 3.25,
|
| 264 |
+
"learning_rate": 1.5575696542463918e-05,
|
| 265 |
+
"loss": 1.739,
|
| 266 |
+
"step": 4300
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 3.32,
|
| 270 |
+
"learning_rate": 1.4904330312185298e-05,
|
| 271 |
+
"loss": 1.7283,
|
| 272 |
+
"step": 4400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 3.4,
|
| 276 |
+
"learning_rate": 1.4232964081906682e-05,
|
| 277 |
+
"loss": 1.7479,
|
| 278 |
+
"step": 4500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 3.47,
|
| 282 |
+
"learning_rate": 1.3561597851628063e-05,
|
| 283 |
+
"loss": 1.7429,
|
| 284 |
+
"step": 4600
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 3.55,
|
| 288 |
+
"learning_rate": 1.2890231621349447e-05,
|
| 289 |
+
"loss": 1.7446,
|
| 290 |
+
"step": 4700
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 3.63,
|
| 294 |
+
"learning_rate": 1.2218865391070831e-05,
|
| 295 |
+
"loss": 1.7533,
|
| 296 |
+
"step": 4800
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 3.7,
|
| 300 |
+
"learning_rate": 1.1547499160792213e-05,
|
| 301 |
+
"loss": 1.7455,
|
| 302 |
+
"step": 4900
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 3.78,
|
| 306 |
+
"learning_rate": 1.0876132930513597e-05,
|
| 307 |
+
"loss": 1.7368,
|
| 308 |
+
"step": 5000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 3.78,
|
| 312 |
+
"eval_bleu1": 29.9409,
|
| 313 |
+
"eval_bleu2": 17.5869,
|
| 314 |
+
"eval_bleu3": 10.4195,
|
| 315 |
+
"eval_bleu4": 6.1345,
|
| 316 |
+
"eval_gen_len": 50.5967,
|
| 317 |
+
"eval_loss": 2.613112688064575,
|
| 318 |
+
"eval_rouge1": 35.7499,
|
| 319 |
+
"eval_rouge2": 13.0188,
|
| 320 |
+
"eval_rougeL": 23.5089,
|
| 321 |
+
"eval_runtime": 67.2771,
|
| 322 |
+
"eval_samples_per_second": 6.377,
|
| 323 |
+
"eval_steps_per_second": 0.059,
|
| 324 |
+
"step": 5000
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 3.85,
|
| 328 |
+
"learning_rate": 1.0204766700234978e-05,
|
| 329 |
+
"loss": 1.7469,
|
| 330 |
+
"step": 5100
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 3.93,
|
| 334 |
+
"learning_rate": 9.533400469956362e-06,
|
| 335 |
+
"loss": 1.7526,
|
| 336 |
+
"step": 5200
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 4.0,
|
| 340 |
+
"learning_rate": 8.862034239677744e-06,
|
| 341 |
+
"loss": 1.7479,
|
| 342 |
+
"step": 5300
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 4.08,
|
| 346 |
+
"learning_rate": 8.190668009399128e-06,
|
| 347 |
+
"loss": 1.6487,
|
| 348 |
+
"step": 5400
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 4.15,
|
| 352 |
+
"learning_rate": 7.5193017791205116e-06,
|
| 353 |
+
"loss": 1.6374,
|
| 354 |
+
"step": 5500
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 4.23,
|
| 358 |
+
"learning_rate": 6.847935548841894e-06,
|
| 359 |
+
"loss": 1.652,
|
| 360 |
+
"step": 5600
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 4.31,
|
| 364 |
+
"learning_rate": 6.176569318563277e-06,
|
| 365 |
+
"loss": 1.6542,
|
| 366 |
+
"step": 5700
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 4.38,
|
| 370 |
+
"learning_rate": 5.505203088284659e-06,
|
| 371 |
+
"loss": 1.6443,
|
| 372 |
+
"step": 5800
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 4.46,
|
| 376 |
+
"learning_rate": 4.833836858006043e-06,
|
| 377 |
+
"loss": 1.6469,
|
| 378 |
+
"step": 5900
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 4.53,
|
| 382 |
+
"learning_rate": 4.162470627727426e-06,
|
| 383 |
+
"loss": 1.6569,
|
| 384 |
+
"step": 6000
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 4.61,
|
| 388 |
+
"learning_rate": 3.4911043974488085e-06,
|
| 389 |
+
"loss": 1.6499,
|
| 390 |
+
"step": 6100
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 4.68,
|
| 394 |
+
"learning_rate": 2.8197381671701917e-06,
|
| 395 |
+
"loss": 1.6694,
|
| 396 |
+
"step": 6200
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 4.76,
|
| 400 |
+
"learning_rate": 2.1483719368915745e-06,
|
| 401 |
+
"loss": 1.6437,
|
| 402 |
+
"step": 6300
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 4.83,
|
| 406 |
+
"learning_rate": 1.4770057066129575e-06,
|
| 407 |
+
"loss": 1.6704,
|
| 408 |
+
"step": 6400
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 4.91,
|
| 412 |
+
"learning_rate": 8.056394763343404e-07,
|
| 413 |
+
"loss": 1.6467,
|
| 414 |
+
"step": 6500
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 4.98,
|
| 418 |
+
"learning_rate": 1.342732460557234e-07,
|
| 419 |
+
"loss": 1.6622,
|
| 420 |
+
"step": 6600
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 5.0,
|
| 424 |
+
"step": 6620,
|
| 425 |
+
"total_flos": 4.347230642786304e+16,
|
| 426 |
+
"train_loss": 2.009976102938465,
|
| 427 |
+
"train_runtime": 3338.0012,
|
| 428 |
+
"train_samples_per_second": 63.462,
|
| 429 |
+
"train_steps_per_second": 1.983
|
| 430 |
+
}
|
| 431 |
+
],
|
| 432 |
+
"max_steps": 6620,
|
| 433 |
+
"num_train_epochs": 5,
|
| 434 |
+
"total_flos": 4.347230642786304e+16,
|
| 435 |
+
"trial_name": null,
|
| 436 |
+
"trial_params": null
|
| 437 |
+
}
|