checkpoint 10000
Browse files- config.json +1 -1
- optimizer.pt +1 -1
- pytorch_model.bin +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +443 -83
- training_args.bin +1 -1
config.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "t5-base",
|
| 3 |
"architectures": [
|
| 4 |
"T5ForConditionalGeneration"
|
| 5 |
],
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "longcld/t5-base-squad-visquad-aqg",
|
| 3 |
"architectures": [
|
| 4 |
"T5ForConditionalGeneration"
|
| 5 |
],
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1585984129
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:88b9b49019310f798e1923345ce9ec6c14037f0f82e82a08a0d09c39865b1e92
|
| 3 |
size 1585984129
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1561347007
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a8020cd7f123221d4ad2b91e2eace50fd3c94ae3404506f834040aa0c81780ee
|
| 3 |
size 1561347007
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14503
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7c72fa0aadf13ba9bc8d7c1ce47b1a767e46d98c4f2f7feef608e3058c94b662
|
| 3 |
size 14503
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 623
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:18154713747c9de3ce13a33d982ae6adbee296332147d7512416d0719c5c7899
|
| 3 |
size 623
|
trainer_state.json
CHANGED
|
@@ -1,256 +1,616 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
| 10 |
{
|
| 11 |
"epoch": 0.02,
|
| 12 |
-
"learning_rate": 9.
|
| 13 |
-
"loss":
|
| 14 |
"step": 100
|
| 15 |
},
|
| 16 |
{
|
| 17 |
"epoch": 0.04,
|
| 18 |
-
"learning_rate": 9.
|
| 19 |
-
"loss":
|
| 20 |
"step": 200
|
| 21 |
},
|
| 22 |
{
|
| 23 |
"epoch": 0.05,
|
| 24 |
-
"learning_rate": 9.
|
| 25 |
-
"loss":
|
| 26 |
"step": 300
|
| 27 |
},
|
| 28 |
{
|
| 29 |
"epoch": 0.07,
|
| 30 |
-
"learning_rate": 9.
|
| 31 |
-
"loss":
|
| 32 |
"step": 400
|
| 33 |
},
|
| 34 |
{
|
| 35 |
"epoch": 0.09,
|
| 36 |
-
"learning_rate": 9.
|
| 37 |
-
"loss":
|
| 38 |
"step": 500
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.11,
|
| 42 |
-
"learning_rate": 9.
|
| 43 |
-
"loss":
|
| 44 |
"step": 600
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.12,
|
| 48 |
-
"learning_rate": 9.
|
| 49 |
-
"loss":
|
| 50 |
"step": 700
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"epoch": 0.14,
|
| 54 |
-
"learning_rate": 9.
|
| 55 |
-
"loss":
|
| 56 |
"step": 800
|
| 57 |
},
|
| 58 |
{
|
| 59 |
"epoch": 0.16,
|
| 60 |
-
"learning_rate": 9.840566873339238e-
|
| 61 |
-
"loss":
|
| 62 |
"step": 900
|
| 63 |
},
|
| 64 |
{
|
| 65 |
"epoch": 0.18,
|
| 66 |
-
"learning_rate": 9.
|
| 67 |
-
"loss":
|
| 68 |
"step": 1000
|
| 69 |
},
|
| 70 |
{
|
| 71 |
"epoch": 0.19,
|
| 72 |
-
"learning_rate": 9.
|
| 73 |
-
"loss":
|
| 74 |
"step": 1100
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"epoch": 0.21,
|
| 78 |
-
"learning_rate": 9.
|
| 79 |
-
"loss": 4.
|
| 80 |
"step": 1200
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.23,
|
| 84 |
-
"learning_rate": 9.
|
| 85 |
-
"loss": 4.
|
| 86 |
"step": 1300
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.25,
|
| 90 |
-
"learning_rate": 9.
|
| 91 |
-
"loss": 4.
|
| 92 |
"step": 1400
|
| 93 |
},
|
| 94 |
{
|
| 95 |
"epoch": 0.27,
|
| 96 |
-
"learning_rate": 9.
|
| 97 |
-
"loss":
|
| 98 |
"step": 1500
|
| 99 |
},
|
| 100 |
{
|
| 101 |
"epoch": 0.28,
|
| 102 |
-
"learning_rate": 9.
|
| 103 |
-
"loss": 4.
|
| 104 |
"step": 1600
|
| 105 |
},
|
| 106 |
{
|
| 107 |
"epoch": 0.3,
|
| 108 |
-
"learning_rate": 9.
|
| 109 |
-
"loss": 4.
|
| 110 |
"step": 1700
|
| 111 |
},
|
| 112 |
{
|
| 113 |
"epoch": 0.32,
|
| 114 |
-
"learning_rate": 9.681133746678477e-
|
| 115 |
-
"loss":
|
| 116 |
"step": 1800
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 0.34,
|
| 120 |
-
"learning_rate": 9.
|
| 121 |
-
"loss": 4.
|
| 122 |
"step": 1900
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.35,
|
| 126 |
-
"learning_rate": 9.
|
| 127 |
-
"loss": 4.
|
| 128 |
"step": 2000
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.37,
|
| 132 |
-
"learning_rate": 9.
|
| 133 |
-
"loss": 4.
|
| 134 |
"step": 2100
|
| 135 |
},
|
| 136 |
{
|
| 137 |
"epoch": 0.39,
|
| 138 |
-
"learning_rate": 9.
|
| 139 |
-
"loss":
|
| 140 |
"step": 2200
|
| 141 |
},
|
| 142 |
{
|
| 143 |
"epoch": 0.41,
|
| 144 |
-
"learning_rate": 9.
|
| 145 |
-
"loss":
|
| 146 |
"step": 2300
|
| 147 |
},
|
| 148 |
{
|
| 149 |
"epoch": 0.43,
|
| 150 |
-
"learning_rate": 9.
|
| 151 |
-
"loss":
|
| 152 |
"step": 2400
|
| 153 |
},
|
| 154 |
{
|
| 155 |
"epoch": 0.44,
|
| 156 |
-
"learning_rate": 9.
|
| 157 |
-
"loss":
|
| 158 |
"step": 2500
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 0.46,
|
| 162 |
-
"learning_rate": 9.
|
| 163 |
-
"loss":
|
| 164 |
"step": 2600
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.48,
|
| 168 |
-
"learning_rate": 9.521700620017715e-
|
| 169 |
-
"loss":
|
| 170 |
"step": 2700
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.5,
|
| 174 |
-
"learning_rate": 9.
|
| 175 |
-
"loss":
|
| 176 |
"step": 2800
|
| 177 |
},
|
| 178 |
{
|
| 179 |
"epoch": 0.51,
|
| 180 |
-
"learning_rate": 9.
|
| 181 |
-
"loss":
|
| 182 |
"step": 2900
|
| 183 |
},
|
| 184 |
{
|
| 185 |
"epoch": 0.53,
|
| 186 |
-
"learning_rate": 9.
|
| 187 |
-
"loss":
|
| 188 |
"step": 3000
|
| 189 |
},
|
| 190 |
{
|
| 191 |
"epoch": 0.55,
|
| 192 |
-
"learning_rate": 9.
|
| 193 |
-
"loss":
|
| 194 |
"step": 3100
|
| 195 |
},
|
| 196 |
{
|
| 197 |
"epoch": 0.57,
|
| 198 |
-
"learning_rate": 9.433126660761736e-
|
| 199 |
-
"loss":
|
| 200 |
"step": 3200
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 0.58,
|
| 204 |
-
"learning_rate": 9.
|
| 205 |
-
"loss":
|
| 206 |
"step": 3300
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.6,
|
| 210 |
-
"learning_rate": 9.
|
| 211 |
-
"loss":
|
| 212 |
"step": 3400
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.62,
|
| 216 |
-
"learning_rate": 9.
|
| 217 |
-
"loss": 4.
|
| 218 |
"step": 3500
|
| 219 |
},
|
| 220 |
{
|
| 221 |
"epoch": 0.64,
|
| 222 |
-
"learning_rate": 9.
|
| 223 |
-
"loss":
|
| 224 |
"step": 3600
|
| 225 |
},
|
| 226 |
{
|
| 227 |
"epoch": 0.66,
|
| 228 |
-
"learning_rate": 9.
|
| 229 |
-
"loss":
|
| 230 |
"step": 3700
|
| 231 |
},
|
| 232 |
{
|
| 233 |
"epoch": 0.67,
|
| 234 |
-
"learning_rate": 9.
|
| 235 |
-
"loss":
|
| 236 |
"step": 3800
|
| 237 |
},
|
| 238 |
{
|
| 239 |
"epoch": 0.69,
|
| 240 |
-
"learning_rate": 9.
|
| 241 |
-
"loss":
|
| 242 |
"step": 3900
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 0.71,
|
| 246 |
-
"learning_rate": 9.29140832595217e-
|
| 247 |
-
"loss":
|
| 248 |
"step": 4000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 249 |
}
|
| 250 |
],
|
| 251 |
"max_steps": 56450,
|
| 252 |
"num_train_epochs": 10,
|
| 253 |
-
"total_flos":
|
| 254 |
"trial_name": null,
|
| 255 |
"trial_params": null
|
| 256 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.7714450201496834,
|
| 5 |
+
"global_step": 10000,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
| 10 |
{
|
| 11 |
"epoch": 0.02,
|
| 12 |
+
"learning_rate": 9.982285208148804e-07,
|
| 13 |
+
"loss": 4.0734,
|
| 14 |
"step": 100
|
| 15 |
},
|
| 16 |
{
|
| 17 |
"epoch": 0.04,
|
| 18 |
+
"learning_rate": 9.964570416297608e-07,
|
| 19 |
+
"loss": 4.0569,
|
| 20 |
"step": 200
|
| 21 |
},
|
| 22 |
{
|
| 23 |
"epoch": 0.05,
|
| 24 |
+
"learning_rate": 9.946855624446412e-07,
|
| 25 |
+
"loss": 4.0876,
|
| 26 |
"step": 300
|
| 27 |
},
|
| 28 |
{
|
| 29 |
"epoch": 0.07,
|
| 30 |
+
"learning_rate": 9.929140832595216e-07,
|
| 31 |
+
"loss": 4.0653,
|
| 32 |
"step": 400
|
| 33 |
},
|
| 34 |
{
|
| 35 |
"epoch": 0.09,
|
| 36 |
+
"learning_rate": 9.91142604074402e-07,
|
| 37 |
+
"loss": 4.0145,
|
| 38 |
"step": 500
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.11,
|
| 42 |
+
"learning_rate": 9.893711248892826e-07,
|
| 43 |
+
"loss": 4.0653,
|
| 44 |
"step": 600
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.12,
|
| 48 |
+
"learning_rate": 9.875996457041628e-07,
|
| 49 |
+
"loss": 4.0487,
|
| 50 |
"step": 700
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"epoch": 0.14,
|
| 54 |
+
"learning_rate": 9.858281665190434e-07,
|
| 55 |
+
"loss": 4.0187,
|
| 56 |
"step": 800
|
| 57 |
},
|
| 58 |
{
|
| 59 |
"epoch": 0.16,
|
| 60 |
+
"learning_rate": 9.840566873339238e-07,
|
| 61 |
+
"loss": 4.0423,
|
| 62 |
"step": 900
|
| 63 |
},
|
| 64 |
{
|
| 65 |
"epoch": 0.18,
|
| 66 |
+
"learning_rate": 9.822852081488042e-07,
|
| 67 |
+
"loss": 3.9971,
|
| 68 |
"step": 1000
|
| 69 |
},
|
| 70 |
{
|
| 71 |
"epoch": 0.19,
|
| 72 |
+
"learning_rate": 9.805137289636846e-07,
|
| 73 |
+
"loss": 3.946,
|
| 74 |
"step": 1100
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"epoch": 0.21,
|
| 78 |
+
"learning_rate": 9.78742249778565e-07,
|
| 79 |
+
"loss": 4.0005,
|
| 80 |
"step": 1200
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.23,
|
| 84 |
+
"learning_rate": 9.769707705934455e-07,
|
| 85 |
+
"loss": 4.0426,
|
| 86 |
"step": 1300
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.25,
|
| 90 |
+
"learning_rate": 9.751992914083259e-07,
|
| 91 |
+
"loss": 4.0244,
|
| 92 |
"step": 1400
|
| 93 |
},
|
| 94 |
{
|
| 95 |
"epoch": 0.27,
|
| 96 |
+
"learning_rate": 9.734278122232063e-07,
|
| 97 |
+
"loss": 3.9692,
|
| 98 |
"step": 1500
|
| 99 |
},
|
| 100 |
{
|
| 101 |
"epoch": 0.28,
|
| 102 |
+
"learning_rate": 9.716563330380869e-07,
|
| 103 |
+
"loss": 4.0203,
|
| 104 |
"step": 1600
|
| 105 |
},
|
| 106 |
{
|
| 107 |
"epoch": 0.3,
|
| 108 |
+
"learning_rate": 9.69884853852967e-07,
|
| 109 |
+
"loss": 4.0316,
|
| 110 |
"step": 1700
|
| 111 |
},
|
| 112 |
{
|
| 113 |
"epoch": 0.32,
|
| 114 |
+
"learning_rate": 9.681133746678477e-07,
|
| 115 |
+
"loss": 3.9496,
|
| 116 |
"step": 1800
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 0.34,
|
| 120 |
+
"learning_rate": 9.66341895482728e-07,
|
| 121 |
+
"loss": 4.035,
|
| 122 |
"step": 1900
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.35,
|
| 126 |
+
"learning_rate": 9.645704162976085e-07,
|
| 127 |
+
"loss": 4.0034,
|
| 128 |
"step": 2000
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.37,
|
| 132 |
+
"learning_rate": 9.62798937112489e-07,
|
| 133 |
+
"loss": 4.0186,
|
| 134 |
"step": 2100
|
| 135 |
},
|
| 136 |
{
|
| 137 |
"epoch": 0.39,
|
| 138 |
+
"learning_rate": 9.610274579273693e-07,
|
| 139 |
+
"loss": 3.9873,
|
| 140 |
"step": 2200
|
| 141 |
},
|
| 142 |
{
|
| 143 |
"epoch": 0.41,
|
| 144 |
+
"learning_rate": 9.592559787422497e-07,
|
| 145 |
+
"loss": 3.9791,
|
| 146 |
"step": 2300
|
| 147 |
},
|
| 148 |
{
|
| 149 |
"epoch": 0.43,
|
| 150 |
+
"learning_rate": 9.574844995571301e-07,
|
| 151 |
+
"loss": 3.9925,
|
| 152 |
"step": 2400
|
| 153 |
},
|
| 154 |
{
|
| 155 |
"epoch": 0.44,
|
| 156 |
+
"learning_rate": 9.557130203720105e-07,
|
| 157 |
+
"loss": 3.9703,
|
| 158 |
"step": 2500
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 0.46,
|
| 162 |
+
"learning_rate": 9.53941541186891e-07,
|
| 163 |
+
"loss": 3.9798,
|
| 164 |
"step": 2600
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.48,
|
| 168 |
+
"learning_rate": 9.521700620017715e-07,
|
| 169 |
+
"loss": 3.9557,
|
| 170 |
"step": 2700
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.5,
|
| 174 |
+
"learning_rate": 9.503985828166519e-07,
|
| 175 |
+
"loss": 3.9183,
|
| 176 |
"step": 2800
|
| 177 |
},
|
| 178 |
{
|
| 179 |
"epoch": 0.51,
|
| 180 |
+
"learning_rate": 9.486271036315323e-07,
|
| 181 |
+
"loss": 3.9861,
|
| 182 |
"step": 2900
|
| 183 |
},
|
| 184 |
{
|
| 185 |
"epoch": 0.53,
|
| 186 |
+
"learning_rate": 9.468556244464128e-07,
|
| 187 |
+
"loss": 3.9041,
|
| 188 |
"step": 3000
|
| 189 |
},
|
| 190 |
{
|
| 191 |
"epoch": 0.55,
|
| 192 |
+
"learning_rate": 9.450841452612931e-07,
|
| 193 |
+
"loss": 3.911,
|
| 194 |
"step": 3100
|
| 195 |
},
|
| 196 |
{
|
| 197 |
"epoch": 0.57,
|
| 198 |
+
"learning_rate": 9.433126660761736e-07,
|
| 199 |
+
"loss": 3.9845,
|
| 200 |
"step": 3200
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 0.58,
|
| 204 |
+
"learning_rate": 9.41541186891054e-07,
|
| 205 |
+
"loss": 3.9742,
|
| 206 |
"step": 3300
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.6,
|
| 210 |
+
"learning_rate": 9.397697077059344e-07,
|
| 211 |
+
"loss": 3.9125,
|
| 212 |
"step": 3400
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.62,
|
| 216 |
+
"learning_rate": 9.379982285208148e-07,
|
| 217 |
+
"loss": 4.0199,
|
| 218 |
"step": 3500
|
| 219 |
},
|
| 220 |
{
|
| 221 |
"epoch": 0.64,
|
| 222 |
+
"learning_rate": 9.362267493356952e-07,
|
| 223 |
+
"loss": 3.9746,
|
| 224 |
"step": 3600
|
| 225 |
},
|
| 226 |
{
|
| 227 |
"epoch": 0.66,
|
| 228 |
+
"learning_rate": 9.344552701505756e-07,
|
| 229 |
+
"loss": 3.9881,
|
| 230 |
"step": 3700
|
| 231 |
},
|
| 232 |
{
|
| 233 |
"epoch": 0.67,
|
| 234 |
+
"learning_rate": 9.326837909654561e-07,
|
| 235 |
+
"loss": 3.966,
|
| 236 |
"step": 3800
|
| 237 |
},
|
| 238 |
{
|
| 239 |
"epoch": 0.69,
|
| 240 |
+
"learning_rate": 9.309123117803365e-07,
|
| 241 |
+
"loss": 3.8963,
|
| 242 |
"step": 3900
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 0.71,
|
| 246 |
+
"learning_rate": 9.29140832595217e-07,
|
| 247 |
+
"loss": 3.9228,
|
| 248 |
"step": 4000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.73,
|
| 252 |
+
"learning_rate": 9.273693534100974e-07,
|
| 253 |
+
"loss": 4.0232,
|
| 254 |
+
"step": 4100
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.74,
|
| 258 |
+
"learning_rate": 9.255978742249778e-07,
|
| 259 |
+
"loss": 3.9901,
|
| 260 |
+
"step": 4200
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.76,
|
| 264 |
+
"learning_rate": 9.238263950398583e-07,
|
| 265 |
+
"loss": 3.9779,
|
| 266 |
+
"step": 4300
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.78,
|
| 270 |
+
"learning_rate": 9.220549158547387e-07,
|
| 271 |
+
"loss": 3.9652,
|
| 272 |
+
"step": 4400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.8,
|
| 276 |
+
"learning_rate": 9.202834366696191e-07,
|
| 277 |
+
"loss": 4.0147,
|
| 278 |
+
"step": 4500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.81,
|
| 282 |
+
"learning_rate": 9.185119574844995e-07,
|
| 283 |
+
"loss": 4.0175,
|
| 284 |
+
"step": 4600
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.83,
|
| 288 |
+
"learning_rate": 9.167404782993799e-07,
|
| 289 |
+
"loss": 4.0071,
|
| 290 |
+
"step": 4700
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.85,
|
| 294 |
+
"learning_rate": 9.149689991142604e-07,
|
| 295 |
+
"loss": 3.9689,
|
| 296 |
+
"step": 4800
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.87,
|
| 300 |
+
"learning_rate": 9.131975199291407e-07,
|
| 301 |
+
"loss": 4.0114,
|
| 302 |
+
"step": 4900
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.89,
|
| 306 |
+
"learning_rate": 9.114260407440212e-07,
|
| 307 |
+
"loss": 3.9757,
|
| 308 |
+
"step": 5000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.9,
|
| 312 |
+
"learning_rate": 9.096545615589016e-07,
|
| 313 |
+
"loss": 4.0132,
|
| 314 |
+
"step": 5100
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.92,
|
| 318 |
+
"learning_rate": 9.07883082373782e-07,
|
| 319 |
+
"loss": 4.035,
|
| 320 |
+
"step": 5200
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.94,
|
| 324 |
+
"learning_rate": 9.061116031886626e-07,
|
| 325 |
+
"loss": 4.0601,
|
| 326 |
+
"step": 5300
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.96,
|
| 330 |
+
"learning_rate": 9.043401240035429e-07,
|
| 331 |
+
"loss": 3.9797,
|
| 332 |
+
"step": 5400
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.97,
|
| 336 |
+
"learning_rate": 9.025686448184234e-07,
|
| 337 |
+
"loss": 3.9721,
|
| 338 |
+
"step": 5500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.99,
|
| 342 |
+
"learning_rate": 9.007971656333038e-07,
|
| 343 |
+
"loss": 3.9189,
|
| 344 |
+
"step": 5600
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 1.01,
|
| 348 |
+
"learning_rate": 8.990256864481842e-07,
|
| 349 |
+
"loss": 3.9462,
|
| 350 |
+
"step": 5700
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.03,
|
| 354 |
+
"learning_rate": 8.972542072630647e-07,
|
| 355 |
+
"loss": 4.0232,
|
| 356 |
+
"step": 5800
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 1.05,
|
| 360 |
+
"learning_rate": 8.95482728077945e-07,
|
| 361 |
+
"loss": 3.9716,
|
| 362 |
+
"step": 5900
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.06,
|
| 366 |
+
"learning_rate": 8.937112488928254e-07,
|
| 367 |
+
"loss": 3.9541,
|
| 368 |
+
"step": 6000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.08,
|
| 372 |
+
"learning_rate": 8.919397697077059e-07,
|
| 373 |
+
"loss": 3.9548,
|
| 374 |
+
"step": 6100
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.1,
|
| 378 |
+
"learning_rate": 8.901682905225863e-07,
|
| 379 |
+
"loss": 3.9162,
|
| 380 |
+
"step": 6200
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.12,
|
| 384 |
+
"learning_rate": 8.883968113374667e-07,
|
| 385 |
+
"loss": 3.9678,
|
| 386 |
+
"step": 6300
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 1.13,
|
| 390 |
+
"learning_rate": 8.866253321523471e-07,
|
| 391 |
+
"loss": 3.9135,
|
| 392 |
+
"step": 6400
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.15,
|
| 396 |
+
"learning_rate": 8.848538529672275e-07,
|
| 397 |
+
"loss": 3.9899,
|
| 398 |
+
"step": 6500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 1.17,
|
| 402 |
+
"learning_rate": 8.83082373782108e-07,
|
| 403 |
+
"loss": 4.017,
|
| 404 |
+
"step": 6600
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.19,
|
| 408 |
+
"learning_rate": 8.813108945969885e-07,
|
| 409 |
+
"loss": 3.9619,
|
| 410 |
+
"step": 6700
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.2,
|
| 414 |
+
"learning_rate": 8.795394154118689e-07,
|
| 415 |
+
"loss": 3.9598,
|
| 416 |
+
"step": 6800
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.22,
|
| 420 |
+
"learning_rate": 8.777679362267493e-07,
|
| 421 |
+
"loss": 3.984,
|
| 422 |
+
"step": 6900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.24,
|
| 426 |
+
"learning_rate": 8.759964570416297e-07,
|
| 427 |
+
"loss": 3.9314,
|
| 428 |
+
"step": 7000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.26,
|
| 432 |
+
"learning_rate": 8.742249778565102e-07,
|
| 433 |
+
"loss": 3.9481,
|
| 434 |
+
"step": 7100
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.28,
|
| 438 |
+
"learning_rate": 8.724534986713906e-07,
|
| 439 |
+
"loss": 3.9485,
|
| 440 |
+
"step": 7200
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 1.29,
|
| 444 |
+
"learning_rate": 8.70682019486271e-07,
|
| 445 |
+
"loss": 4.0013,
|
| 446 |
+
"step": 7300
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 1.31,
|
| 450 |
+
"learning_rate": 8.689105403011514e-07,
|
| 451 |
+
"loss": 3.9394,
|
| 452 |
+
"step": 7400
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.33,
|
| 456 |
+
"learning_rate": 8.671390611160318e-07,
|
| 457 |
+
"loss": 3.947,
|
| 458 |
+
"step": 7500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.35,
|
| 462 |
+
"learning_rate": 8.653675819309123e-07,
|
| 463 |
+
"loss": 3.9169,
|
| 464 |
+
"step": 7600
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.36,
|
| 468 |
+
"learning_rate": 8.635961027457926e-07,
|
| 469 |
+
"loss": 3.9976,
|
| 470 |
+
"step": 7700
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 1.38,
|
| 474 |
+
"learning_rate": 8.618246235606731e-07,
|
| 475 |
+
"loss": 3.8938,
|
| 476 |
+
"step": 7800
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.4,
|
| 480 |
+
"learning_rate": 8.600531443755535e-07,
|
| 481 |
+
"loss": 3.9433,
|
| 482 |
+
"step": 7900
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 1.42,
|
| 486 |
+
"learning_rate": 8.58281665190434e-07,
|
| 487 |
+
"loss": 3.9907,
|
| 488 |
+
"step": 8000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 1.43,
|
| 492 |
+
"learning_rate": 8.565101860053145e-07,
|
| 493 |
+
"loss": 3.9559,
|
| 494 |
+
"step": 8100
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.45,
|
| 498 |
+
"learning_rate": 8.547387068201948e-07,
|
| 499 |
+
"loss": 3.9386,
|
| 500 |
+
"step": 8200
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.47,
|
| 504 |
+
"learning_rate": 8.529672276350752e-07,
|
| 505 |
+
"loss": 3.969,
|
| 506 |
+
"step": 8300
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.49,
|
| 510 |
+
"learning_rate": 8.511957484499557e-07,
|
| 511 |
+
"loss": 3.932,
|
| 512 |
+
"step": 8400
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 1.51,
|
| 516 |
+
"learning_rate": 8.494242692648361e-07,
|
| 517 |
+
"loss": 3.9155,
|
| 518 |
+
"step": 8500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 1.52,
|
| 522 |
+
"learning_rate": 8.476527900797166e-07,
|
| 523 |
+
"loss": 3.9502,
|
| 524 |
+
"step": 8600
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 1.54,
|
| 528 |
+
"learning_rate": 8.458813108945969e-07,
|
| 529 |
+
"loss": 3.9153,
|
| 530 |
+
"step": 8700
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 1.56,
|
| 534 |
+
"learning_rate": 8.441098317094773e-07,
|
| 535 |
+
"loss": 3.945,
|
| 536 |
+
"step": 8800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.58,
|
| 540 |
+
"learning_rate": 8.423383525243578e-07,
|
| 541 |
+
"loss": 3.8924,
|
| 542 |
+
"step": 8900
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.59,
|
| 546 |
+
"learning_rate": 8.405668733392382e-07,
|
| 547 |
+
"loss": 3.9133,
|
| 548 |
+
"step": 9000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.61,
|
| 552 |
+
"learning_rate": 8.387953941541186e-07,
|
| 553 |
+
"loss": 3.8818,
|
| 554 |
+
"step": 9100
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 1.63,
|
| 558 |
+
"learning_rate": 8.37023914968999e-07,
|
| 559 |
+
"loss": 3.9661,
|
| 560 |
+
"step": 9200
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.65,
|
| 564 |
+
"learning_rate": 8.352524357838795e-07,
|
| 565 |
+
"loss": 3.9492,
|
| 566 |
+
"step": 9300
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 1.67,
|
| 570 |
+
"learning_rate": 8.3348095659876e-07,
|
| 571 |
+
"loss": 3.9046,
|
| 572 |
+
"step": 9400
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 1.68,
|
| 576 |
+
"learning_rate": 8.317094774136404e-07,
|
| 577 |
+
"loss": 3.9662,
|
| 578 |
+
"step": 9500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.7,
|
| 582 |
+
"learning_rate": 8.299379982285208e-07,
|
| 583 |
+
"loss": 3.95,
|
| 584 |
+
"step": 9600
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.72,
|
| 588 |
+
"learning_rate": 8.281665190434012e-07,
|
| 589 |
+
"loss": 3.9136,
|
| 590 |
+
"step": 9700
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.74,
|
| 594 |
+
"learning_rate": 8.263950398582816e-07,
|
| 595 |
+
"loss": 3.9176,
|
| 596 |
+
"step": 9800
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 1.75,
|
| 600 |
+
"learning_rate": 8.246235606731621e-07,
|
| 601 |
+
"loss": 3.8828,
|
| 602 |
+
"step": 9900
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 1.77,
|
| 606 |
+
"learning_rate": 8.228520814880425e-07,
|
| 607 |
+
"loss": 3.9087,
|
| 608 |
+
"step": 10000
|
| 609 |
}
|
| 610 |
],
|
| 611 |
"max_steps": 56450,
|
| 612 |
"num_train_epochs": 10,
|
| 613 |
+
"total_flos": 3.7155143735731814e+17,
|
| 614 |
"trial_name": null,
|
| 615 |
"trial_params": null
|
| 616 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2607
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f854c59c995d85e209b722bb65fba80fd8acac98f5b5019ebbee26de24fb5e0b
|
| 3 |
size 2607
|