Commit
·
7690249
1
Parent(s):
dd1f686
Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +547 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,547 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 44368,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.05,
|
| 13 |
+
"learning_rate": 1.9776866209880996e-05,
|
| 14 |
+
"loss": 1.6544,
|
| 15 |
+
"step": 500
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.09,
|
| 19 |
+
"learning_rate": 1.9551478543094125e-05,
|
| 20 |
+
"loss": 1.4272,
|
| 21 |
+
"step": 1000
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.14,
|
| 25 |
+
"learning_rate": 1.932609087630725e-05,
|
| 26 |
+
"loss": 1.3783,
|
| 27 |
+
"step": 1500
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.18,
|
| 31 |
+
"learning_rate": 1.9100703209520376e-05,
|
| 32 |
+
"loss": 1.324,
|
| 33 |
+
"step": 2000
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.23,
|
| 37 |
+
"learning_rate": 1.8875315542733504e-05,
|
| 38 |
+
"loss": 1.2637,
|
| 39 |
+
"step": 2500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.27,
|
| 43 |
+
"learning_rate": 1.864992787594663e-05,
|
| 44 |
+
"loss": 1.2255,
|
| 45 |
+
"step": 3000
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.32,
|
| 49 |
+
"learning_rate": 1.8424540209159756e-05,
|
| 50 |
+
"loss": 1.211,
|
| 51 |
+
"step": 3500
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.36,
|
| 55 |
+
"learning_rate": 1.8199603317706455e-05,
|
| 56 |
+
"loss": 1.1977,
|
| 57 |
+
"step": 4000
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.41,
|
| 61 |
+
"learning_rate": 1.7974215650919584e-05,
|
| 62 |
+
"loss": 1.1879,
|
| 63 |
+
"step": 4500
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.45,
|
| 67 |
+
"learning_rate": 1.774882798413271e-05,
|
| 68 |
+
"loss": 1.1487,
|
| 69 |
+
"step": 5000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.5,
|
| 73 |
+
"learning_rate": 1.7523440317345835e-05,
|
| 74 |
+
"loss": 1.1427,
|
| 75 |
+
"step": 5500
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.54,
|
| 79 |
+
"learning_rate": 1.7298052650558964e-05,
|
| 80 |
+
"loss": 1.1066,
|
| 81 |
+
"step": 6000
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.59,
|
| 85 |
+
"learning_rate": 1.707266498377209e-05,
|
| 86 |
+
"loss": 1.1576,
|
| 87 |
+
"step": 6500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.63,
|
| 91 |
+
"learning_rate": 1.6847277316985215e-05,
|
| 92 |
+
"loss": 1.0895,
|
| 93 |
+
"step": 7000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.68,
|
| 97 |
+
"learning_rate": 1.662188965019834e-05,
|
| 98 |
+
"loss": 1.0893,
|
| 99 |
+
"step": 7500
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.72,
|
| 103 |
+
"learning_rate": 1.639650198341147e-05,
|
| 104 |
+
"loss": 1.0693,
|
| 105 |
+
"step": 8000
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.77,
|
| 109 |
+
"learning_rate": 1.6171114316624595e-05,
|
| 110 |
+
"loss": 1.0979,
|
| 111 |
+
"step": 8500
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.81,
|
| 115 |
+
"learning_rate": 1.5945726649837724e-05,
|
| 116 |
+
"loss": 1.0202,
|
| 117 |
+
"step": 9000
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.86,
|
| 121 |
+
"learning_rate": 1.572033898305085e-05,
|
| 122 |
+
"loss": 1.0574,
|
| 123 |
+
"step": 9500
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.9,
|
| 127 |
+
"learning_rate": 1.549540209159755e-05,
|
| 128 |
+
"loss": 0.9914,
|
| 129 |
+
"step": 10000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.95,
|
| 133 |
+
"learning_rate": 1.5270014424810675e-05,
|
| 134 |
+
"loss": 1.0377,
|
| 135 |
+
"step": 10500
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.99,
|
| 139 |
+
"learning_rate": 1.5045077533357375e-05,
|
| 140 |
+
"loss": 0.9989,
|
| 141 |
+
"step": 11000
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 1.04,
|
| 145 |
+
"learning_rate": 1.4820140641904077e-05,
|
| 146 |
+
"loss": 0.8036,
|
| 147 |
+
"step": 11500
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.08,
|
| 151 |
+
"learning_rate": 1.4594752975117204e-05,
|
| 152 |
+
"loss": 0.7445,
|
| 153 |
+
"step": 12000
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 1.13,
|
| 157 |
+
"learning_rate": 1.436936530833033e-05,
|
| 158 |
+
"loss": 0.7566,
|
| 159 |
+
"step": 12500
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 1.17,
|
| 163 |
+
"learning_rate": 1.4143977641543456e-05,
|
| 164 |
+
"loss": 0.7529,
|
| 165 |
+
"step": 13000
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 1.22,
|
| 169 |
+
"learning_rate": 1.3918589974756582e-05,
|
| 170 |
+
"loss": 0.7446,
|
| 171 |
+
"step": 13500
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 1.26,
|
| 175 |
+
"learning_rate": 1.369320230796971e-05,
|
| 176 |
+
"loss": 0.7539,
|
| 177 |
+
"step": 14000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.31,
|
| 181 |
+
"learning_rate": 1.3468265416516409e-05,
|
| 182 |
+
"loss": 0.7389,
|
| 183 |
+
"step": 14500
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 1.35,
|
| 187 |
+
"learning_rate": 1.3242877749729536e-05,
|
| 188 |
+
"loss": 0.7502,
|
| 189 |
+
"step": 15000
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.4,
|
| 193 |
+
"learning_rate": 1.3017940858276236e-05,
|
| 194 |
+
"loss": 0.7319,
|
| 195 |
+
"step": 15500
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 1.44,
|
| 199 |
+
"learning_rate": 1.2792553191489363e-05,
|
| 200 |
+
"loss": 0.7732,
|
| 201 |
+
"step": 16000
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 1.49,
|
| 205 |
+
"learning_rate": 1.2567165524702489e-05,
|
| 206 |
+
"loss": 0.7458,
|
| 207 |
+
"step": 16500
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 1.53,
|
| 211 |
+
"learning_rate": 1.2341777857915616e-05,
|
| 212 |
+
"loss": 0.7277,
|
| 213 |
+
"step": 17000
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.58,
|
| 217 |
+
"learning_rate": 1.2116390191128742e-05,
|
| 218 |
+
"loss": 0.7559,
|
| 219 |
+
"step": 17500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.62,
|
| 223 |
+
"learning_rate": 1.1891002524341869e-05,
|
| 224 |
+
"loss": 0.7572,
|
| 225 |
+
"step": 18000
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 1.67,
|
| 229 |
+
"learning_rate": 1.1665614857554996e-05,
|
| 230 |
+
"loss": 0.7387,
|
| 231 |
+
"step": 18500
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.71,
|
| 235 |
+
"learning_rate": 1.1440227190768123e-05,
|
| 236 |
+
"loss": 0.7502,
|
| 237 |
+
"step": 19000
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 1.76,
|
| 241 |
+
"learning_rate": 1.1214839523981249e-05,
|
| 242 |
+
"loss": 0.7007,
|
| 243 |
+
"step": 19500
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 1.8,
|
| 247 |
+
"learning_rate": 1.0989451857194376e-05,
|
| 248 |
+
"loss": 0.7435,
|
| 249 |
+
"step": 20000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.85,
|
| 253 |
+
"learning_rate": 1.0764514965741076e-05,
|
| 254 |
+
"loss": 0.7114,
|
| 255 |
+
"step": 20500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.89,
|
| 259 |
+
"learning_rate": 1.0539127298954201e-05,
|
| 260 |
+
"loss": 0.7574,
|
| 261 |
+
"step": 21000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.94,
|
| 265 |
+
"learning_rate": 1.0313739632167327e-05,
|
| 266 |
+
"loss": 0.7184,
|
| 267 |
+
"step": 21500
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 1.98,
|
| 271 |
+
"learning_rate": 1.0088351965380456e-05,
|
| 272 |
+
"loss": 0.6812,
|
| 273 |
+
"step": 22000
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 2.03,
|
| 277 |
+
"learning_rate": 9.862964298593581e-06,
|
| 278 |
+
"loss": 0.5916,
|
| 279 |
+
"step": 22500
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 2.07,
|
| 283 |
+
"learning_rate": 9.637576631806709e-06,
|
| 284 |
+
"loss": 0.5152,
|
| 285 |
+
"step": 23000
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 2.12,
|
| 289 |
+
"learning_rate": 9.412188965019836e-06,
|
| 290 |
+
"loss": 0.4892,
|
| 291 |
+
"step": 23500
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.16,
|
| 295 |
+
"learning_rate": 9.187252073566536e-06,
|
| 296 |
+
"loss": 0.5003,
|
| 297 |
+
"step": 24000
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 2.21,
|
| 301 |
+
"learning_rate": 8.961864406779663e-06,
|
| 302 |
+
"loss": 0.4962,
|
| 303 |
+
"step": 24500
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 2.25,
|
| 307 |
+
"learning_rate": 8.736476739992788e-06,
|
| 308 |
+
"loss": 0.4955,
|
| 309 |
+
"step": 25000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 2.3,
|
| 313 |
+
"learning_rate": 8.511089073205916e-06,
|
| 314 |
+
"loss": 0.5183,
|
| 315 |
+
"step": 25500
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 2.34,
|
| 319 |
+
"learning_rate": 8.285701406419041e-06,
|
| 320 |
+
"loss": 0.501,
|
| 321 |
+
"step": 26000
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 2.39,
|
| 325 |
+
"learning_rate": 8.060313739632168e-06,
|
| 326 |
+
"loss": 0.4932,
|
| 327 |
+
"step": 26500
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 2.43,
|
| 331 |
+
"learning_rate": 7.835376848178868e-06,
|
| 332 |
+
"loss": 0.5092,
|
| 333 |
+
"step": 27000
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 2.48,
|
| 337 |
+
"learning_rate": 7.609989181391995e-06,
|
| 338 |
+
"loss": 0.5327,
|
| 339 |
+
"step": 27500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 2.52,
|
| 343 |
+
"learning_rate": 7.384601514605121e-06,
|
| 344 |
+
"loss": 0.4913,
|
| 345 |
+
"step": 28000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 2.57,
|
| 349 |
+
"learning_rate": 7.159213847818248e-06,
|
| 350 |
+
"loss": 0.4862,
|
| 351 |
+
"step": 28500
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 2.61,
|
| 355 |
+
"learning_rate": 6.9338261810313745e-06,
|
| 356 |
+
"loss": 0.5005,
|
| 357 |
+
"step": 29000
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 2.66,
|
| 361 |
+
"learning_rate": 6.708889289578074e-06,
|
| 362 |
+
"loss": 0.498,
|
| 363 |
+
"step": 29500
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 2.7,
|
| 367 |
+
"learning_rate": 6.483952398124775e-06,
|
| 368 |
+
"loss": 0.5037,
|
| 369 |
+
"step": 30000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 2.75,
|
| 373 |
+
"learning_rate": 6.2585647313379015e-06,
|
| 374 |
+
"loss": 0.4922,
|
| 375 |
+
"step": 30500
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 2.79,
|
| 379 |
+
"learning_rate": 6.033177064551028e-06,
|
| 380 |
+
"loss": 0.4944,
|
| 381 |
+
"step": 31000
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 2.84,
|
| 385 |
+
"learning_rate": 5.807789397764155e-06,
|
| 386 |
+
"loss": 0.4795,
|
| 387 |
+
"step": 31500
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 2.88,
|
| 391 |
+
"learning_rate": 5.5824017309772815e-06,
|
| 392 |
+
"loss": 0.498,
|
| 393 |
+
"step": 32000
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 2.93,
|
| 397 |
+
"learning_rate": 5.357014064190408e-06,
|
| 398 |
+
"loss": 0.4985,
|
| 399 |
+
"step": 32500
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 2.98,
|
| 403 |
+
"learning_rate": 5.131626397403534e-06,
|
| 404 |
+
"loss": 0.4915,
|
| 405 |
+
"step": 33000
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 3.02,
|
| 409 |
+
"learning_rate": 4.9062387306166605e-06,
|
| 410 |
+
"loss": 0.4185,
|
| 411 |
+
"step": 33500
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 3.07,
|
| 415 |
+
"learning_rate": 4.680851063829788e-06,
|
| 416 |
+
"loss": 0.3421,
|
| 417 |
+
"step": 34000
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 3.11,
|
| 421 |
+
"learning_rate": 4.455463397042914e-06,
|
| 422 |
+
"loss": 0.353,
|
| 423 |
+
"step": 34500
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 3.16,
|
| 427 |
+
"learning_rate": 4.2300757302560405e-06,
|
| 428 |
+
"loss": 0.33,
|
| 429 |
+
"step": 35000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 3.2,
|
| 433 |
+
"learning_rate": 4.004688063469167e-06,
|
| 434 |
+
"loss": 0.3277,
|
| 435 |
+
"step": 35500
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 3.25,
|
| 439 |
+
"learning_rate": 3.779300396682294e-06,
|
| 440 |
+
"loss": 0.3456,
|
| 441 |
+
"step": 36000
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 3.29,
|
| 445 |
+
"learning_rate": 3.553912729895421e-06,
|
| 446 |
+
"loss": 0.3295,
|
| 447 |
+
"step": 36500
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 3.34,
|
| 451 |
+
"learning_rate": 3.3289758384421207e-06,
|
| 452 |
+
"loss": 0.3293,
|
| 453 |
+
"step": 37000
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 3.38,
|
| 457 |
+
"learning_rate": 3.1035881716552475e-06,
|
| 458 |
+
"loss": 0.3478,
|
| 459 |
+
"step": 37500
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 3.43,
|
| 463 |
+
"learning_rate": 2.8782005048683735e-06,
|
| 464 |
+
"loss": 0.3639,
|
| 465 |
+
"step": 38000
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 3.47,
|
| 469 |
+
"learning_rate": 2.6528128380815007e-06,
|
| 470 |
+
"loss": 0.3333,
|
| 471 |
+
"step": 38500
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 3.52,
|
| 475 |
+
"learning_rate": 2.427425171294627e-06,
|
| 476 |
+
"loss": 0.3324,
|
| 477 |
+
"step": 39000
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 3.56,
|
| 481 |
+
"learning_rate": 2.2024882798413274e-06,
|
| 482 |
+
"loss": 0.3435,
|
| 483 |
+
"step": 39500
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 3.61,
|
| 487 |
+
"learning_rate": 1.9771006130544537e-06,
|
| 488 |
+
"loss": 0.3512,
|
| 489 |
+
"step": 40000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 3.65,
|
| 493 |
+
"learning_rate": 1.7517129462675803e-06,
|
| 494 |
+
"loss": 0.3388,
|
| 495 |
+
"step": 40500
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 3.7,
|
| 499 |
+
"learning_rate": 1.526325279480707e-06,
|
| 500 |
+
"loss": 0.3421,
|
| 501 |
+
"step": 41000
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 3.74,
|
| 505 |
+
"learning_rate": 1.3009376126938337e-06,
|
| 506 |
+
"loss": 0.3648,
|
| 507 |
+
"step": 41500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 3.79,
|
| 511 |
+
"learning_rate": 1.0760007212405338e-06,
|
| 512 |
+
"loss": 0.3271,
|
| 513 |
+
"step": 42000
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 3.83,
|
| 517 |
+
"learning_rate": 8.506130544536603e-07,
|
| 518 |
+
"loss": 0.321,
|
| 519 |
+
"step": 42500
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 3.88,
|
| 523 |
+
"learning_rate": 6.256761630003607e-07,
|
| 524 |
+
"loss": 0.3432,
|
| 525 |
+
"step": 43000
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 3.92,
|
| 529 |
+
"learning_rate": 4.0028849621348723e-07,
|
| 530 |
+
"loss": 0.322,
|
| 531 |
+
"step": 43500
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 3.97,
|
| 535 |
+
"learning_rate": 1.7490082942661378e-07,
|
| 536 |
+
"loss": 0.3316,
|
| 537 |
+
"step": 44000
|
| 538 |
+
}
|
| 539 |
+
],
|
| 540 |
+
"logging_steps": 500,
|
| 541 |
+
"max_steps": 44368,
|
| 542 |
+
"num_train_epochs": 4,
|
| 543 |
+
"save_steps": 500,
|
| 544 |
+
"total_flos": 6.955379978528563e+16,
|
| 545 |
+
"trial_name": null,
|
| 546 |
+
"trial_params": null
|
| 547 |
+
}
|