update loss 2.6
Browse files- optimizer.pt +1 -1
- pytorch_model.bin +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +513 -63
- training_args.bin +1 -1
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:d1fb31fbea31bbfe6b18f3d4365f1ee475802779ad933635eb16e050553bd591
|
| 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 sha256:23e2bc4f447c0ad25e35dd41b6355a29c94ff2ac8a9769ac4434883d89693542
|
| 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:a3ab1697cb3299e8934b6a15403084174c2197dcc391d6efe7ba0b672f2686c9
|
| 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 sha256:55ee0954cb418fe695e4c943d4e61496aec15e4c69de4bfbcbf1d67b5f134d04
|
| 3 |
size 623
|
trainer_state.json
CHANGED
|
@@ -1,196 +1,646 @@
|
|
| 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":
|
| 13 |
-
"loss": 3.
|
| 14 |
"step": 100
|
| 15 |
},
|
| 16 |
{
|
| 17 |
"epoch": 0.04,
|
| 18 |
-
"learning_rate":
|
| 19 |
-
"loss": 3.
|
| 20 |
"step": 200
|
| 21 |
},
|
| 22 |
{
|
| 23 |
"epoch": 0.05,
|
| 24 |
-
"learning_rate":
|
| 25 |
-
"loss": 3.
|
| 26 |
"step": 300
|
| 27 |
},
|
| 28 |
{
|
| 29 |
"epoch": 0.07,
|
| 30 |
-
"learning_rate":
|
| 31 |
-
"loss": 3.
|
| 32 |
"step": 400
|
| 33 |
},
|
| 34 |
{
|
| 35 |
"epoch": 0.09,
|
| 36 |
-
"learning_rate":
|
| 37 |
-
"loss": 3.
|
| 38 |
"step": 500
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.11,
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss": 3.
|
| 44 |
"step": 600
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.12,
|
| 48 |
-
"learning_rate":
|
| 49 |
-
"loss": 3.
|
| 50 |
"step": 700
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"epoch": 0.14,
|
| 54 |
-
"learning_rate":
|
| 55 |
-
"loss": 3.
|
| 56 |
"step": 800
|
| 57 |
},
|
| 58 |
{
|
| 59 |
"epoch": 0.16,
|
| 60 |
-
"learning_rate":
|
| 61 |
-
"loss": 3.
|
| 62 |
"step": 900
|
| 63 |
},
|
| 64 |
{
|
| 65 |
"epoch": 0.18,
|
| 66 |
-
"learning_rate":
|
| 67 |
-
"loss": 3.
|
| 68 |
"step": 1000
|
| 69 |
},
|
| 70 |
{
|
| 71 |
"epoch": 0.19,
|
| 72 |
-
"learning_rate":
|
| 73 |
-
"loss": 3.
|
| 74 |
"step": 1100
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"epoch": 0.21,
|
| 78 |
-
"learning_rate":
|
| 79 |
-
"loss": 3.
|
| 80 |
"step": 1200
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.23,
|
| 84 |
-
"learning_rate":
|
| 85 |
-
"loss": 3.
|
| 86 |
"step": 1300
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.25,
|
| 90 |
-
"learning_rate":
|
| 91 |
-
"loss": 3.
|
| 92 |
"step": 1400
|
| 93 |
},
|
| 94 |
{
|
| 95 |
"epoch": 0.27,
|
| 96 |
-
"learning_rate":
|
| 97 |
-
"loss": 3.
|
| 98 |
"step": 1500
|
| 99 |
},
|
| 100 |
{
|
| 101 |
"epoch": 0.28,
|
| 102 |
-
"learning_rate":
|
| 103 |
-
"loss": 3.
|
| 104 |
"step": 1600
|
| 105 |
},
|
| 106 |
{
|
| 107 |
"epoch": 0.3,
|
| 108 |
-
"learning_rate":
|
| 109 |
-
"loss": 3.
|
| 110 |
"step": 1700
|
| 111 |
},
|
| 112 |
{
|
| 113 |
"epoch": 0.32,
|
| 114 |
-
"learning_rate":
|
| 115 |
-
"loss": 3.
|
| 116 |
"step": 1800
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 0.34,
|
| 120 |
-
"learning_rate":
|
| 121 |
-
"loss": 3.
|
| 122 |
"step": 1900
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.35,
|
| 126 |
-
"learning_rate":
|
| 127 |
-
"loss": 3.
|
| 128 |
"step": 2000
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.37,
|
| 132 |
-
"learning_rate":
|
| 133 |
-
"loss": 3.
|
| 134 |
"step": 2100
|
| 135 |
},
|
| 136 |
{
|
| 137 |
"epoch": 0.39,
|
| 138 |
-
"learning_rate":
|
| 139 |
-
"loss":
|
| 140 |
"step": 2200
|
| 141 |
},
|
| 142 |
{
|
| 143 |
"epoch": 0.41,
|
| 144 |
-
"learning_rate":
|
| 145 |
-
"loss":
|
| 146 |
"step": 2300
|
| 147 |
},
|
| 148 |
{
|
| 149 |
"epoch": 0.43,
|
| 150 |
-
"learning_rate":
|
| 151 |
-
"loss":
|
| 152 |
"step": 2400
|
| 153 |
},
|
| 154 |
{
|
| 155 |
"epoch": 0.44,
|
| 156 |
-
"learning_rate":
|
| 157 |
-
"loss":
|
| 158 |
"step": 2500
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 0.46,
|
| 162 |
-
"learning_rate":
|
| 163 |
-
"loss":
|
| 164 |
"step": 2600
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.48,
|
| 168 |
-
"learning_rate":
|
| 169 |
-
"loss":
|
| 170 |
"step": 2700
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.5,
|
| 174 |
-
"learning_rate":
|
| 175 |
-
"loss":
|
| 176 |
"step": 2800
|
| 177 |
},
|
| 178 |
{
|
| 179 |
"epoch": 0.51,
|
| 180 |
-
"learning_rate":
|
| 181 |
-
"loss":
|
| 182 |
"step": 2900
|
| 183 |
},
|
| 184 |
{
|
| 185 |
"epoch": 0.53,
|
| 186 |
-
"learning_rate":
|
| 187 |
-
"loss":
|
| 188 |
"step": 3000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 189 |
}
|
| 190 |
],
|
| 191 |
"max_steps": 56450,
|
| 192 |
"num_train_epochs": 10,
|
| 193 |
-
"total_flos":
|
| 194 |
"trial_name": null,
|
| 195 |
"trial_params": null
|
| 196 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.8600150569062486,
|
| 5 |
+
"global_step": 10500,
|
| 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": 4.991142604074402e-05,
|
| 13 |
+
"loss": 3.2628,
|
| 14 |
"step": 100
|
| 15 |
},
|
| 16 |
{
|
| 17 |
"epoch": 0.04,
|
| 18 |
+
"learning_rate": 4.982285208148805e-05,
|
| 19 |
+
"loss": 3.2467,
|
| 20 |
"step": 200
|
| 21 |
},
|
| 22 |
{
|
| 23 |
"epoch": 0.05,
|
| 24 |
+
"learning_rate": 4.973427812223206e-05,
|
| 25 |
+
"loss": 3.2697,
|
| 26 |
"step": 300
|
| 27 |
},
|
| 28 |
{
|
| 29 |
"epoch": 0.07,
|
| 30 |
+
"learning_rate": 4.964570416297609e-05,
|
| 31 |
+
"loss": 3.2468,
|
| 32 |
"step": 400
|
| 33 |
},
|
| 34 |
{
|
| 35 |
"epoch": 0.09,
|
| 36 |
+
"learning_rate": 4.9557130203720107e-05,
|
| 37 |
+
"loss": 3.1687,
|
| 38 |
"step": 500
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.11,
|
| 42 |
+
"learning_rate": 4.946855624446413e-05,
|
| 43 |
+
"loss": 3.22,
|
| 44 |
"step": 600
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.12,
|
| 48 |
+
"learning_rate": 4.9379982285208146e-05,
|
| 49 |
+
"loss": 3.1855,
|
| 50 |
"step": 700
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"epoch": 0.14,
|
| 54 |
+
"learning_rate": 4.929140832595217e-05,
|
| 55 |
+
"loss": 3.1569,
|
| 56 |
"step": 800
|
| 57 |
},
|
| 58 |
{
|
| 59 |
"epoch": 0.16,
|
| 60 |
+
"learning_rate": 4.920283436669619e-05,
|
| 61 |
+
"loss": 3.1495,
|
| 62 |
"step": 900
|
| 63 |
},
|
| 64 |
{
|
| 65 |
"epoch": 0.18,
|
| 66 |
+
"learning_rate": 4.911426040744022e-05,
|
| 67 |
+
"loss": 3.1099,
|
| 68 |
"step": 1000
|
| 69 |
},
|
| 70 |
{
|
| 71 |
"epoch": 0.19,
|
| 72 |
+
"learning_rate": 4.902568644818424e-05,
|
| 73 |
+
"loss": 3.0485,
|
| 74 |
"step": 1100
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"epoch": 0.21,
|
| 78 |
+
"learning_rate": 4.8937112488928256e-05,
|
| 79 |
+
"loss": 3.1011,
|
| 80 |
"step": 1200
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.23,
|
| 84 |
+
"learning_rate": 4.8848538529672276e-05,
|
| 85 |
+
"loss": 3.1181,
|
| 86 |
"step": 1300
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.25,
|
| 90 |
+
"learning_rate": 4.87599645704163e-05,
|
| 91 |
+
"loss": 3.0912,
|
| 92 |
"step": 1400
|
| 93 |
},
|
| 94 |
{
|
| 95 |
"epoch": 0.27,
|
| 96 |
+
"learning_rate": 4.867139061116032e-05,
|
| 97 |
+
"loss": 3.0443,
|
| 98 |
"step": 1500
|
| 99 |
},
|
| 100 |
{
|
| 101 |
"epoch": 0.28,
|
| 102 |
+
"learning_rate": 4.858281665190435e-05,
|
| 103 |
+
"loss": 3.0694,
|
| 104 |
"step": 1600
|
| 105 |
},
|
| 106 |
{
|
| 107 |
"epoch": 0.3,
|
| 108 |
+
"learning_rate": 4.849424269264836e-05,
|
| 109 |
+
"loss": 3.0743,
|
| 110 |
"step": 1700
|
| 111 |
},
|
| 112 |
{
|
| 113 |
"epoch": 0.32,
|
| 114 |
+
"learning_rate": 4.840566873339239e-05,
|
| 115 |
+
"loss": 3.0058,
|
| 116 |
"step": 1800
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 0.34,
|
| 120 |
+
"learning_rate": 4.8317094774136406e-05,
|
| 121 |
+
"loss": 3.0656,
|
| 122 |
"step": 1900
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.35,
|
| 126 |
+
"learning_rate": 4.8228520814880426e-05,
|
| 127 |
+
"loss": 3.0223,
|
| 128 |
"step": 2000
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.37,
|
| 132 |
+
"learning_rate": 4.8139946855624445e-05,
|
| 133 |
+
"loss": 3.0125,
|
| 134 |
"step": 2100
|
| 135 |
},
|
| 136 |
{
|
| 137 |
"epoch": 0.39,
|
| 138 |
+
"learning_rate": 4.805137289636847e-05,
|
| 139 |
+
"loss": 2.9976,
|
| 140 |
"step": 2200
|
| 141 |
},
|
| 142 |
{
|
| 143 |
"epoch": 0.41,
|
| 144 |
+
"learning_rate": 4.796279893711249e-05,
|
| 145 |
+
"loss": 2.9885,
|
| 146 |
"step": 2300
|
| 147 |
},
|
| 148 |
{
|
| 149 |
"epoch": 0.43,
|
| 150 |
+
"learning_rate": 4.787422497785651e-05,
|
| 151 |
+
"loss": 2.9921,
|
| 152 |
"step": 2400
|
| 153 |
},
|
| 154 |
{
|
| 155 |
"epoch": 0.44,
|
| 156 |
+
"learning_rate": 4.778565101860054e-05,
|
| 157 |
+
"loss": 2.9764,
|
| 158 |
"step": 2500
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 0.46,
|
| 162 |
+
"learning_rate": 4.769707705934455e-05,
|
| 163 |
+
"loss": 2.9695,
|
| 164 |
"step": 2600
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.48,
|
| 168 |
+
"learning_rate": 4.7608503100088576e-05,
|
| 169 |
+
"loss": 2.9583,
|
| 170 |
"step": 2700
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.5,
|
| 174 |
+
"learning_rate": 4.7519929140832595e-05,
|
| 175 |
+
"loss": 2.9083,
|
| 176 |
"step": 2800
|
| 177 |
},
|
| 178 |
{
|
| 179 |
"epoch": 0.51,
|
| 180 |
+
"learning_rate": 4.743135518157662e-05,
|
| 181 |
+
"loss": 2.9681,
|
| 182 |
"step": 2900
|
| 183 |
},
|
| 184 |
{
|
| 185 |
"epoch": 0.53,
|
| 186 |
+
"learning_rate": 4.734278122232064e-05,
|
| 187 |
+
"loss": 2.8814,
|
| 188 |
"step": 3000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.55,
|
| 192 |
+
"learning_rate": 4.725420726306466e-05,
|
| 193 |
+
"loss": 2.9217,
|
| 194 |
+
"step": 3100
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.57,
|
| 198 |
+
"learning_rate": 4.716563330380868e-05,
|
| 199 |
+
"loss": 2.965,
|
| 200 |
+
"step": 3200
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.58,
|
| 204 |
+
"learning_rate": 4.7077059344552706e-05,
|
| 205 |
+
"loss": 2.9396,
|
| 206 |
+
"step": 3300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.6,
|
| 210 |
+
"learning_rate": 4.6988485385296726e-05,
|
| 211 |
+
"loss": 2.903,
|
| 212 |
+
"step": 3400
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.62,
|
| 216 |
+
"learning_rate": 4.6899911426040745e-05,
|
| 217 |
+
"loss": 2.9866,
|
| 218 |
+
"step": 3500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.64,
|
| 222 |
+
"learning_rate": 4.6811337466784765e-05,
|
| 223 |
+
"loss": 2.9458,
|
| 224 |
+
"step": 3600
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.66,
|
| 228 |
+
"learning_rate": 4.672276350752879e-05,
|
| 229 |
+
"loss": 2.9344,
|
| 230 |
+
"step": 3700
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.67,
|
| 234 |
+
"learning_rate": 4.663418954827281e-05,
|
| 235 |
+
"loss": 2.9194,
|
| 236 |
+
"step": 3800
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.69,
|
| 240 |
+
"learning_rate": 4.6545615589016837e-05,
|
| 241 |
+
"loss": 2.8582,
|
| 242 |
+
"step": 3900
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.71,
|
| 246 |
+
"learning_rate": 4.645704162976085e-05,
|
| 247 |
+
"loss": 2.8646,
|
| 248 |
+
"step": 4000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.73,
|
| 252 |
+
"learning_rate": 4.6368467670504875e-05,
|
| 253 |
+
"loss": 2.8961,
|
| 254 |
+
"step": 4100
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.74,
|
| 258 |
+
"learning_rate": 4.6279893711248895e-05,
|
| 259 |
+
"loss": 2.8787,
|
| 260 |
+
"step": 4200
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.76,
|
| 264 |
+
"learning_rate": 4.6191319751992914e-05,
|
| 265 |
+
"loss": 2.8538,
|
| 266 |
+
"step": 4300
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.78,
|
| 270 |
+
"learning_rate": 4.610274579273694e-05,
|
| 271 |
+
"loss": 2.8209,
|
| 272 |
+
"step": 4400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.8,
|
| 276 |
+
"learning_rate": 4.601417183348096e-05,
|
| 277 |
+
"loss": 2.8836,
|
| 278 |
+
"step": 4500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.81,
|
| 282 |
+
"learning_rate": 4.592559787422498e-05,
|
| 283 |
+
"loss": 2.8679,
|
| 284 |
+
"step": 4600
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.83,
|
| 288 |
+
"learning_rate": 4.5837023914969e-05,
|
| 289 |
+
"loss": 2.8646,
|
| 290 |
+
"step": 4700
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.85,
|
| 294 |
+
"learning_rate": 4.5748449955713025e-05,
|
| 295 |
+
"loss": 2.8165,
|
| 296 |
+
"step": 4800
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.87,
|
| 300 |
+
"learning_rate": 4.565987599645704e-05,
|
| 301 |
+
"loss": 2.858,
|
| 302 |
+
"step": 4900
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.89,
|
| 306 |
+
"learning_rate": 4.5571302037201064e-05,
|
| 307 |
+
"loss": 2.8192,
|
| 308 |
+
"step": 5000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.9,
|
| 312 |
+
"learning_rate": 4.5482728077945084e-05,
|
| 313 |
+
"loss": 2.8457,
|
| 314 |
+
"step": 5100
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.92,
|
| 318 |
+
"learning_rate": 4.539415411868911e-05,
|
| 319 |
+
"loss": 2.8423,
|
| 320 |
+
"step": 5200
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.94,
|
| 324 |
+
"learning_rate": 4.530558015943313e-05,
|
| 325 |
+
"loss": 2.8714,
|
| 326 |
+
"step": 5300
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.96,
|
| 330 |
+
"learning_rate": 4.521700620017715e-05,
|
| 331 |
+
"loss": 2.8131,
|
| 332 |
+
"step": 5400
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.97,
|
| 336 |
+
"learning_rate": 4.512843224092117e-05,
|
| 337 |
+
"loss": 2.7967,
|
| 338 |
+
"step": 5500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.99,
|
| 342 |
+
"learning_rate": 4.5039858281665195e-05,
|
| 343 |
+
"loss": 2.7518,
|
| 344 |
+
"step": 5600
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 1.01,
|
| 348 |
+
"learning_rate": 4.4951284322409214e-05,
|
| 349 |
+
"loss": 2.7624,
|
| 350 |
+
"step": 5700
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.03,
|
| 354 |
+
"learning_rate": 4.486271036315324e-05,
|
| 355 |
+
"loss": 2.8126,
|
| 356 |
+
"step": 5800
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 1.05,
|
| 360 |
+
"learning_rate": 4.477413640389725e-05,
|
| 361 |
+
"loss": 2.7712,
|
| 362 |
+
"step": 5900
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.06,
|
| 366 |
+
"learning_rate": 4.468556244464128e-05,
|
| 367 |
+
"loss": 2.7397,
|
| 368 |
+
"step": 6000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.08,
|
| 372 |
+
"learning_rate": 4.45969884853853e-05,
|
| 373 |
+
"loss": 2.7596,
|
| 374 |
+
"step": 6100
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.1,
|
| 378 |
+
"learning_rate": 4.450841452612932e-05,
|
| 379 |
+
"loss": 2.7189,
|
| 380 |
+
"step": 6200
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.12,
|
| 384 |
+
"learning_rate": 4.441984056687334e-05,
|
| 385 |
+
"loss": 2.7625,
|
| 386 |
+
"step": 6300
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 1.13,
|
| 390 |
+
"learning_rate": 4.4331266607617364e-05,
|
| 391 |
+
"loss": 2.7192,
|
| 392 |
+
"step": 6400
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.15,
|
| 396 |
+
"learning_rate": 4.4242692648361384e-05,
|
| 397 |
+
"loss": 2.7639,
|
| 398 |
+
"step": 6500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 1.17,
|
| 402 |
+
"learning_rate": 4.41541186891054e-05,
|
| 403 |
+
"loss": 2.7702,
|
| 404 |
+
"step": 6600
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.19,
|
| 408 |
+
"learning_rate": 4.406554472984943e-05,
|
| 409 |
+
"loss": 2.7265,
|
| 410 |
+
"step": 6700
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.2,
|
| 414 |
+
"learning_rate": 4.397697077059344e-05,
|
| 415 |
+
"loss": 2.7297,
|
| 416 |
+
"step": 6800
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.22,
|
| 420 |
+
"learning_rate": 4.388839681133747e-05,
|
| 421 |
+
"loss": 2.7263,
|
| 422 |
+
"step": 6900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.24,
|
| 426 |
+
"learning_rate": 4.379982285208149e-05,
|
| 427 |
+
"loss": 2.6887,
|
| 428 |
+
"step": 7000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.26,
|
| 432 |
+
"learning_rate": 4.3711248892825514e-05,
|
| 433 |
+
"loss": 2.7184,
|
| 434 |
+
"step": 7100
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.28,
|
| 438 |
+
"learning_rate": 4.3622674933569533e-05,
|
| 439 |
+
"loss": 2.7053,
|
| 440 |
+
"step": 7200
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 1.29,
|
| 444 |
+
"learning_rate": 4.353410097431355e-05,
|
| 445 |
+
"loss": 2.737,
|
| 446 |
+
"step": 7300
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 1.31,
|
| 450 |
+
"learning_rate": 4.344552701505757e-05,
|
| 451 |
+
"loss": 2.692,
|
| 452 |
+
"step": 7400
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.33,
|
| 456 |
+
"learning_rate": 4.33569530558016e-05,
|
| 457 |
+
"loss": 2.7134,
|
| 458 |
+
"step": 7500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.35,
|
| 462 |
+
"learning_rate": 4.326837909654562e-05,
|
| 463 |
+
"loss": 2.6883,
|
| 464 |
+
"step": 7600
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.36,
|
| 468 |
+
"learning_rate": 4.317980513728964e-05,
|
| 469 |
+
"loss": 2.7302,
|
| 470 |
+
"step": 7700
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 1.38,
|
| 474 |
+
"learning_rate": 4.309123117803366e-05,
|
| 475 |
+
"loss": 2.646,
|
| 476 |
+
"step": 7800
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.4,
|
| 480 |
+
"learning_rate": 4.300265721877768e-05,
|
| 481 |
+
"loss": 2.6761,
|
| 482 |
+
"step": 7900
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 1.42,
|
| 486 |
+
"learning_rate": 4.29140832595217e-05,
|
| 487 |
+
"loss": 2.7138,
|
| 488 |
+
"step": 8000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 1.43,
|
| 492 |
+
"learning_rate": 4.282550930026573e-05,
|
| 493 |
+
"loss": 2.6977,
|
| 494 |
+
"step": 8100
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.45,
|
| 498 |
+
"learning_rate": 4.273693534100974e-05,
|
| 499 |
+
"loss": 2.6718,
|
| 500 |
+
"step": 8200
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.47,
|
| 504 |
+
"learning_rate": 4.264836138175377e-05,
|
| 505 |
+
"loss": 2.6965,
|
| 506 |
+
"step": 8300
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.49,
|
| 510 |
+
"learning_rate": 4.255978742249779e-05,
|
| 511 |
+
"loss": 2.6604,
|
| 512 |
+
"step": 8400
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 1.51,
|
| 516 |
+
"learning_rate": 4.247121346324181e-05,
|
| 517 |
+
"loss": 2.6504,
|
| 518 |
+
"step": 8500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 1.52,
|
| 522 |
+
"learning_rate": 4.238263950398583e-05,
|
| 523 |
+
"loss": 2.659,
|
| 524 |
+
"step": 8600
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 1.54,
|
| 528 |
+
"learning_rate": 4.229406554472985e-05,
|
| 529 |
+
"loss": 2.6395,
|
| 530 |
+
"step": 8700
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 1.56,
|
| 534 |
+
"learning_rate": 4.220549158547387e-05,
|
| 535 |
+
"loss": 2.6709,
|
| 536 |
+
"step": 8800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.58,
|
| 540 |
+
"learning_rate": 4.211691762621789e-05,
|
| 541 |
+
"loss": 2.6267,
|
| 542 |
+
"step": 8900
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.59,
|
| 546 |
+
"learning_rate": 4.202834366696192e-05,
|
| 547 |
+
"loss": 2.631,
|
| 548 |
+
"step": 9000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.61,
|
| 552 |
+
"learning_rate": 4.193976970770593e-05,
|
| 553 |
+
"loss": 2.6071,
|
| 554 |
+
"step": 9100
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 1.63,
|
| 558 |
+
"learning_rate": 4.185119574844996e-05,
|
| 559 |
+
"loss": 2.6651,
|
| 560 |
+
"step": 9200
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.65,
|
| 564 |
+
"learning_rate": 4.1762621789193976e-05,
|
| 565 |
+
"loss": 2.6422,
|
| 566 |
+
"step": 9300
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 1.67,
|
| 570 |
+
"learning_rate": 4.1674047829938e-05,
|
| 571 |
+
"loss": 2.6112,
|
| 572 |
+
"step": 9400
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 1.68,
|
| 576 |
+
"learning_rate": 4.158547387068202e-05,
|
| 577 |
+
"loss": 2.6552,
|
| 578 |
+
"step": 9500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.7,
|
| 582 |
+
"learning_rate": 4.149689991142604e-05,
|
| 583 |
+
"loss": 2.6484,
|
| 584 |
+
"step": 9600
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.72,
|
| 588 |
+
"learning_rate": 4.140832595217006e-05,
|
| 589 |
+
"loss": 2.634,
|
| 590 |
+
"step": 9700
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.74,
|
| 594 |
+
"learning_rate": 4.131975199291409e-05,
|
| 595 |
+
"loss": 2.6383,
|
| 596 |
+
"step": 9800
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 1.75,
|
| 600 |
+
"learning_rate": 4.123117803365811e-05,
|
| 601 |
+
"loss": 2.5874,
|
| 602 |
+
"step": 9900
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 1.77,
|
| 606 |
+
"learning_rate": 4.114260407440213e-05,
|
| 607 |
+
"loss": 2.6141,
|
| 608 |
+
"step": 10000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 1.79,
|
| 612 |
+
"learning_rate": 4.1054030115146146e-05,
|
| 613 |
+
"loss": 2.6204,
|
| 614 |
+
"step": 10100
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 1.81,
|
| 618 |
+
"learning_rate": 4.096545615589017e-05,
|
| 619 |
+
"loss": 2.6344,
|
| 620 |
+
"step": 10200
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.82,
|
| 624 |
+
"learning_rate": 4.087688219663419e-05,
|
| 625 |
+
"loss": 2.6132,
|
| 626 |
+
"step": 10300
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.84,
|
| 630 |
+
"learning_rate": 4.078830823737822e-05,
|
| 631 |
+
"loss": 2.6309,
|
| 632 |
+
"step": 10400
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.86,
|
| 636 |
+
"learning_rate": 4.069973427812223e-05,
|
| 637 |
+
"loss": 2.6399,
|
| 638 |
+
"step": 10500
|
| 639 |
}
|
| 640 |
],
|
| 641 |
"max_steps": 56450,
|
| 642 |
"num_train_epochs": 10,
|
| 643 |
+
"total_flos": 3.901024976606024e+17,
|
| 644 |
"trial_name": null,
|
| 645 |
"trial_params": null
|
| 646 |
}
|
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:f57f6d86a29403946f27226ad8d197fb8ade690fd2784dc6db158755be06bf03
|
| 3 |
size 2607
|