jonabur
commited on
Commit
•
5da6497
1
Parent(s):
c9757cd
update chat version
Browse files- all_results.json +10 -10
- config.json +1 -1
- eval_results.json +5 -5
- model-00001-of-00014.safetensors +1 -1
- model-00002-of-00014.safetensors +1 -1
- model-00003-of-00014.safetensors +1 -1
- model-00004-of-00014.safetensors +1 -1
- model-00005-of-00014.safetensors +1 -1
- model-00006-of-00014.safetensors +1 -1
- model-00007-of-00014.safetensors +1 -1
- model-00008-of-00014.safetensors +1 -1
- model-00009-of-00014.safetensors +1 -1
- model-00010-of-00014.safetensors +1 -1
- model-00011-of-00014.safetensors +1 -1
- model-00012-of-00014.safetensors +1 -1
- model-00013-of-00014.safetensors +1 -1
- model-00014-of-00014.safetensors +1 -1
- train_results.json +5 -5
- trainer_state.json +1077 -1329
- training_args.bin +2 -2
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 3.0,
|
| 3 |
-
"eval_loss": 1.
|
| 4 |
-
"eval_runtime":
|
| 5 |
-
"eval_samples":
|
| 6 |
-
"eval_samples_per_second": 10.
|
| 7 |
-
"eval_steps_per_second": 0.
|
| 8 |
-
"train_loss": 0.
|
| 9 |
-
"train_runtime":
|
| 10 |
-
"train_samples":
|
| 11 |
-
"train_samples_per_second":
|
| 12 |
-
"train_steps_per_second": 0.
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 3.0,
|
| 3 |
+
"eval_loss": 1.7387598752975464,
|
| 4 |
+
"eval_runtime": 133.2357,
|
| 5 |
+
"eval_samples": 1363,
|
| 6 |
+
"eval_samples_per_second": 10.23,
|
| 7 |
+
"eval_steps_per_second": 0.113,
|
| 8 |
+
"train_loss": 0.872046828952715,
|
| 9 |
+
"train_runtime": 30104.3205,
|
| 10 |
+
"train_samples": 27928,
|
| 11 |
+
"train_samples_per_second": 2.783,
|
| 12 |
+
"train_steps_per_second": 0.058
|
| 13 |
}
|
config.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "/scratch/
|
| 3 |
"apply_residual_connection_post_layernorm": false,
|
| 4 |
"architectures": [
|
| 5 |
"BloomForCausalLM"
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "/scratch/project_462000444/zosaelai2/models/Poro-34b-Chat",
|
| 3 |
"apply_residual_connection_post_layernorm": false,
|
| 4 |
"architectures": [
|
| 5 |
"BloomForCausalLM"
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 3.0,
|
| 3 |
-
"eval_loss": 1.
|
| 4 |
-
"eval_runtime":
|
| 5 |
-
"eval_samples":
|
| 6 |
-
"eval_samples_per_second": 10.
|
| 7 |
-
"eval_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 3.0,
|
| 3 |
+
"eval_loss": 1.7387598752975464,
|
| 4 |
+
"eval_runtime": 133.2357,
|
| 5 |
+
"eval_samples": 1363,
|
| 6 |
+
"eval_samples_per_second": 10.23,
|
| 7 |
+
"eval_steps_per_second": 0.113
|
| 8 |
}
|
model-00001-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4712820784
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e26d4d99ffee422d8f22c4b8c684bb9fcd52910ae87d605f77589754c86c9754
|
| 3 |
size 4712820784
|
model-00002-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4933252680
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9cb05b88fdd0c8c2c7abba8ef4b99edcd3d52210b6e942cc27c61bbad5fb45a0
|
| 3 |
size 4933252680
|
model-00003-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4933252648
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:04b141df4d05f4f0019b3e038415c3caa0958db4f3c00724242c160920811916
|
| 3 |
size 4933252648
|
model-00004-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4933252728
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9639ee942617714ed0ec5e6036c4696906b7dfa2ddc20416b2f53fc6245d1213
|
| 3 |
size 4933252728
|
model-00005-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4933252728
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:504d621932697cee74218cdfd052cc76f62c469d1f5a644c089919be72b14ef6
|
| 3 |
size 4933252728
|
model-00006-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4933252728
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a9035eabc590f2497b2c413d1b382476e67b304d27312d2ddd1fd66ba4f46b5b
|
| 3 |
size 4933252728
|
model-00007-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4933252728
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2dc30447b516474700d4bb860155da2862907b1f0da94d72f776e7fb278e505d
|
| 3 |
size 4933252728
|
model-00008-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4933252728
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4ec50bc952bfe8c69d3e588fc5c85345e8aecd696d5032631e1eb356ed605997
|
| 3 |
size 4933252728
|
model-00009-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4933252728
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a48cae7f730aa02e5f746b0fafe1ac3efb36e7e6a05b7e4739c6ff6378421acf
|
| 3 |
size 4933252728
|
model-00010-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4933252728
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b97164bb97f8f4d89cc6ed3922dadb38e6b3436ec24197583a62befcdb679162
|
| 3 |
size 4933252728
|
model-00011-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4933252728
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:647d99c8c910a23392a09237dabb427ff62927e42d52e6896dd4f8e0585fb350
|
| 3 |
size 4933252728
|
model-00012-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4933252728
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8f8aba7688dcb24fbc6dc691a19d2ba0389edc76ca37be7c935d712f628467f3
|
| 3 |
size 4933252728
|
model-00013-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4933252728
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d3b7d4e3b8cef88db8863317b471398ee777ae289ffc20bb849bffeb3d53fb84
|
| 3 |
size 4933252728
|
model-00014-of-00014.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4522124144
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea783f2f5dc63b28b2768bb8dc1ea66ffd914f5b13b55fa95ed124025a408435
|
| 3 |
size 4522124144
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 3.0,
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 3.0,
|
| 3 |
+
"train_loss": 0.872046828952715,
|
| 4 |
+
"train_runtime": 30104.3205,
|
| 5 |
+
"train_samples": 27928,
|
| 6 |
+
"train_samples_per_second": 2.783,
|
| 7 |
+
"train_steps_per_second": 0.058
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,2404 +1,2152 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
-
"eval_steps":
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0,
|
| 13 |
-
"learning_rate": 1.
|
| 14 |
-
"loss": 2.
|
| 15 |
"step": 1
|
| 16 |
},
|
| 17 |
{
|
| 18 |
"epoch": 0.01,
|
| 19 |
-
"learning_rate": 5.
|
| 20 |
-
"loss":
|
| 21 |
"step": 5
|
| 22 |
},
|
| 23 |
{
|
| 24 |
"epoch": 0.02,
|
| 25 |
-
"learning_rate": 1.
|
| 26 |
-
"loss":
|
| 27 |
"step": 10
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch": 0.
|
| 31 |
-
"learning_rate": 1.
|
| 32 |
-
"loss": 2.
|
| 33 |
"step": 15
|
| 34 |
},
|
| 35 |
{
|
| 36 |
"epoch": 0.03,
|
| 37 |
-
"learning_rate": 2.
|
| 38 |
-
"loss": 2.
|
| 39 |
"step": 20
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.04,
|
| 43 |
-
"learning_rate": 2.
|
| 44 |
-
"loss":
|
| 45 |
"step": 25
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.05,
|
| 49 |
-
"learning_rate": 3.
|
| 50 |
-
"loss": 1.
|
| 51 |
"step": 30
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"learning_rate":
|
| 56 |
-
"loss": 1.
|
| 57 |
"step": 35
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch": 0.
|
| 61 |
-
"learning_rate": 4.
|
| 62 |
-
"loss": 1.
|
| 63 |
"step": 40
|
| 64 |
},
|
| 65 |
{
|
| 66 |
-
"epoch": 0.
|
| 67 |
-
"learning_rate":
|
| 68 |
-
"loss": 1.
|
| 69 |
"step": 45
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch": 0.
|
| 73 |
-
"learning_rate": 5.
|
| 74 |
-
"loss": 1.
|
| 75 |
"step": 50
|
| 76 |
},
|
| 77 |
{
|
| 78 |
"epoch": 0.09,
|
| 79 |
-
"learning_rate":
|
| 80 |
-
"loss": 1.
|
| 81 |
"step": 55
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch": 0.
|
| 85 |
-
"learning_rate": 6.
|
| 86 |
-
"loss": 1.
|
| 87 |
"step": 60
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch": 0.
|
| 91 |
-
"learning_rate":
|
| 92 |
-
"loss": 1.
|
| 93 |
"step": 65
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch": 0.
|
| 97 |
-
"learning_rate":
|
| 98 |
-
"loss": 1.
|
| 99 |
"step": 70
|
| 100 |
},
|
| 101 |
{
|
| 102 |
-
"epoch": 0.
|
| 103 |
-
"learning_rate":
|
| 104 |
-
"loss": 1.
|
| 105 |
"step": 75
|
| 106 |
},
|
| 107 |
{
|
| 108 |
-
"epoch": 0.
|
| 109 |
-
"learning_rate":
|
| 110 |
-
"loss": 1.
|
| 111 |
"step": 80
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch": 0.
|
| 115 |
-
"learning_rate":
|
| 116 |
-
"loss": 1.
|
| 117 |
"step": 85
|
| 118 |
},
|
| 119 |
{
|
| 120 |
-
"epoch": 0.
|
| 121 |
-
"learning_rate":
|
| 122 |
-
"loss": 1.
|
| 123 |
"step": 90
|
| 124 |
},
|
| 125 |
{
|
| 126 |
-
"epoch": 0.
|
| 127 |
-
"learning_rate":
|
| 128 |
-
"loss": 1.
|
| 129 |
"step": 95
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch": 0.
|
| 133 |
-
"learning_rate": 1.
|
| 134 |
-
"loss": 1.
|
| 135 |
"step": 100
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch": 0.
|
| 139 |
-
"learning_rate": 1.
|
| 140 |
-
"loss": 1.
|
| 141 |
"step": 105
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch": 0.
|
| 145 |
-
"learning_rate": 1.
|
| 146 |
-
"loss": 1.
|
| 147 |
"step": 110
|
| 148 |
},
|
| 149 |
{
|
| 150 |
-
"epoch": 0.
|
| 151 |
-
"learning_rate": 1.
|
| 152 |
-
"loss": 1.
|
| 153 |
"step": 115
|
| 154 |
},
|
| 155 |
{
|
| 156 |
-
"epoch": 0.
|
| 157 |
-
"learning_rate": 1.
|
| 158 |
-
"loss": 1.
|
| 159 |
"step": 120
|
| 160 |
},
|
| 161 |
{
|
| 162 |
-
"epoch": 0.
|
| 163 |
-
"learning_rate": 1.
|
| 164 |
-
"loss": 1.
|
| 165 |
"step": 125
|
| 166 |
},
|
| 167 |
{
|
| 168 |
-
"epoch": 0.
|
| 169 |
-
"learning_rate": 1.
|
| 170 |
-
"loss": 1.
|
| 171 |
"step": 130
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch": 0.
|
| 175 |
-
"learning_rate": 1.
|
| 176 |
-
"loss": 1.
|
| 177 |
"step": 135
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch": 0.
|
| 181 |
-
"learning_rate": 1.
|
| 182 |
-
"loss": 1.
|
| 183 |
"step": 140
|
| 184 |
},
|
| 185 |
{
|
| 186 |
-
"epoch": 0.
|
| 187 |
-
"learning_rate": 1.
|
| 188 |
-
"loss": 1.
|
| 189 |
"step": 145
|
| 190 |
},
|
| 191 |
{
|
| 192 |
-
"epoch": 0.
|
| 193 |
-
"learning_rate": 1.
|
| 194 |
-
"loss": 1.
|
| 195 |
"step": 150
|
| 196 |
},
|
| 197 |
{
|
| 198 |
-
"epoch": 0.
|
| 199 |
-
"learning_rate": 1.
|
| 200 |
-
"loss": 1.
|
| 201 |
"step": 155
|
| 202 |
},
|
| 203 |
{
|
| 204 |
-
"epoch": 0.
|
| 205 |
-
"learning_rate": 1.
|
| 206 |
-
"loss": 1.
|
| 207 |
"step": 160
|
| 208 |
},
|
| 209 |
{
|
| 210 |
-
"epoch": 0.
|
| 211 |
-
"learning_rate": 1.
|
| 212 |
-
"loss": 1.
|
| 213 |
"step": 165
|
| 214 |
},
|
| 215 |
{
|
| 216 |
-
"epoch": 0.
|
| 217 |
-
"learning_rate": 1.
|
| 218 |
-
"loss": 1.
|
| 219 |
"step": 170
|
| 220 |
},
|
| 221 |
{
|
| 222 |
-
"epoch": 0.
|
| 223 |
-
"learning_rate":
|
| 224 |
-
"loss": 1.
|
| 225 |
"step": 175
|
| 226 |
},
|
| 227 |
{
|
| 228 |
-
"epoch": 0.
|
| 229 |
-
"learning_rate": 1.
|
| 230 |
-
"loss": 1.
|
| 231 |
"step": 180
|
| 232 |
},
|
| 233 |
{
|
| 234 |
-
"epoch": 0.
|
| 235 |
-
"learning_rate": 1.
|
| 236 |
-
"loss": 1.
|
| 237 |
"step": 185
|
| 238 |
},
|
| 239 |
{
|
| 240 |
-
"epoch": 0.
|
| 241 |
-
"learning_rate": 1.
|
| 242 |
-
"loss": 1.
|
| 243 |
"step": 190
|
| 244 |
},
|
| 245 |
{
|
| 246 |
-
"epoch": 0.
|
| 247 |
-
"learning_rate": 1.
|
| 248 |
-
"loss": 1.
|
| 249 |
"step": 195
|
| 250 |
},
|
| 251 |
{
|
| 252 |
-
"epoch": 0.
|
| 253 |
-
"learning_rate": 1.
|
| 254 |
-
"loss": 1.
|
| 255 |
-
"step": 200
|
| 256 |
-
},
|
| 257 |
-
{
|
| 258 |
-
"epoch": 0.31,
|
| 259 |
-
"eval_loss": 1.5671130418777466,
|
| 260 |
-
"eval_runtime": 157.1346,
|
| 261 |
-
"eval_samples_per_second": 10.691,
|
| 262 |
-
"eval_steps_per_second": 0.115,
|
| 263 |
"step": 200
|
| 264 |
},
|
| 265 |
{
|
| 266 |
-
"epoch": 0.
|
| 267 |
-
"learning_rate": 1.
|
| 268 |
-
"loss": 1.
|
| 269 |
"step": 205
|
| 270 |
},
|
| 271 |
{
|
| 272 |
-
"epoch": 0.
|
| 273 |
-
"learning_rate": 1.
|
| 274 |
-
"loss": 1.
|
| 275 |
"step": 210
|
| 276 |
},
|
| 277 |
{
|
| 278 |
-
"epoch": 0.
|
| 279 |
-
"learning_rate": 1.
|
| 280 |
-
"loss": 1.
|
| 281 |
"step": 215
|
| 282 |
},
|
| 283 |
{
|
| 284 |
-
"epoch": 0.
|
| 285 |
-
"learning_rate": 1.
|
| 286 |
-
"loss": 1.
|
| 287 |
"step": 220
|
| 288 |
},
|
| 289 |
{
|
| 290 |
-
"epoch": 0.
|
| 291 |
-
"learning_rate": 1.
|
| 292 |
-
"loss": 1.
|
| 293 |
"step": 225
|
| 294 |
},
|
| 295 |
{
|
| 296 |
-
"epoch": 0.
|
| 297 |
-
"learning_rate": 1.
|
| 298 |
-
"loss": 1.
|
| 299 |
"step": 230
|
| 300 |
},
|
| 301 |
{
|
| 302 |
-
"epoch": 0.
|
| 303 |
-
"learning_rate": 1.
|
| 304 |
-
"loss": 1.
|
| 305 |
"step": 235
|
| 306 |
},
|
| 307 |
{
|
| 308 |
-
"epoch": 0.
|
| 309 |
-
"learning_rate": 1.
|
| 310 |
-
"loss": 1.
|
| 311 |
"step": 240
|
| 312 |
},
|
| 313 |
{
|
| 314 |
-
"epoch": 0.
|
| 315 |
-
"learning_rate": 1.
|
| 316 |
-
"loss": 1.
|
| 317 |
"step": 245
|
| 318 |
},
|
| 319 |
{
|
| 320 |
-
"epoch": 0.
|
| 321 |
-
"learning_rate": 1.
|
| 322 |
-
"loss": 1.
|
| 323 |
"step": 250
|
| 324 |
},
|
| 325 |
{
|
| 326 |
-
"epoch": 0.
|
| 327 |
-
"learning_rate": 1.
|
| 328 |
-
"loss": 1.
|
| 329 |
"step": 255
|
| 330 |
},
|
| 331 |
{
|
| 332 |
-
"epoch": 0.
|
| 333 |
-
"learning_rate": 1.
|
| 334 |
-
"loss": 1.
|
| 335 |
"step": 260
|
| 336 |
},
|
| 337 |
{
|
| 338 |
-
"epoch": 0.
|
| 339 |
-
"learning_rate": 1.
|
| 340 |
-
"loss": 1.
|
| 341 |
"step": 265
|
| 342 |
},
|
| 343 |
{
|
| 344 |
-
"epoch": 0.
|
| 345 |
-
"learning_rate": 1.
|
| 346 |
-
"loss": 1.
|
| 347 |
"step": 270
|
| 348 |
},
|
| 349 |
{
|
| 350 |
-
"epoch": 0.
|
| 351 |
-
"learning_rate": 1.
|
| 352 |
-
"loss": 1.
|
| 353 |
"step": 275
|
| 354 |
},
|
| 355 |
{
|
| 356 |
-
"epoch": 0.
|
| 357 |
-
"learning_rate": 1.
|
| 358 |
-
"loss": 1.
|
| 359 |
"step": 280
|
| 360 |
},
|
| 361 |
{
|
| 362 |
-
"epoch": 0.
|
| 363 |
-
"learning_rate": 1.
|
| 364 |
-
"loss": 1.
|
| 365 |
"step": 285
|
| 366 |
},
|
| 367 |
{
|
| 368 |
-
"epoch": 0.
|
| 369 |
-
"learning_rate": 1.
|
| 370 |
-
"loss": 1.
|
| 371 |
"step": 290
|
| 372 |
},
|
| 373 |
{
|
| 374 |
-
"epoch": 0.
|
| 375 |
-
"learning_rate": 1.
|
| 376 |
-
"loss": 1.
|
| 377 |
"step": 295
|
| 378 |
},
|
| 379 |
{
|
| 380 |
-
"epoch": 0.
|
| 381 |
-
"learning_rate": 1.
|
| 382 |
-
"loss": 1.
|
| 383 |
"step": 300
|
| 384 |
},
|
| 385 |
{
|
| 386 |
-
"epoch": 0.
|
| 387 |
-
"learning_rate": 1.
|
| 388 |
-
"loss": 1.
|
| 389 |
"step": 305
|
| 390 |
},
|
| 391 |
{
|
| 392 |
-
"epoch": 0.
|
| 393 |
-
"learning_rate": 1.
|
| 394 |
-
"loss": 1.
|
| 395 |
"step": 310
|
| 396 |
},
|
| 397 |
{
|
| 398 |
-
"epoch": 0.
|
| 399 |
-
"learning_rate": 1.
|
| 400 |
-
"loss": 1.
|
| 401 |
"step": 315
|
| 402 |
},
|
| 403 |
{
|
| 404 |
-
"epoch": 0.
|
| 405 |
-
"learning_rate": 1.
|
| 406 |
-
"loss": 1.
|
| 407 |
"step": 320
|
| 408 |
},
|
| 409 |
{
|
| 410 |
-
"epoch": 0.
|
| 411 |
-
"learning_rate": 1.
|
| 412 |
-
"loss": 1.
|
| 413 |
"step": 325
|
| 414 |
},
|
| 415 |
{
|
| 416 |
-
"epoch": 0.
|
| 417 |
-
"learning_rate": 1.
|
| 418 |
-
"loss": 1.
|
| 419 |
"step": 330
|
| 420 |
},
|
| 421 |
{
|
| 422 |
-
"epoch": 0.
|
| 423 |
-
"learning_rate": 1.
|
| 424 |
-
"loss": 1.
|
| 425 |
"step": 335
|
| 426 |
},
|
| 427 |
{
|
| 428 |
-
"epoch": 0.
|
| 429 |
-
"learning_rate": 1.
|
| 430 |
-
"loss": 1.
|
| 431 |
"step": 340
|
| 432 |
},
|
| 433 |
{
|
| 434 |
-
"epoch": 0.
|
| 435 |
-
"learning_rate": 1.
|
| 436 |
-
"loss": 1.
|
| 437 |
"step": 345
|
| 438 |
},
|
| 439 |
{
|
| 440 |
-
"epoch": 0.
|
| 441 |
-
"learning_rate": 1.
|
| 442 |
-
"loss": 1.
|
| 443 |
"step": 350
|
| 444 |
},
|
| 445 |
{
|
| 446 |
-
"epoch": 0.
|
| 447 |
-
"learning_rate": 1.
|
| 448 |
-
"loss": 1.
|
| 449 |
"step": 355
|
| 450 |
},
|
| 451 |
{
|
| 452 |
-
"epoch": 0.
|
| 453 |
-
"learning_rate": 1.
|
| 454 |
-
"loss": 1.
|
| 455 |
"step": 360
|
| 456 |
},
|
| 457 |
{
|
| 458 |
-
"epoch": 0.
|
| 459 |
-
"learning_rate": 1.
|
| 460 |
-
"loss": 1.
|
| 461 |
"step": 365
|
| 462 |
},
|
| 463 |
{
|
| 464 |
-
"epoch": 0.
|
| 465 |
-
"learning_rate": 1.
|
| 466 |
-
"loss": 1.
|
| 467 |
"step": 370
|
| 468 |
},
|
| 469 |
{
|
| 470 |
-
"epoch": 0.
|
| 471 |
-
"learning_rate": 1.
|
| 472 |
-
"loss": 1.
|
| 473 |
"step": 375
|
| 474 |
},
|
| 475 |
{
|
| 476 |
-
"epoch": 0.
|
| 477 |
-
"learning_rate": 1.
|
| 478 |
-
"loss": 1.
|
| 479 |
"step": 380
|
| 480 |
},
|
| 481 |
{
|
| 482 |
-
"epoch": 0.
|
| 483 |
-
"learning_rate": 1.
|
| 484 |
-
"loss": 1.
|
| 485 |
"step": 385
|
| 486 |
},
|
| 487 |
{
|
| 488 |
-
"epoch": 0.
|
| 489 |
-
"learning_rate": 1.
|
| 490 |
-
"loss": 1.
|
| 491 |
"step": 390
|
| 492 |
},
|
| 493 |
{
|
| 494 |
-
"epoch": 0.
|
| 495 |
-
"learning_rate": 1.
|
| 496 |
-
"loss": 1.
|
| 497 |
"step": 395
|
| 498 |
},
|
| 499 |
{
|
| 500 |
-
"epoch": 0.
|
| 501 |
-
"learning_rate": 1.
|
| 502 |
-
"loss": 1.
|
| 503 |
-
"step": 400
|
| 504 |
-
},
|
| 505 |
-
{
|
| 506 |
-
"epoch": 0.63,
|
| 507 |
-
"eval_loss": 1.551980972290039,
|
| 508 |
-
"eval_runtime": 156.7574,
|
| 509 |
-
"eval_samples_per_second": 10.717,
|
| 510 |
-
"eval_steps_per_second": 0.115,
|
| 511 |
"step": 400
|
| 512 |
},
|
| 513 |
{
|
| 514 |
-
"epoch": 0.
|
| 515 |
-
"learning_rate": 1.
|
| 516 |
-
"loss": 1.
|
| 517 |
"step": 405
|
| 518 |
},
|
| 519 |
{
|
| 520 |
-
"epoch": 0.
|
| 521 |
-
"learning_rate": 1.
|
| 522 |
-
"loss": 1.
|
| 523 |
"step": 410
|
| 524 |
},
|
| 525 |
{
|
| 526 |
-
"epoch": 0.
|
| 527 |
-
"learning_rate": 1.
|
| 528 |
-
"loss": 1.
|
| 529 |
"step": 415
|
| 530 |
},
|
| 531 |
{
|
| 532 |
-
"epoch": 0.
|
| 533 |
-
"learning_rate": 1.
|
| 534 |
-
"loss": 1.
|
| 535 |
"step": 420
|
| 536 |
},
|
| 537 |
{
|
| 538 |
-
"epoch": 0.
|
| 539 |
-
"learning_rate": 1.
|
| 540 |
-
"loss": 1.
|
| 541 |
"step": 425
|
| 542 |
},
|
| 543 |
{
|
| 544 |
-
"epoch": 0.
|
| 545 |
-
"learning_rate": 1.
|
| 546 |
-
"loss": 1.
|
| 547 |
"step": 430
|
| 548 |
},
|
| 549 |
{
|
| 550 |
-
"epoch": 0.
|
| 551 |
-
"learning_rate": 1.
|
| 552 |
-
"loss": 1.
|
| 553 |
"step": 435
|
| 554 |
},
|
| 555 |
{
|
| 556 |
-
"epoch": 0.
|
| 557 |
-
"learning_rate": 1.
|
| 558 |
-
"loss": 1.
|
| 559 |
"step": 440
|
| 560 |
},
|
| 561 |
{
|
| 562 |
-
"epoch": 0.
|
| 563 |
-
"learning_rate": 1.
|
| 564 |
-
"loss": 1.
|
| 565 |
"step": 445
|
| 566 |
},
|
| 567 |
{
|
| 568 |
-
"epoch": 0.
|
| 569 |
-
"learning_rate": 1.
|
| 570 |
-
"loss": 1.
|
| 571 |
"step": 450
|
| 572 |
},
|
| 573 |
{
|
| 574 |
-
"epoch": 0.
|
| 575 |
-
"learning_rate": 1.
|
| 576 |
-
"loss": 1.
|
| 577 |
"step": 455
|
| 578 |
},
|
| 579 |
{
|
| 580 |
-
"epoch": 0.
|
| 581 |
-
"learning_rate": 1.
|
| 582 |
-
"loss": 1.
|
| 583 |
"step": 460
|
| 584 |
},
|
| 585 |
{
|
| 586 |
-
"epoch": 0.
|
| 587 |
-
"learning_rate": 1.
|
| 588 |
-
"loss": 1.
|
| 589 |
"step": 465
|
| 590 |
},
|
| 591 |
{
|
| 592 |
-
"epoch": 0.
|
| 593 |
-
"learning_rate": 1.
|
| 594 |
-
"loss": 1.
|
| 595 |
"step": 470
|
| 596 |
},
|
| 597 |
{
|
| 598 |
-
"epoch": 0.
|
| 599 |
-
"learning_rate": 1.
|
| 600 |
-
"loss": 1.
|
| 601 |
"step": 475
|
| 602 |
},
|
| 603 |
{
|
| 604 |
-
"epoch": 0.
|
| 605 |
-
"learning_rate": 1.
|
| 606 |
-
"loss": 1.
|
| 607 |
"step": 480
|
| 608 |
},
|
| 609 |
{
|
| 610 |
-
"epoch": 0.
|
| 611 |
-
"learning_rate": 1.
|
| 612 |
-
"loss": 1.
|
| 613 |
"step": 485
|
| 614 |
},
|
| 615 |
{
|
| 616 |
-
"epoch": 0.
|
| 617 |
-
"learning_rate": 1.
|
| 618 |
-
"loss": 1.
|
| 619 |
"step": 490
|
| 620 |
},
|
| 621 |
{
|
| 622 |
-
"epoch": 0.
|
| 623 |
-
"learning_rate": 1.
|
| 624 |
-
"loss": 1.
|
| 625 |
"step": 495
|
| 626 |
},
|
| 627 |
{
|
| 628 |
-
"epoch": 0.
|
| 629 |
-
"learning_rate": 1.
|
| 630 |
-
"loss": 1.
|
| 631 |
"step": 500
|
| 632 |
},
|
| 633 |
{
|
| 634 |
-
"epoch": 0.
|
| 635 |
-
"learning_rate": 1.
|
| 636 |
-
"loss": 1.
|
| 637 |
"step": 505
|
| 638 |
},
|
| 639 |
{
|
| 640 |
-
"epoch": 0.
|
| 641 |
-
"learning_rate": 1.
|
| 642 |
-
"loss": 1.
|
| 643 |
"step": 510
|
| 644 |
},
|
| 645 |
{
|
| 646 |
-
"epoch": 0.
|
| 647 |
-
"learning_rate": 1.
|
| 648 |
-
"loss": 1.
|
| 649 |
"step": 515
|
| 650 |
},
|
| 651 |
{
|
| 652 |
-
"epoch": 0.
|
| 653 |
-
"learning_rate": 1.
|
| 654 |
-
"loss": 1.
|
| 655 |
"step": 520
|
| 656 |
},
|
| 657 |
{
|
| 658 |
-
"epoch": 0.
|
| 659 |
-
"learning_rate": 1.
|
| 660 |
-
"loss": 1.
|
| 661 |
"step": 525
|
| 662 |
},
|
| 663 |
{
|
| 664 |
-
"epoch": 0.
|
| 665 |
-
"learning_rate": 1.
|
| 666 |
-
"loss": 1.
|
| 667 |
"step": 530
|
| 668 |
},
|
| 669 |
{
|
| 670 |
-
"epoch": 0.
|
| 671 |
-
"learning_rate": 1.
|
| 672 |
-
"loss": 1.
|
| 673 |
"step": 535
|
| 674 |
},
|
| 675 |
{
|
| 676 |
-
"epoch": 0.
|
| 677 |
-
"learning_rate": 1.
|
| 678 |
-
"loss": 1.
|
| 679 |
"step": 540
|
| 680 |
},
|
| 681 |
{
|
| 682 |
-
"epoch": 0.
|
| 683 |
-
"learning_rate": 1.
|
| 684 |
-
"loss": 1.
|
| 685 |
"step": 545
|
| 686 |
},
|
| 687 |
{
|
| 688 |
-
"epoch": 0.
|
| 689 |
-
"learning_rate": 1.
|
| 690 |
-
"loss": 1.
|
| 691 |
"step": 550
|
| 692 |
},
|
| 693 |
{
|
| 694 |
-
"epoch": 0.
|
| 695 |
-
"learning_rate": 1.
|
| 696 |
-
"loss": 1.
|
| 697 |
"step": 555
|
| 698 |
},
|
| 699 |
{
|
| 700 |
-
"epoch": 0.
|
| 701 |
-
"learning_rate": 1.
|
| 702 |
-
"loss": 1.
|
| 703 |
"step": 560
|
| 704 |
},
|
| 705 |
{
|
| 706 |
-
"epoch": 0.
|
| 707 |
-
"learning_rate": 1.
|
| 708 |
-
"loss": 1.
|
| 709 |
"step": 565
|
| 710 |
},
|
| 711 |
{
|
| 712 |
-
"epoch": 0.
|
| 713 |
-
"learning_rate": 1.
|
| 714 |
-
"loss": 1.
|
| 715 |
"step": 570
|
| 716 |
},
|
| 717 |
{
|
| 718 |
-
"epoch": 0.
|
| 719 |
-
"learning_rate": 1.
|
| 720 |
-
"loss": 1.
|
| 721 |
"step": 575
|
| 722 |
},
|
| 723 |
{
|
| 724 |
-
"epoch": 0
|
| 725 |
-
"learning_rate": 1.
|
| 726 |
-
"loss": 1.
|
| 727 |
"step": 580
|
| 728 |
},
|
| 729 |
{
|
| 730 |
-
"epoch": 0
|
| 731 |
-
"
|
| 732 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 733 |
"step": 585
|
| 734 |
},
|
| 735 |
{
|
| 736 |
-
"epoch":
|
| 737 |
-
"learning_rate": 1.
|
| 738 |
-
"loss": 1.
|
| 739 |
"step": 590
|
| 740 |
},
|
| 741 |
{
|
| 742 |
-
"epoch":
|
| 743 |
-
"learning_rate": 1.
|
| 744 |
-
"loss": 1.
|
| 745 |
"step": 595
|
| 746 |
},
|
| 747 |
{
|
| 748 |
-
"epoch":
|
| 749 |
-
"learning_rate": 1.
|
| 750 |
-
"loss": 1.
|
| 751 |
-
"step": 600
|
| 752 |
-
},
|
| 753 |
-
{
|
| 754 |
-
"epoch": 0.94,
|
| 755 |
-
"eval_loss": 1.5388672351837158,
|
| 756 |
-
"eval_runtime": 156.7467,
|
| 757 |
-
"eval_samples_per_second": 10.718,
|
| 758 |
-
"eval_steps_per_second": 0.115,
|
| 759 |
"step": 600
|
| 760 |
},
|
| 761 |
{
|
| 762 |
-
"epoch":
|
| 763 |
-
"learning_rate": 1.
|
| 764 |
-
"loss": 1.
|
| 765 |
"step": 605
|
| 766 |
},
|
| 767 |
{
|
| 768 |
-
"epoch":
|
| 769 |
-
"learning_rate": 1.
|
| 770 |
-
"loss": 1.
|
| 771 |
"step": 610
|
| 772 |
},
|
| 773 |
{
|
| 774 |
-
"epoch":
|
| 775 |
-
"learning_rate": 1.
|
| 776 |
-
"loss": 1.
|
| 777 |
"step": 615
|
| 778 |
},
|
| 779 |
{
|
| 780 |
-
"epoch":
|
| 781 |
-
"learning_rate": 1.
|
| 782 |
-
"loss": 1.
|
| 783 |
"step": 620
|
| 784 |
},
|
| 785 |
{
|
| 786 |
-
"epoch":
|
| 787 |
-
"learning_rate": 1.
|
| 788 |
-
"loss": 1.
|
| 789 |
"step": 625
|
| 790 |
},
|
| 791 |
{
|
| 792 |
-
"epoch":
|
| 793 |
-
"learning_rate": 1.
|
| 794 |
-
"loss": 1.
|
| 795 |
"step": 630
|
| 796 |
},
|
| 797 |
{
|
| 798 |
-
"epoch":
|
| 799 |
-
"learning_rate": 1.
|
| 800 |
-
"loss": 1.
|
| 801 |
"step": 635
|
| 802 |
},
|
| 803 |
{
|
| 804 |
-
"epoch": 1.
|
| 805 |
-
"learning_rate": 1.
|
| 806 |
-
"loss": 1.
|
| 807 |
"step": 640
|
| 808 |
},
|
| 809 |
{
|
| 810 |
-
"epoch": 1.
|
| 811 |
-
"learning_rate": 1.
|
| 812 |
-
"loss": 1.
|
| 813 |
"step": 645
|
| 814 |
},
|
| 815 |
{
|
| 816 |
-
"epoch": 1.
|
| 817 |
-
"learning_rate": 1.
|
| 818 |
-
"loss": 1.
|
| 819 |
"step": 650
|
| 820 |
},
|
| 821 |
{
|
| 822 |
-
"epoch": 1.
|
| 823 |
-
"learning_rate": 1.
|
| 824 |
-
"loss":
|
| 825 |
"step": 655
|
| 826 |
},
|
| 827 |
{
|
| 828 |
-
"epoch": 1.
|
| 829 |
-
"learning_rate": 1.
|
| 830 |
-
"loss": 1.
|
| 831 |
"step": 660
|
| 832 |
},
|
| 833 |
{
|
| 834 |
-
"epoch": 1.
|
| 835 |
-
"learning_rate": 1.
|
| 836 |
-
"loss":
|
| 837 |
"step": 665
|
| 838 |
},
|
| 839 |
{
|
| 840 |
-
"epoch": 1.
|
| 841 |
-
"learning_rate": 1.
|
| 842 |
-
"loss":
|
| 843 |
"step": 670
|
| 844 |
},
|
| 845 |
{
|
| 846 |
-
"epoch": 1.
|
| 847 |
-
"learning_rate": 1.
|
| 848 |
-
"loss":
|
| 849 |
"step": 675
|
| 850 |
},
|
| 851 |
{
|
| 852 |
-
"epoch": 1.
|
| 853 |
-
"learning_rate": 1.
|
| 854 |
-
"loss":
|
| 855 |
"step": 680
|
| 856 |
},
|
| 857 |
{
|
| 858 |
-
"epoch": 1.
|
| 859 |
-
"learning_rate": 1.
|
| 860 |
-
"loss":
|
| 861 |
"step": 685
|
| 862 |
},
|
| 863 |
{
|
| 864 |
-
"epoch": 1.
|
| 865 |
-
"learning_rate": 1.
|
| 866 |
-
"loss":
|
| 867 |
"step": 690
|
| 868 |
},
|
| 869 |
{
|
| 870 |
-
"epoch": 1.
|
| 871 |
-
"learning_rate": 1.
|
| 872 |
-
"loss":
|
| 873 |
"step": 695
|
| 874 |
},
|
| 875 |
{
|
| 876 |
-
"epoch": 1.
|
| 877 |
-
"learning_rate": 1.
|
| 878 |
-
"loss":
|
| 879 |
"step": 700
|
| 880 |
},
|
| 881 |
{
|
| 882 |
-
"epoch": 1.
|
| 883 |
-
"learning_rate": 1.
|
| 884 |
-
"loss": 0.
|
| 885 |
"step": 705
|
| 886 |
},
|
| 887 |
{
|
| 888 |
-
"epoch": 1.
|
| 889 |
-
"learning_rate": 1.
|
| 890 |
-
"loss": 0.
|
| 891 |
"step": 710
|
| 892 |
},
|
| 893 |
{
|
| 894 |
-
"epoch": 1.
|
| 895 |
-
"learning_rate": 1.
|
| 896 |
-
"loss":
|
| 897 |
"step": 715
|
| 898 |
},
|
| 899 |
{
|
| 900 |
-
"epoch": 1.
|
| 901 |
-
"learning_rate": 1.
|
| 902 |
-
"loss": 0.
|
| 903 |
"step": 720
|
| 904 |
},
|
| 905 |
{
|
| 906 |
-
"epoch": 1.
|
| 907 |
-
"learning_rate": 1.
|
| 908 |
-
"loss": 0.
|
| 909 |
"step": 725
|
| 910 |
},
|
| 911 |
{
|
| 912 |
-
"epoch": 1.
|
| 913 |
-
"learning_rate": 1.
|
| 914 |
-
"loss": 0.
|
| 915 |
"step": 730
|
| 916 |
},
|
| 917 |
{
|
| 918 |
-
"epoch": 1.
|
| 919 |
-
"learning_rate": 1.
|
| 920 |
-
"loss": 0.
|
| 921 |
"step": 735
|
| 922 |
},
|
| 923 |
{
|
| 924 |
-
"epoch": 1.
|
| 925 |
-
"learning_rate": 1.
|
| 926 |
-
"loss": 0.
|
| 927 |
"step": 740
|
| 928 |
},
|
| 929 |
{
|
| 930 |
-
"epoch": 1.
|
| 931 |
-
"learning_rate": 1.
|
| 932 |
-
"loss": 0.
|
| 933 |
"step": 745
|
| 934 |
},
|
| 935 |
{
|
| 936 |
-
"epoch": 1.
|
| 937 |
-
"learning_rate": 1.
|
| 938 |
-
"loss": 0.
|
| 939 |
"step": 750
|
| 940 |
},
|
| 941 |
{
|
| 942 |
-
"epoch": 1.
|
| 943 |
-
"learning_rate": 1.
|
| 944 |
-
"loss": 0.
|
| 945 |
"step": 755
|
| 946 |
},
|
| 947 |
{
|
| 948 |
-
"epoch": 1.
|
| 949 |
-
"learning_rate": 1.
|
| 950 |
-
"loss": 0.
|
| 951 |
"step": 760
|
| 952 |
},
|
| 953 |
{
|
| 954 |
-
"epoch": 1.
|
| 955 |
-
"learning_rate": 1.
|
| 956 |
-
"loss": 0.
|
| 957 |
"step": 765
|
| 958 |
},
|
| 959 |
{
|
| 960 |
-
"epoch": 1.
|
| 961 |
-
"learning_rate": 1.
|
| 962 |
-
"loss": 0.
|
| 963 |
"step": 770
|
| 964 |
},
|
| 965 |
{
|
| 966 |
-
"epoch": 1.
|
| 967 |
-
"learning_rate": 1.
|
| 968 |
-
"loss": 0.
|
| 969 |
"step": 775
|
| 970 |
},
|
| 971 |
{
|
| 972 |
-
"epoch": 1.
|
| 973 |
-
"learning_rate": 1.
|
| 974 |
-
"loss": 0.
|
| 975 |
"step": 780
|
| 976 |
},
|
| 977 |
{
|
| 978 |
-
"epoch": 1.
|
| 979 |
-
"learning_rate": 1.
|
| 980 |
-
"loss": 0.
|
| 981 |
"step": 785
|
| 982 |
},
|
| 983 |
{
|
| 984 |
-
"epoch": 1.
|
| 985 |
-
"learning_rate": 1.
|
| 986 |
-
"loss": 0.
|
| 987 |
"step": 790
|
| 988 |
},
|
| 989 |
{
|
| 990 |
-
"epoch": 1.
|
| 991 |
-
"learning_rate": 1.
|
| 992 |
-
"loss": 0.
|
| 993 |
"step": 795
|
| 994 |
},
|
| 995 |
{
|
| 996 |
-
"epoch": 1.
|
| 997 |
-
"learning_rate": 1.
|
| 998 |
-
"loss": 0.
|
| 999 |
-
"step": 800
|
| 1000 |
-
},
|
| 1001 |
-
{
|
| 1002 |
-
"epoch": 1.25,
|
| 1003 |
-
"eval_loss": 1.6069754362106323,
|
| 1004 |
-
"eval_runtime": 156.7669,
|
| 1005 |
-
"eval_samples_per_second": 10.717,
|
| 1006 |
-
"eval_steps_per_second": 0.115,
|
| 1007 |
"step": 800
|
| 1008 |
},
|
| 1009 |
{
|
| 1010 |
-
"epoch": 1.
|
| 1011 |
-
"learning_rate": 1.
|
| 1012 |
-
"loss": 0.
|
| 1013 |
"step": 805
|
| 1014 |
},
|
| 1015 |
{
|
| 1016 |
-
"epoch": 1.
|
| 1017 |
-
"learning_rate": 1.
|
| 1018 |
-
"loss": 0.
|
| 1019 |
"step": 810
|
| 1020 |
},
|
| 1021 |
{
|
| 1022 |
-
"epoch": 1.
|
| 1023 |
-
"learning_rate": 1.
|
| 1024 |
-
"loss": 0.
|
| 1025 |
"step": 815
|
| 1026 |
},
|
| 1027 |
{
|
| 1028 |
-
"epoch": 1.
|
| 1029 |
-
"learning_rate": 1.
|
| 1030 |
-
"loss": 0.
|
| 1031 |
"step": 820
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
-
"epoch": 1.
|
| 1035 |
-
"learning_rate": 1.
|
| 1036 |
-
"loss": 0.
|
| 1037 |
"step": 825
|
| 1038 |
},
|
| 1039 |
{
|
| 1040 |
-
"epoch": 1.
|
| 1041 |
-
"learning_rate": 1.
|
| 1042 |
-
"loss": 0.
|
| 1043 |
"step": 830
|
| 1044 |
},
|
| 1045 |
{
|
| 1046 |
-
"epoch": 1.
|
| 1047 |
-
"learning_rate": 1.
|
| 1048 |
-
"loss": 0.
|
| 1049 |
"step": 835
|
| 1050 |
},
|
| 1051 |
{
|
| 1052 |
-
"epoch": 1.
|
| 1053 |
-
"learning_rate": 1.
|
| 1054 |
-
"loss": 0.
|
| 1055 |
"step": 840
|
| 1056 |
},
|
| 1057 |
{
|
| 1058 |
-
"epoch": 1.
|
| 1059 |
-
"learning_rate": 1.
|
| 1060 |
-
"loss": 0.
|
| 1061 |
"step": 845
|
| 1062 |
},
|
| 1063 |
{
|
| 1064 |
-
"epoch": 1.
|
| 1065 |
-
"learning_rate": 1.
|
| 1066 |
-
"loss": 0.
|
| 1067 |
"step": 850
|
| 1068 |
},
|
| 1069 |
{
|
| 1070 |
-
"epoch": 1.
|
| 1071 |
-
"learning_rate": 1.
|
| 1072 |
-
"loss": 0.
|
| 1073 |
"step": 855
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
-
"epoch": 1.
|
| 1077 |
-
"learning_rate": 1.
|
| 1078 |
-
"loss": 0.
|
| 1079 |
"step": 860
|
| 1080 |
},
|
| 1081 |
{
|
| 1082 |
-
"epoch": 1.
|
| 1083 |
-
"learning_rate": 1.
|
| 1084 |
-
"loss": 0.
|
| 1085 |
"step": 865
|
| 1086 |
},
|
| 1087 |
{
|
| 1088 |
-
"epoch": 1.
|
| 1089 |
-
"learning_rate": 1.
|
| 1090 |
-
"loss": 0.
|
| 1091 |
"step": 870
|
| 1092 |
},
|
| 1093 |
{
|
| 1094 |
-
"epoch": 1.
|
| 1095 |
-
"learning_rate": 1.
|
| 1096 |
-
"loss": 0.
|
| 1097 |
"step": 875
|
| 1098 |
},
|
| 1099 |
{
|
| 1100 |
-
"epoch": 1.
|
| 1101 |
-
"learning_rate": 1.
|
| 1102 |
-
"loss": 0.
|
| 1103 |
"step": 880
|
| 1104 |
},
|
| 1105 |
{
|
| 1106 |
-
"epoch": 1.
|
| 1107 |
-
"learning_rate": 1.
|
| 1108 |
-
"loss": 0.
|
| 1109 |
"step": 885
|
| 1110 |
},
|
| 1111 |
{
|
| 1112 |
-
"epoch": 1.
|
| 1113 |
-
"learning_rate": 1.
|
| 1114 |
-
"loss": 0.
|
| 1115 |
"step": 890
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
-
"epoch": 1.
|
| 1119 |
-
"learning_rate": 1.
|
| 1120 |
-
"loss": 0.
|
| 1121 |
"step": 895
|
| 1122 |
},
|
| 1123 |
{
|
| 1124 |
-
"epoch": 1.
|
| 1125 |
-
"learning_rate": 1.
|
| 1126 |
-
"loss": 0.
|
| 1127 |
"step": 900
|
| 1128 |
},
|
| 1129 |
{
|
| 1130 |
-
"epoch": 1.
|
| 1131 |
-
"learning_rate": 1.
|
| 1132 |
-
"loss": 0.
|
| 1133 |
"step": 905
|
| 1134 |
},
|
| 1135 |
{
|
| 1136 |
-
"epoch": 1.
|
| 1137 |
-
"learning_rate": 1.
|
| 1138 |
-
"loss": 0.
|
| 1139 |
"step": 910
|
| 1140 |
},
|
| 1141 |
{
|
| 1142 |
-
"epoch": 1.
|
| 1143 |
-
"learning_rate": 1.
|
| 1144 |
-
"loss": 0.
|
| 1145 |
"step": 915
|
| 1146 |
},
|
| 1147 |
{
|
| 1148 |
-
"epoch": 1.
|
| 1149 |
-
"learning_rate": 1.
|
| 1150 |
-
"loss": 0.
|
| 1151 |
"step": 920
|
| 1152 |
},
|
| 1153 |
{
|
| 1154 |
-
"epoch": 1.
|
| 1155 |
-
"learning_rate": 1.
|
| 1156 |
-
"loss": 0.
|
| 1157 |
"step": 925
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
-
"epoch": 1.
|
| 1161 |
-
"learning_rate": 1.
|
| 1162 |
-
"loss": 0.
|
| 1163 |
"step": 930
|
| 1164 |
},
|
| 1165 |
{
|
| 1166 |
-
"epoch": 1.
|
| 1167 |
-
"learning_rate": 1.
|
| 1168 |
-
"loss": 0.
|
| 1169 |
"step": 935
|
| 1170 |
},
|
| 1171 |
{
|
| 1172 |
-
"epoch": 1.
|
| 1173 |
-
"learning_rate": 1.
|
| 1174 |
-
"loss": 0.
|
| 1175 |
"step": 940
|
| 1176 |
},
|
| 1177 |
{
|
| 1178 |
-
"epoch": 1.
|
| 1179 |
-
"learning_rate": 1.
|
| 1180 |
-
"loss": 0.
|
| 1181 |
"step": 945
|
| 1182 |
},
|
| 1183 |
{
|
| 1184 |
-
"epoch": 1.
|
| 1185 |
-
"learning_rate": 1.
|
| 1186 |
-
"loss": 0.
|
| 1187 |
"step": 950
|
| 1188 |
},
|
| 1189 |
{
|
| 1190 |
-
"epoch": 1.
|
| 1191 |
-
"learning_rate": 1.
|
| 1192 |
-
"loss": 0.
|
| 1193 |
"step": 955
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
-
"epoch": 1.
|
| 1197 |
-
"learning_rate": 1.
|
| 1198 |
-
"loss": 0.
|
| 1199 |
"step": 960
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
-
"epoch": 1.
|
| 1203 |
-
"learning_rate":
|
| 1204 |
-
"loss": 0.
|
| 1205 |
"step": 965
|
| 1206 |
},
|
| 1207 |
{
|
| 1208 |
-
"epoch": 1.
|
| 1209 |
-
"learning_rate":
|
| 1210 |
-
"loss": 0.
|
| 1211 |
"step": 970
|
| 1212 |
},
|
| 1213 |
{
|
| 1214 |
-
"epoch": 1.
|
| 1215 |
-
"learning_rate":
|
| 1216 |
-
"loss": 0.
|
| 1217 |
"step": 975
|
| 1218 |
},
|
| 1219 |
{
|
| 1220 |
-
"epoch": 1.
|
| 1221 |
-
"learning_rate":
|
| 1222 |
-
"loss": 0.
|
| 1223 |
"step": 980
|
| 1224 |
},
|
| 1225 |
{
|
| 1226 |
-
"epoch": 1.
|
| 1227 |
-
"learning_rate":
|
| 1228 |
-
"loss": 0.
|
| 1229 |
"step": 985
|
| 1230 |
},
|
| 1231 |
{
|
| 1232 |
-
"epoch": 1.
|
| 1233 |
-
"learning_rate":
|
| 1234 |
-
"loss": 0.
|
| 1235 |
"step": 990
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
-
"epoch": 1.
|
| 1239 |
-
"learning_rate":
|
| 1240 |
-
"loss": 0.
|
| 1241 |
"step": 995
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
-
"epoch": 1.
|
| 1245 |
-
"learning_rate":
|
| 1246 |
-
"loss": 0.
|
| 1247 |
-
"step": 1000
|
| 1248 |
-
},
|
| 1249 |
-
{
|
| 1250 |
-
"epoch": 1.56,
|
| 1251 |
-
"eval_loss": 1.6513299942016602,
|
| 1252 |
-
"eval_runtime": 156.7333,
|
| 1253 |
-
"eval_samples_per_second": 10.719,
|
| 1254 |
-
"eval_steps_per_second": 0.115,
|
| 1255 |
"step": 1000
|
| 1256 |
},
|
| 1257 |
{
|
| 1258 |
-
"epoch": 1.
|
| 1259 |
-
"learning_rate":
|
| 1260 |
-
"loss": 0.
|
| 1261 |
"step": 1005
|
| 1262 |
},
|
| 1263 |
{
|
| 1264 |
-
"epoch": 1.
|
| 1265 |
-
"learning_rate":
|
| 1266 |
-
"loss": 0.
|
| 1267 |
"step": 1010
|
| 1268 |
},
|
| 1269 |
{
|
| 1270 |
-
"epoch": 1.
|
| 1271 |
-
"learning_rate":
|
| 1272 |
-
"loss": 0.
|
| 1273 |
"step": 1015
|
| 1274 |
},
|
| 1275 |
{
|
| 1276 |
-
"epoch": 1.
|
| 1277 |
-
"learning_rate":
|
| 1278 |
-
"loss": 0.
|
| 1279 |
"step": 1020
|
| 1280 |
},
|
| 1281 |
{
|
| 1282 |
-
"epoch": 1.
|
| 1283 |
-
"learning_rate":
|
| 1284 |
-
"loss": 0.
|
| 1285 |
"step": 1025
|
| 1286 |
},
|
| 1287 |
{
|
| 1288 |
-
"epoch": 1.
|
| 1289 |
-
"learning_rate":
|
| 1290 |
-
"loss": 0.
|
| 1291 |
"step": 1030
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
-
"epoch": 1.
|
| 1295 |
-
"learning_rate":
|
| 1296 |
-
"loss": 0.
|
| 1297 |
"step": 1035
|
| 1298 |
},
|
| 1299 |
{
|
| 1300 |
-
"epoch": 1.
|
| 1301 |
-
"learning_rate":
|
| 1302 |
-
"loss": 0.
|
| 1303 |
"step": 1040
|
| 1304 |
},
|
| 1305 |
{
|
| 1306 |
-
"epoch": 1.
|
| 1307 |
-
"learning_rate":
|
| 1308 |
-
"loss": 0.
|
| 1309 |
"step": 1045
|
| 1310 |
},
|
| 1311 |
{
|
| 1312 |
-
"epoch": 1.
|
| 1313 |
-
"learning_rate":
|
| 1314 |
-
"loss": 0.
|
| 1315 |
"step": 1050
|
| 1316 |
},
|
| 1317 |
{
|
| 1318 |
-
"epoch": 1.
|
| 1319 |
-
"learning_rate":
|
| 1320 |
-
"loss": 0.
|
| 1321 |
"step": 1055
|
| 1322 |
},
|
| 1323 |
{
|
| 1324 |
-
"epoch": 1.
|
| 1325 |
-
"learning_rate":
|
| 1326 |
-
"loss": 0.
|
| 1327 |
"step": 1060
|
| 1328 |
},
|
| 1329 |
{
|
| 1330 |
-
"epoch": 1.
|
| 1331 |
-
"learning_rate":
|
| 1332 |
-
"loss": 0.
|
| 1333 |
"step": 1065
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
-
"epoch": 1.
|
| 1337 |
-
"learning_rate":
|
| 1338 |
-
"loss": 0.
|
| 1339 |
"step": 1070
|
| 1340 |
},
|
| 1341 |
{
|
| 1342 |
-
"epoch": 1.
|
| 1343 |
-
"learning_rate":
|
| 1344 |
-
"loss": 0.
|
| 1345 |
"step": 1075
|
| 1346 |
},
|
| 1347 |
{
|
| 1348 |
-
"epoch": 1.
|
| 1349 |
-
"learning_rate":
|
| 1350 |
-
"loss": 0.
|
| 1351 |
"step": 1080
|
| 1352 |
},
|
| 1353 |
{
|
| 1354 |
-
"epoch": 1.
|
| 1355 |
-
"learning_rate":
|
| 1356 |
-
"loss": 0.
|
| 1357 |
"step": 1085
|
| 1358 |
},
|
| 1359 |
{
|
| 1360 |
-
"epoch": 1.
|
| 1361 |
-
"learning_rate":
|
| 1362 |
-
"loss": 0.
|
| 1363 |
"step": 1090
|
| 1364 |
},
|
| 1365 |
{
|
| 1366 |
-
"epoch": 1.
|
| 1367 |
-
"learning_rate":
|
| 1368 |
-
"loss": 0.
|
| 1369 |
"step": 1095
|
| 1370 |
},
|
| 1371 |
{
|
| 1372 |
-
"epoch": 1.
|
| 1373 |
-
"learning_rate":
|
| 1374 |
-
"loss": 0.
|
| 1375 |
"step": 1100
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
-
"epoch": 1.
|
| 1379 |
-
"learning_rate":
|
| 1380 |
-
"loss": 0.
|
| 1381 |
"step": 1105
|
| 1382 |
},
|
| 1383 |
{
|
| 1384 |
-
"epoch": 1.
|
| 1385 |
-
"learning_rate":
|
| 1386 |
-
"loss": 0.
|
| 1387 |
"step": 1110
|
| 1388 |
},
|
| 1389 |
{
|
| 1390 |
-
"epoch": 1.
|
| 1391 |
-
"learning_rate":
|
| 1392 |
-
"loss": 0.
|
| 1393 |
"step": 1115
|
| 1394 |
},
|
| 1395 |
{
|
| 1396 |
-
"epoch": 1.
|
| 1397 |
-
"learning_rate":
|
| 1398 |
-
"loss": 0.
|
| 1399 |
"step": 1120
|
| 1400 |
},
|
| 1401 |
{
|
| 1402 |
-
"epoch": 1.
|
| 1403 |
-
"learning_rate":
|
| 1404 |
-
"loss": 0.
|
| 1405 |
"step": 1125
|
| 1406 |
},
|
| 1407 |
{
|
| 1408 |
-
"epoch": 1.
|
| 1409 |
-
"learning_rate":
|
| 1410 |
-
"loss": 0.
|
| 1411 |
"step": 1130
|
| 1412 |
},
|
| 1413 |
{
|
| 1414 |
-
"epoch": 1.
|
| 1415 |
-
"learning_rate":
|
| 1416 |
-
"loss": 0.
|
| 1417 |
"step": 1135
|
| 1418 |
},
|
| 1419 |
{
|
| 1420 |
-
"epoch": 1.
|
| 1421 |
-
"learning_rate":
|
| 1422 |
-
"loss": 0.
|
| 1423 |
"step": 1140
|
| 1424 |
},
|
| 1425 |
{
|
| 1426 |
-
"epoch": 1.
|
| 1427 |
-
"learning_rate":
|
| 1428 |
-
"loss": 0.
|
| 1429 |
"step": 1145
|
| 1430 |
},
|
| 1431 |
{
|
| 1432 |
-
"epoch": 1.
|
| 1433 |
-
"learning_rate":
|
| 1434 |
-
"loss": 0.
|
| 1435 |
"step": 1150
|
| 1436 |
},
|
| 1437 |
{
|
| 1438 |
-
"epoch": 1.
|
| 1439 |
-
"learning_rate":
|
| 1440 |
-
"loss": 0.
|
| 1441 |
"step": 1155
|
| 1442 |
},
|
| 1443 |
{
|
| 1444 |
-
"epoch": 1.
|
| 1445 |
-
"learning_rate":
|
| 1446 |
-
"loss": 0.
|
| 1447 |
"step": 1160
|
| 1448 |
},
|
| 1449 |
{
|
| 1450 |
-
"epoch":
|
| 1451 |
-
"
|
| 1452 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1453 |
"step": 1165
|
| 1454 |
},
|
| 1455 |
{
|
| 1456 |
-
"epoch":
|
| 1457 |
-
"learning_rate":
|
| 1458 |
-
"loss": 0.
|
| 1459 |
"step": 1170
|
| 1460 |
},
|
| 1461 |
{
|
| 1462 |
-
"epoch":
|
| 1463 |
-
"learning_rate":
|
| 1464 |
-
"loss": 0.
|
| 1465 |
"step": 1175
|
| 1466 |
},
|
| 1467 |
{
|
| 1468 |
-
"epoch":
|
| 1469 |
-
"learning_rate":
|
| 1470 |
-
"loss": 0.
|
| 1471 |
"step": 1180
|
| 1472 |
},
|
| 1473 |
{
|
| 1474 |
-
"epoch":
|
| 1475 |
-
"learning_rate":
|
| 1476 |
-
"loss": 0.
|
| 1477 |
"step": 1185
|
| 1478 |
},
|
| 1479 |
{
|
| 1480 |
-
"epoch":
|
| 1481 |
-
"learning_rate":
|
| 1482 |
-
"loss": 0.
|
| 1483 |
"step": 1190
|
| 1484 |
},
|
| 1485 |
{
|
| 1486 |
-
"epoch":
|
| 1487 |
-
"learning_rate":
|
| 1488 |
-
"loss": 0.
|
| 1489 |
"step": 1195
|
| 1490 |
},
|
| 1491 |
{
|
| 1492 |
-
"epoch":
|
| 1493 |
-
"learning_rate":
|
| 1494 |
-
"loss": 0.
|
| 1495 |
-
"step": 1200
|
| 1496 |
-
},
|
| 1497 |
-
{
|
| 1498 |
-
"epoch": 1.88,
|
| 1499 |
-
"eval_loss": 1.6570870876312256,
|
| 1500 |
-
"eval_runtime": 156.7819,
|
| 1501 |
-
"eval_samples_per_second": 10.716,
|
| 1502 |
-
"eval_steps_per_second": 0.115,
|
| 1503 |
"step": 1200
|
| 1504 |
},
|
| 1505 |
{
|
| 1506 |
-
"epoch":
|
| 1507 |
-
"learning_rate":
|
| 1508 |
-
"loss": 0.
|
| 1509 |
"step": 1205
|
| 1510 |
},
|
| 1511 |
{
|
| 1512 |
-
"epoch":
|
| 1513 |
-
"learning_rate":
|
| 1514 |
-
"loss": 0.
|
| 1515 |
"step": 1210
|
| 1516 |
},
|
| 1517 |
{
|
| 1518 |
-
"epoch":
|
| 1519 |
-
"learning_rate":
|
| 1520 |
-
"loss": 0.
|
| 1521 |
"step": 1215
|
| 1522 |
},
|
| 1523 |
{
|
| 1524 |
-
"epoch": 1
|
| 1525 |
-
"learning_rate":
|
| 1526 |
-
"loss": 0.
|
| 1527 |
"step": 1220
|
| 1528 |
},
|
| 1529 |
{
|
| 1530 |
-
"epoch": 1
|
| 1531 |
-
"learning_rate":
|
| 1532 |
-
"loss": 0.
|
| 1533 |
"step": 1225
|
| 1534 |
},
|
| 1535 |
{
|
| 1536 |
-
"epoch":
|
| 1537 |
-
"learning_rate":
|
| 1538 |
-
"loss": 0.
|
| 1539 |
"step": 1230
|
| 1540 |
},
|
| 1541 |
{
|
| 1542 |
-
"epoch":
|
| 1543 |
-
"learning_rate":
|
| 1544 |
-
"loss": 0.
|
| 1545 |
"step": 1235
|
| 1546 |
},
|
| 1547 |
{
|
| 1548 |
-
"epoch":
|
| 1549 |
-
"learning_rate":
|
| 1550 |
-
"loss": 0.
|
| 1551 |
"step": 1240
|
| 1552 |
},
|
| 1553 |
{
|
| 1554 |
-
"epoch":
|
| 1555 |
-
"learning_rate":
|
| 1556 |
-
"loss": 0.
|
| 1557 |
"step": 1245
|
| 1558 |
},
|
| 1559 |
{
|
| 1560 |
-
"epoch":
|
| 1561 |
-
"learning_rate":
|
| 1562 |
-
"loss": 0.
|
| 1563 |
"step": 1250
|
| 1564 |
},
|
| 1565 |
{
|
| 1566 |
-
"epoch":
|
| 1567 |
-
"learning_rate":
|
| 1568 |
-
"loss": 0.
|
| 1569 |
"step": 1255
|
| 1570 |
},
|
| 1571 |
{
|
| 1572 |
-
"epoch":
|
| 1573 |
-
"learning_rate":
|
| 1574 |
-
"loss": 0.
|
| 1575 |
"step": 1260
|
| 1576 |
},
|
| 1577 |
{
|
| 1578 |
-
"epoch":
|
| 1579 |
-
"learning_rate":
|
| 1580 |
-
"loss": 0.
|
| 1581 |
"step": 1265
|
| 1582 |
},
|
| 1583 |
{
|
| 1584 |
-
"epoch":
|
| 1585 |
-
"learning_rate":
|
| 1586 |
-
"loss": 0.
|
| 1587 |
"step": 1270
|
| 1588 |
},
|
| 1589 |
{
|
| 1590 |
-
"epoch":
|
| 1591 |
-
"learning_rate":
|
| 1592 |
-
"loss": 0.
|
| 1593 |
"step": 1275
|
| 1594 |
},
|
| 1595 |
{
|
| 1596 |
-
"epoch": 2.
|
| 1597 |
-
"learning_rate":
|
| 1598 |
-
"loss": 0.
|
| 1599 |
"step": 1280
|
| 1600 |
},
|
| 1601 |
{
|
| 1602 |
-
"epoch": 2.
|
| 1603 |
-
"learning_rate":
|
| 1604 |
-
"loss": 0.
|
| 1605 |
"step": 1285
|
| 1606 |
},
|
| 1607 |
{
|
| 1608 |
-
"epoch": 2.
|
| 1609 |
-
"learning_rate":
|
| 1610 |
-
"loss": 0.
|
| 1611 |
"step": 1290
|
| 1612 |
},
|
| 1613 |
{
|
| 1614 |
-
"epoch": 2.
|
| 1615 |
-
"learning_rate":
|
| 1616 |
-
"loss": 0.
|
| 1617 |
"step": 1295
|
| 1618 |
},
|
| 1619 |
{
|
| 1620 |
-
"epoch": 2.
|
| 1621 |
-
"learning_rate":
|
| 1622 |
-
"loss": 0.
|
| 1623 |
"step": 1300
|
| 1624 |
},
|
| 1625 |
{
|
| 1626 |
-
"epoch": 2.
|
| 1627 |
-
"learning_rate":
|
| 1628 |
-
"loss": 0.
|
| 1629 |
"step": 1305
|
| 1630 |
},
|
| 1631 |
{
|
| 1632 |
-
"epoch": 2.
|
| 1633 |
-
"learning_rate":
|
| 1634 |
-
"loss": 0.
|
| 1635 |
"step": 1310
|
| 1636 |
},
|
| 1637 |
{
|
| 1638 |
-
"epoch": 2.
|
| 1639 |
-
"learning_rate":
|
| 1640 |
-
"loss": 0.
|
| 1641 |
"step": 1315
|
| 1642 |
},
|
| 1643 |
{
|
| 1644 |
-
"epoch": 2.
|
| 1645 |
-
"learning_rate":
|
| 1646 |
-
"loss": 0.
|
| 1647 |
"step": 1320
|
| 1648 |
},
|
| 1649 |
{
|
| 1650 |
-
"epoch": 2.
|
| 1651 |
-
"learning_rate":
|
| 1652 |
-
"loss": 0.
|
| 1653 |
"step": 1325
|
| 1654 |
},
|
| 1655 |
{
|
| 1656 |
-
"epoch": 2.
|
| 1657 |
-
"learning_rate":
|
| 1658 |
-
"loss": 0.
|
| 1659 |
"step": 1330
|
| 1660 |
},
|
| 1661 |
{
|
| 1662 |
-
"epoch": 2.
|
| 1663 |
-
"learning_rate":
|
| 1664 |
-
"loss": 0.
|
| 1665 |
"step": 1335
|
| 1666 |
},
|
| 1667 |
{
|
| 1668 |
-
"epoch": 2.
|
| 1669 |
-
"learning_rate":
|
| 1670 |
-
"loss": 0.
|
| 1671 |
"step": 1340
|
| 1672 |
},
|
| 1673 |
{
|
| 1674 |
-
"epoch": 2.
|
| 1675 |
-
"learning_rate":
|
| 1676 |
-
"loss": 0.
|
| 1677 |
"step": 1345
|
| 1678 |
},
|
| 1679 |
{
|
| 1680 |
-
"epoch": 2.
|
| 1681 |
-
"learning_rate":
|
| 1682 |
-
"loss": 0.
|
| 1683 |
"step": 1350
|
| 1684 |
},
|
| 1685 |
{
|
| 1686 |
-
"epoch": 2.
|
| 1687 |
-
"learning_rate":
|
| 1688 |
-
"loss": 0.
|
| 1689 |
"step": 1355
|
| 1690 |
},
|
| 1691 |
{
|
| 1692 |
-
"epoch": 2.
|
| 1693 |
-
"learning_rate":
|
| 1694 |
-
"loss": 0.
|
| 1695 |
"step": 1360
|
| 1696 |
},
|
| 1697 |
{
|
| 1698 |
-
"epoch": 2.
|
| 1699 |
-
"learning_rate":
|
| 1700 |
-
"loss": 0.
|
| 1701 |
"step": 1365
|
| 1702 |
},
|
| 1703 |
{
|
| 1704 |
-
"epoch": 2.
|
| 1705 |
-
"learning_rate":
|
| 1706 |
-
"loss": 0.
|
| 1707 |
"step": 1370
|
| 1708 |
},
|
| 1709 |
{
|
| 1710 |
-
"epoch": 2.
|
| 1711 |
-
"learning_rate":
|
| 1712 |
-
"loss": 0.
|
| 1713 |
"step": 1375
|
| 1714 |
},
|
| 1715 |
{
|
| 1716 |
-
"epoch": 2.
|
| 1717 |
-
"learning_rate":
|
| 1718 |
-
"loss": 0.
|
| 1719 |
"step": 1380
|
| 1720 |
},
|
| 1721 |
{
|
| 1722 |
-
"epoch": 2.
|
| 1723 |
-
"learning_rate":
|
| 1724 |
-
"loss": 0.
|
| 1725 |
"step": 1385
|
| 1726 |
},
|
| 1727 |
{
|
| 1728 |
-
"epoch": 2.
|
| 1729 |
-
"learning_rate":
|
| 1730 |
-
"loss": 0.
|
| 1731 |
"step": 1390
|
| 1732 |
},
|
| 1733 |
{
|
| 1734 |
-
"epoch": 2.
|
| 1735 |
-
"learning_rate":
|
| 1736 |
-
"loss": 0.
|
| 1737 |
"step": 1395
|
| 1738 |
},
|
| 1739 |
{
|
| 1740 |
-
"epoch": 2.
|
| 1741 |
-
"learning_rate":
|
| 1742 |
-
"loss": 0.
|
| 1743 |
-
"step": 1400
|
| 1744 |
-
},
|
| 1745 |
-
{
|
| 1746 |
-
"epoch": 2.19,
|
| 1747 |
-
"eval_loss": 1.7886532545089722,
|
| 1748 |
-
"eval_runtime": 156.7662,
|
| 1749 |
-
"eval_samples_per_second": 10.717,
|
| 1750 |
-
"eval_steps_per_second": 0.115,
|
| 1751 |
"step": 1400
|
| 1752 |
},
|
| 1753 |
{
|
| 1754 |
-
"epoch": 2.
|
| 1755 |
-
"learning_rate":
|
| 1756 |
-
"loss": 0.
|
| 1757 |
"step": 1405
|
| 1758 |
},
|
| 1759 |
{
|
| 1760 |
-
"epoch": 2.
|
| 1761 |
-
"learning_rate":
|
| 1762 |
-
"loss": 0.
|
| 1763 |
"step": 1410
|
| 1764 |
},
|
| 1765 |
{
|
| 1766 |
-
"epoch": 2.
|
| 1767 |
-
"learning_rate":
|
| 1768 |
-
"loss": 0.
|
| 1769 |
"step": 1415
|
| 1770 |
},
|
| 1771 |
{
|
| 1772 |
-
"epoch": 2.
|
| 1773 |
-
"learning_rate":
|
| 1774 |
-
"loss": 0.
|
| 1775 |
"step": 1420
|
| 1776 |
},
|
| 1777 |
{
|
| 1778 |
-
"epoch": 2.
|
| 1779 |
-
"learning_rate":
|
| 1780 |
-
"loss": 0.
|
| 1781 |
"step": 1425
|
| 1782 |
},
|
| 1783 |
{
|
| 1784 |
-
"epoch": 2.
|
| 1785 |
-
"learning_rate":
|
| 1786 |
-
"loss": 0.
|
| 1787 |
"step": 1430
|
| 1788 |
},
|
| 1789 |
{
|
| 1790 |
-
"epoch": 2.
|
| 1791 |
-
"learning_rate":
|
| 1792 |
-
"loss": 0.
|
| 1793 |
"step": 1435
|
| 1794 |
},
|
| 1795 |
{
|
| 1796 |
-
"epoch": 2.
|
| 1797 |
-
"learning_rate":
|
| 1798 |
-
"loss": 0.
|
| 1799 |
"step": 1440
|
| 1800 |
},
|
| 1801 |
{
|
| 1802 |
-
"epoch": 2.
|
| 1803 |
-
"learning_rate":
|
| 1804 |
-
"loss": 0.
|
| 1805 |
"step": 1445
|
| 1806 |
},
|
| 1807 |
{
|
| 1808 |
-
"epoch": 2.
|
| 1809 |
-
"learning_rate":
|
| 1810 |
-
"loss": 0.
|
| 1811 |
"step": 1450
|
| 1812 |
},
|
| 1813 |
{
|
| 1814 |
-
"epoch": 2.
|
| 1815 |
-
"learning_rate":
|
| 1816 |
-
"loss": 0.
|
| 1817 |
"step": 1455
|
| 1818 |
},
|
| 1819 |
{
|
| 1820 |
-
"epoch": 2.
|
| 1821 |
-
"learning_rate":
|
| 1822 |
-
"loss": 0.
|
| 1823 |
"step": 1460
|
| 1824 |
},
|
| 1825 |
{
|
| 1826 |
-
"epoch": 2.
|
| 1827 |
-
"learning_rate":
|
| 1828 |
-
"loss": 0.
|
| 1829 |
"step": 1465
|
| 1830 |
},
|
| 1831 |
{
|
| 1832 |
-
"epoch": 2.
|
| 1833 |
-
"learning_rate":
|
| 1834 |
-
"loss": 0.
|
| 1835 |
"step": 1470
|
| 1836 |
},
|
| 1837 |
{
|
| 1838 |
-
"epoch": 2.
|
| 1839 |
-
"learning_rate":
|
| 1840 |
-
"loss": 0.
|
| 1841 |
"step": 1475
|
| 1842 |
},
|
| 1843 |
{
|
| 1844 |
-
"epoch": 2.
|
| 1845 |
-
"learning_rate":
|
| 1846 |
-
"loss": 0.
|
| 1847 |
"step": 1480
|
| 1848 |
},
|
| 1849 |
{
|
| 1850 |
-
"epoch": 2.
|
| 1851 |
-
"learning_rate":
|
| 1852 |
-
"loss": 0.
|
| 1853 |
"step": 1485
|
| 1854 |
},
|
| 1855 |
{
|
| 1856 |
-
"epoch": 2.
|
| 1857 |
-
"learning_rate":
|
| 1858 |
-
"loss": 0.
|
| 1859 |
"step": 1490
|
| 1860 |
},
|
| 1861 |
{
|
| 1862 |
-
"epoch": 2.
|
| 1863 |
-
"learning_rate":
|
| 1864 |
-
"loss": 0.
|
| 1865 |
"step": 1495
|
| 1866 |
},
|
| 1867 |
{
|
| 1868 |
-
"epoch": 2.
|
| 1869 |
-
"learning_rate":
|
| 1870 |
-
"loss": 0.
|
| 1871 |
"step": 1500
|
| 1872 |
},
|
| 1873 |
{
|
| 1874 |
-
"epoch": 2.
|
| 1875 |
-
"learning_rate":
|
| 1876 |
-
"loss": 0.
|
| 1877 |
"step": 1505
|
| 1878 |
},
|
| 1879 |
{
|
| 1880 |
-
"epoch": 2.
|
| 1881 |
-
"learning_rate":
|
| 1882 |
-
"loss": 0.
|
| 1883 |
"step": 1510
|
| 1884 |
},
|
| 1885 |
{
|
| 1886 |
-
"epoch": 2.
|
| 1887 |
-
"learning_rate":
|
| 1888 |
-
"loss": 0.
|
| 1889 |
"step": 1515
|
| 1890 |
},
|
| 1891 |
{
|
| 1892 |
-
"epoch": 2.
|
| 1893 |
-
"learning_rate":
|
| 1894 |
-
"loss": 0.
|
| 1895 |
"step": 1520
|
| 1896 |
},
|
| 1897 |
{
|
| 1898 |
-
"epoch": 2.
|
| 1899 |
-
"learning_rate":
|
| 1900 |
-
"loss": 0.
|
| 1901 |
"step": 1525
|
| 1902 |
},
|
| 1903 |
{
|
| 1904 |
-
"epoch": 2.
|
| 1905 |
-
"learning_rate":
|
| 1906 |
-
"loss": 0.
|
| 1907 |
"step": 1530
|
| 1908 |
},
|
| 1909 |
{
|
| 1910 |
-
"epoch": 2.
|
| 1911 |
-
"learning_rate":
|
| 1912 |
-
"loss": 0.
|
| 1913 |
"step": 1535
|
| 1914 |
},
|
| 1915 |
{
|
| 1916 |
-
"epoch": 2.
|
| 1917 |
-
"learning_rate":
|
| 1918 |
-
"loss": 0.
|
| 1919 |
"step": 1540
|
| 1920 |
},
|
| 1921 |
{
|
| 1922 |
-
"epoch": 2.
|
| 1923 |
-
"learning_rate":
|
| 1924 |
-
"loss": 0.
|
| 1925 |
"step": 1545
|
| 1926 |
},
|
| 1927 |
{
|
| 1928 |
-
"epoch": 2.
|
| 1929 |
-
"learning_rate":
|
| 1930 |
-
"loss": 0.
|
| 1931 |
"step": 1550
|
| 1932 |
},
|
| 1933 |
{
|
| 1934 |
-
"epoch": 2.
|
| 1935 |
-
"learning_rate":
|
| 1936 |
-
"loss": 0.
|
| 1937 |
"step": 1555
|
| 1938 |
},
|
| 1939 |
{
|
| 1940 |
-
"epoch": 2.
|
| 1941 |
-
"learning_rate":
|
| 1942 |
-
"loss": 0.
|
| 1943 |
"step": 1560
|
| 1944 |
},
|
| 1945 |
{
|
| 1946 |
-
"epoch": 2.
|
| 1947 |
-
"learning_rate":
|
| 1948 |
-
"loss": 0.
|
| 1949 |
"step": 1565
|
| 1950 |
},
|
| 1951 |
{
|
| 1952 |
-
"epoch": 2.
|
| 1953 |
-
"learning_rate":
|
| 1954 |
-
"loss": 0.
|
| 1955 |
"step": 1570
|
| 1956 |
},
|
| 1957 |
{
|
| 1958 |
-
"epoch": 2.
|
| 1959 |
-
"learning_rate":
|
| 1960 |
-
"loss": 0.
|
| 1961 |
"step": 1575
|
| 1962 |
},
|
| 1963 |
{
|
| 1964 |
-
"epoch": 2.
|
| 1965 |
-
"learning_rate":
|
| 1966 |
-
"loss": 0.
|
| 1967 |
"step": 1580
|
| 1968 |
},
|
| 1969 |
{
|
| 1970 |
-
"epoch": 2.
|
| 1971 |
-
"learning_rate":
|
| 1972 |
-
"loss": 0.
|
| 1973 |
"step": 1585
|
| 1974 |
},
|
| 1975 |
{
|
| 1976 |
-
"epoch": 2.
|
| 1977 |
-
"learning_rate":
|
| 1978 |
-
"loss": 0.
|
| 1979 |
"step": 1590
|
| 1980 |
},
|
| 1981 |
{
|
| 1982 |
-
"epoch": 2.
|
| 1983 |
-
"learning_rate":
|
| 1984 |
-
"loss": 0.
|
| 1985 |
"step": 1595
|
| 1986 |
},
|
| 1987 |
{
|
| 1988 |
-
"epoch": 2.
|
| 1989 |
-
"learning_rate":
|
| 1990 |
-
"loss": 0.
|
| 1991 |
-
"step": 1600
|
| 1992 |
-
},
|
| 1993 |
-
{
|
| 1994 |
-
"epoch": 2.5,
|
| 1995 |
-
"eval_loss": 1.8198195695877075,
|
| 1996 |
-
"eval_runtime": 156.7807,
|
| 1997 |
-
"eval_samples_per_second": 10.716,
|
| 1998 |
-
"eval_steps_per_second": 0.115,
|
| 1999 |
"step": 1600
|
| 2000 |
},
|
| 2001 |
{
|
| 2002 |
-
"epoch": 2.
|
| 2003 |
-
"learning_rate":
|
| 2004 |
-
"loss": 0.
|
| 2005 |
"step": 1605
|
| 2006 |
},
|
| 2007 |
{
|
| 2008 |
-
"epoch": 2.
|
| 2009 |
-
"learning_rate":
|
| 2010 |
-
"loss": 0.
|
| 2011 |
"step": 1610
|
| 2012 |
},
|
| 2013 |
{
|
| 2014 |
-
"epoch": 2.
|
| 2015 |
-
"learning_rate":
|
| 2016 |
-
"loss": 0.
|
| 2017 |
"step": 1615
|
| 2018 |
},
|
| 2019 |
{
|
| 2020 |
-
"epoch": 2.
|
| 2021 |
-
"learning_rate":
|
| 2022 |
-
"loss": 0.
|
| 2023 |
"step": 1620
|
| 2024 |
},
|
| 2025 |
{
|
| 2026 |
-
"epoch": 2.
|
| 2027 |
-
"learning_rate":
|
| 2028 |
-
"loss": 0.
|
| 2029 |
"step": 1625
|
| 2030 |
},
|
| 2031 |
{
|
| 2032 |
-
"epoch": 2.
|
| 2033 |
-
"learning_rate":
|
| 2034 |
-
"loss": 0.
|
| 2035 |
"step": 1630
|
| 2036 |
},
|
| 2037 |
{
|
| 2038 |
-
"epoch": 2.
|
| 2039 |
-
"learning_rate":
|
| 2040 |
-
"loss": 0.
|
| 2041 |
"step": 1635
|
| 2042 |
},
|
| 2043 |
{
|
| 2044 |
-
"epoch": 2.
|
| 2045 |
-
"learning_rate":
|
| 2046 |
-
"loss": 0.
|
| 2047 |
"step": 1640
|
| 2048 |
},
|
| 2049 |
{
|
| 2050 |
-
"epoch": 2.
|
| 2051 |
-
"learning_rate":
|
| 2052 |
-
"loss": 0.
|
| 2053 |
"step": 1645
|
| 2054 |
},
|
| 2055 |
{
|
| 2056 |
-
"epoch": 2.
|
| 2057 |
-
"learning_rate": 1.
|
| 2058 |
-
"loss": 0.
|
| 2059 |
"step": 1650
|
| 2060 |
},
|
| 2061 |
{
|
| 2062 |
-
"epoch": 2.
|
| 2063 |
-
"learning_rate": 1.
|
| 2064 |
-
"loss": 0.
|
| 2065 |
"step": 1655
|
| 2066 |
},
|
| 2067 |
{
|
| 2068 |
-
"epoch": 2.
|
| 2069 |
-
"learning_rate": 1.
|
| 2070 |
-
"loss": 0.
|
| 2071 |
"step": 1660
|
| 2072 |
},
|
| 2073 |
{
|
| 2074 |
-
"epoch": 2.
|
| 2075 |
-
"learning_rate": 1.
|
| 2076 |
-
"loss": 0.
|
| 2077 |
"step": 1665
|
| 2078 |
},
|
| 2079 |
{
|
| 2080 |
-
"epoch": 2.
|
| 2081 |
-
"learning_rate":
|
| 2082 |
-
"loss": 0.
|
| 2083 |
"step": 1670
|
| 2084 |
},
|
| 2085 |
{
|
| 2086 |
-
"epoch": 2.
|
| 2087 |
-
"learning_rate":
|
| 2088 |
-
"loss": 0.
|
| 2089 |
"step": 1675
|
| 2090 |
},
|
| 2091 |
{
|
| 2092 |
-
"epoch": 2.
|
| 2093 |
-
"learning_rate":
|
| 2094 |
-
"loss": 0.
|
| 2095 |
"step": 1680
|
| 2096 |
},
|
| 2097 |
{
|
| 2098 |
-
"epoch": 2.
|
| 2099 |
-
"learning_rate":
|
| 2100 |
-
"loss": 0.
|
| 2101 |
"step": 1685
|
| 2102 |
},
|
| 2103 |
{
|
| 2104 |
-
"epoch": 2.
|
| 2105 |
-
"learning_rate":
|
| 2106 |
-
"loss": 0.
|
| 2107 |
"step": 1690
|
| 2108 |
},
|
| 2109 |
{
|
| 2110 |
-
"epoch": 2.
|
| 2111 |
-
"learning_rate":
|
| 2112 |
-
"loss": 0.
|
| 2113 |
"step": 1695
|
| 2114 |
},
|
| 2115 |
{
|
| 2116 |
-
"epoch": 2.
|
| 2117 |
-
"learning_rate":
|
| 2118 |
-
"loss": 0.
|
| 2119 |
"step": 1700
|
| 2120 |
},
|
| 2121 |
{
|
| 2122 |
-
"epoch": 2.
|
| 2123 |
-
"learning_rate":
|
| 2124 |
-
"loss": 0.
|
| 2125 |
"step": 1705
|
| 2126 |
},
|
| 2127 |
{
|
| 2128 |
-
"epoch": 2.
|
| 2129 |
-
"learning_rate":
|
| 2130 |
-
"loss": 0.
|
| 2131 |
"step": 1710
|
| 2132 |
},
|
| 2133 |
{
|
| 2134 |
-
"epoch": 2.
|
| 2135 |
-
"learning_rate":
|
| 2136 |
-
"loss": 0.
|
| 2137 |
"step": 1715
|
| 2138 |
},
|
| 2139 |
{
|
| 2140 |
-
"epoch": 2.
|
| 2141 |
-
"learning_rate":
|
| 2142 |
-
"loss": 0.
|
| 2143 |
"step": 1720
|
| 2144 |
},
|
| 2145 |
{
|
| 2146 |
-
"epoch": 2.
|
| 2147 |
-
"learning_rate":
|
| 2148 |
-
"loss": 0.
|
| 2149 |
"step": 1725
|
| 2150 |
},
|
| 2151 |
{
|
| 2152 |
-
"epoch": 2.
|
| 2153 |
-
"learning_rate": 5.
|
| 2154 |
-
"loss": 0.
|
| 2155 |
"step": 1730
|
| 2156 |
},
|
| 2157 |
{
|
| 2158 |
-
"epoch": 2.
|
| 2159 |
-
"learning_rate":
|
| 2160 |
-
"loss": 0.
|
| 2161 |
"step": 1735
|
| 2162 |
},
|
| 2163 |
{
|
| 2164 |
-
"epoch": 2.
|
| 2165 |
-
"learning_rate":
|
| 2166 |
-
"loss": 0.
|
| 2167 |
"step": 1740
|
| 2168 |
},
|
| 2169 |
{
|
| 2170 |
-
"epoch":
|
| 2171 |
-
"learning_rate":
|
| 2172 |
-
"loss": 0.
|
| 2173 |
"step": 1745
|
| 2174 |
},
|
| 2175 |
{
|
| 2176 |
-
"epoch":
|
| 2177 |
-
"
|
| 2178 |
-
"
|
| 2179 |
-
"
|
| 2180 |
-
|
| 2181 |
-
|
| 2182 |
-
"epoch": 2.74,
|
| 2183 |
-
"learning_rate": 4.3208478307738e-07,
|
| 2184 |
-
"loss": 0.3614,
|
| 2185 |
-
"step": 1755
|
| 2186 |
-
},
|
| 2187 |
-
{
|
| 2188 |
-
"epoch": 2.75,
|
| 2189 |
-
"learning_rate": 4.0600368992376074e-07,
|
| 2190 |
-
"loss": 0.3555,
|
| 2191 |
-
"step": 1760
|
| 2192 |
-
},
|
| 2193 |
-
{
|
| 2194 |
-
"epoch": 2.76,
|
| 2195 |
-
"learning_rate": 3.8071812959016965e-07,
|
| 2196 |
-
"loss": 0.3618,
|
| 2197 |
-
"step": 1765
|
| 2198 |
-
},
|
| 2199 |
-
{
|
| 2200 |
-
"epoch": 2.77,
|
| 2201 |
-
"learning_rate": 3.5623019875169916e-07,
|
| 2202 |
-
"loss": 0.3667,
|
| 2203 |
-
"step": 1770
|
| 2204 |
-
},
|
| 2205 |
-
{
|
| 2206 |
-
"epoch": 2.78,
|
| 2207 |
-
"learning_rate": 3.325419279441211e-07,
|
| 2208 |
-
"loss": 0.364,
|
| 2209 |
-
"step": 1775
|
| 2210 |
-
},
|
| 2211 |
-
{
|
| 2212 |
-
"epoch": 2.78,
|
| 2213 |
-
"learning_rate": 3.096552813955045e-07,
|
| 2214 |
-
"loss": 0.3693,
|
| 2215 |
-
"step": 1780
|
| 2216 |
-
},
|
| 2217 |
-
{
|
| 2218 |
-
"epoch": 2.79,
|
| 2219 |
-
"learning_rate": 2.875721568633527e-07,
|
| 2220 |
-
"loss": 0.3265,
|
| 2221 |
-
"step": 1785
|
| 2222 |
-
},
|
| 2223 |
-
{
|
| 2224 |
-
"epoch": 2.8,
|
| 2225 |
-
"learning_rate": 2.6629438547723484e-07,
|
| 2226 |
-
"loss": 0.3447,
|
| 2227 |
-
"step": 1790
|
| 2228 |
-
},
|
| 2229 |
-
{
|
| 2230 |
-
"epoch": 2.81,
|
| 2231 |
-
"learning_rate": 2.458237315869527e-07,
|
| 2232 |
-
"loss": 0.3602,
|
| 2233 |
-
"step": 1795
|
| 2234 |
-
},
|
| 2235 |
-
{
|
| 2236 |
-
"epoch": 2.81,
|
| 2237 |
-
"learning_rate": 2.2616189261623568e-07,
|
| 2238 |
-
"loss": 0.3506,
|
| 2239 |
-
"step": 1800
|
| 2240 |
-
},
|
| 2241 |
-
{
|
| 2242 |
-
"epoch": 2.81,
|
| 2243 |
-
"eval_loss": 1.8299107551574707,
|
| 2244 |
-
"eval_runtime": 156.7929,
|
| 2245 |
-
"eval_samples_per_second": 10.715,
|
| 2246 |
-
"eval_steps_per_second": 0.115,
|
| 2247 |
-
"step": 1800
|
| 2248 |
-
},
|
| 2249 |
-
{
|
| 2250 |
-
"epoch": 2.82,
|
| 2251 |
-
"learning_rate": 2.0731049892199873e-07,
|
| 2252 |
-
"loss": 0.3721,
|
| 2253 |
-
"step": 1805
|
| 2254 |
-
},
|
| 2255 |
-
{
|
| 2256 |
-
"epoch": 2.83,
|
| 2257 |
-
"learning_rate": 1.8927111365914297e-07,
|
| 2258 |
-
"loss": 0.3383,
|
| 2259 |
-
"step": 1810
|
| 2260 |
-
},
|
| 2261 |
-
{
|
| 2262 |
-
"epoch": 2.84,
|
| 2263 |
-
"learning_rate": 1.7204523265095252e-07,
|
| 2264 |
-
"loss": 0.3455,
|
| 2265 |
-
"step": 1815
|
| 2266 |
-
},
|
| 2267 |
-
{
|
| 2268 |
-
"epoch": 2.85,
|
| 2269 |
-
"learning_rate": 1.55634284265046e-07,
|
| 2270 |
-
"loss": 0.3635,
|
| 2271 |
-
"step": 1820
|
| 2272 |
-
},
|
| 2273 |
-
{
|
| 2274 |
-
"epoch": 2.85,
|
| 2275 |
-
"learning_rate": 1.400396292949513e-07,
|
| 2276 |
-
"loss": 0.3151,
|
| 2277 |
-
"step": 1825
|
| 2278 |
-
},
|
| 2279 |
-
{
|
| 2280 |
-
"epoch": 2.86,
|
| 2281 |
-
"learning_rate": 1.2526256084725351e-07,
|
| 2282 |
-
"loss": 0.3347,
|
| 2283 |
-
"step": 1830
|
| 2284 |
-
},
|
| 2285 |
-
{
|
| 2286 |
-
"epoch": 2.87,
|
| 2287 |
-
"learning_rate": 1.1130430423438642e-07,
|
| 2288 |
-
"loss": 0.3795,
|
| 2289 |
-
"step": 1835
|
| 2290 |
-
},
|
| 2291 |
-
{
|
| 2292 |
-
"epoch": 2.88,
|
| 2293 |
-
"learning_rate": 9.816601687301585e-08,
|
| 2294 |
-
"loss": 0.3486,
|
| 2295 |
-
"step": 1840
|
| 2296 |
-
},
|
| 2297 |
-
{
|
| 2298 |
-
"epoch": 2.89,
|
| 2299 |
-
"learning_rate": 8.584878818807652e-08,
|
| 2300 |
-
"loss": 0.3646,
|
| 2301 |
-
"step": 1845
|
| 2302 |
-
},
|
| 2303 |
-
{
|
| 2304 |
-
"epoch": 2.89,
|
| 2305 |
-
"learning_rate": 7.435363952243313e-08,
|
| 2306 |
-
"loss": 0.3665,
|
| 2307 |
-
"step": 1850
|
| 2308 |
-
},
|
| 2309 |
-
{
|
| 2310 |
-
"epoch": 2.9,
|
| 2311 |
-
"learning_rate": 6.368152405219042e-08,
|
| 2312 |
-
"loss": 0.3876,
|
| 2313 |
-
"step": 1855
|
| 2314 |
-
},
|
| 2315 |
-
{
|
| 2316 |
-
"epoch": 2.91,
|
| 2317 |
-
"learning_rate": 5.3833326707656284e-08,
|
| 2318 |
-
"loss": 0.3748,
|
| 2319 |
-
"step": 1860
|
| 2320 |
-
},
|
| 2321 |
-
{
|
| 2322 |
-
"epoch": 2.92,
|
| 2323 |
-
"learning_rate": 4.4809864099965014e-08,
|
| 2324 |
-
"loss": 0.3486,
|
| 2325 |
-
"step": 1865
|
| 2326 |
-
},
|
| 2327 |
-
{
|
| 2328 |
-
"epoch": 2.92,
|
| 2329 |
-
"learning_rate": 3.661188445336361e-08,
|
| 2330 |
-
"loss": 0.3427,
|
| 2331 |
-
"step": 1870
|
| 2332 |
-
},
|
| 2333 |
-
{
|
| 2334 |
-
"epoch": 2.93,
|
| 2335 |
-
"learning_rate": 2.9240067543165883e-08,
|
| 2336 |
-
"loss": 0.3381,
|
| 2337 |
-
"step": 1875
|
| 2338 |
-
},
|
| 2339 |
-
{
|
| 2340 |
-
"epoch": 2.94,
|
| 2341 |
-
"learning_rate": 2.2695024639390884e-08,
|
| 2342 |
-
"loss": 0.3717,
|
| 2343 |
-
"step": 1880
|
| 2344 |
-
},
|
| 2345 |
-
{
|
| 2346 |
-
"epoch": 2.95,
|
| 2347 |
-
"learning_rate": 1.697729845607121e-08,
|
| 2348 |
-
"loss": 0.3411,
|
| 2349 |
-
"step": 1885
|
| 2350 |
-
},
|
| 2351 |
-
{
|
| 2352 |
-
"epoch": 2.96,
|
| 2353 |
-
"learning_rate": 1.20873631062568e-08,
|
| 2354 |
-
"loss": 0.3595,
|
| 2355 |
-
"step": 1890
|
| 2356 |
-
},
|
| 2357 |
-
{
|
| 2358 |
-
"epoch": 2.96,
|
| 2359 |
-
"learning_rate": 8.025624062695247e-09,
|
| 2360 |
-
"loss": 0.3628,
|
| 2361 |
-
"step": 1895
|
| 2362 |
-
},
|
| 2363 |
-
{
|
| 2364 |
-
"epoch": 2.97,
|
| 2365 |
-
"learning_rate": 4.792418124214271e-09,
|
| 2366 |
-
"loss": 0.3785,
|
| 2367 |
-
"step": 1900
|
| 2368 |
-
},
|
| 2369 |
-
{
|
| 2370 |
-
"epoch": 2.98,
|
| 2371 |
-
"learning_rate": 2.3880133877962617e-09,
|
| 2372 |
-
"loss": 0.3523,
|
| 2373 |
-
"step": 1905
|
| 2374 |
-
},
|
| 2375 |
-
{
|
| 2376 |
-
"epoch": 2.99,
|
| 2377 |
-
"learning_rate": 8.126092263405217e-10,
|
| 2378 |
-
"loss": 0.3398,
|
| 2379 |
-
"step": 1910
|
| 2380 |
-
},
|
| 2381 |
-
{
|
| 2382 |
-
"epoch": 2.99,
|
| 2383 |
-
"learning_rate": 6.633627213870064e-11,
|
| 2384 |
-
"loss": 0.3745,
|
| 2385 |
-
"step": 1915
|
| 2386 |
},
|
| 2387 |
{
|
| 2388 |
"epoch": 3.0,
|
| 2389 |
-
"step":
|
| 2390 |
-
"total_flos":
|
| 2391 |
-
"train_loss": 0.
|
| 2392 |
-
"train_runtime":
|
| 2393 |
-
"train_samples_per_second":
|
| 2394 |
-
"train_steps_per_second": 0.
|
| 2395 |
}
|
| 2396 |
],
|
| 2397 |
"logging_steps": 5,
|
| 2398 |
-
"max_steps":
|
| 2399 |
"num_train_epochs": 3,
|
| 2400 |
-
"save_steps":
|
| 2401 |
-
"total_flos":
|
| 2402 |
"trial_name": null,
|
| 2403 |
"trial_params": null
|
| 2404 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1746,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0,
|
| 13 |
+
"learning_rate": 1.142857142857143e-07,
|
| 14 |
+
"loss": 2.7168,
|
| 15 |
"step": 1
|
| 16 |
},
|
| 17 |
{
|
| 18 |
"epoch": 0.01,
|
| 19 |
+
"learning_rate": 5.714285714285715e-07,
|
| 20 |
+
"loss": 2.5728,
|
| 21 |
"step": 5
|
| 22 |
},
|
| 23 |
{
|
| 24 |
"epoch": 0.02,
|
| 25 |
+
"learning_rate": 1.142857142857143e-06,
|
| 26 |
+
"loss": 2.5422,
|
| 27 |
"step": 10
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.03,
|
| 31 |
+
"learning_rate": 1.7142857142857145e-06,
|
| 32 |
+
"loss": 2.3792,
|
| 33 |
"step": 15
|
| 34 |
},
|
| 35 |
{
|
| 36 |
"epoch": 0.03,
|
| 37 |
+
"learning_rate": 2.285714285714286e-06,
|
| 38 |
+
"loss": 2.2176,
|
| 39 |
"step": 20
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.04,
|
| 43 |
+
"learning_rate": 2.8571428571428573e-06,
|
| 44 |
+
"loss": 1.9282,
|
| 45 |
"step": 25
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.05,
|
| 49 |
+
"learning_rate": 3.428571428571429e-06,
|
| 50 |
+
"loss": 1.7601,
|
| 51 |
"step": 30
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.06,
|
| 55 |
+
"learning_rate": 4.000000000000001e-06,
|
| 56 |
+
"loss": 1.6887,
|
| 57 |
"step": 35
|
| 58 |
},
|
| 59 |
{
|
| 60 |
+
"epoch": 0.07,
|
| 61 |
+
"learning_rate": 4.571428571428572e-06,
|
| 62 |
+
"loss": 1.6597,
|
| 63 |
"step": 40
|
| 64 |
},
|
| 65 |
{
|
| 66 |
+
"epoch": 0.08,
|
| 67 |
+
"learning_rate": 5.142857142857142e-06,
|
| 68 |
+
"loss": 1.5152,
|
| 69 |
"step": 45
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 0.09,
|
| 73 |
+
"learning_rate": 5.7142857142857145e-06,
|
| 74 |
+
"loss": 1.5448,
|
| 75 |
"step": 50
|
| 76 |
},
|
| 77 |
{
|
| 78 |
"epoch": 0.09,
|
| 79 |
+
"learning_rate": 6.285714285714286e-06,
|
| 80 |
+
"loss": 1.5993,
|
| 81 |
"step": 55
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 0.1,
|
| 85 |
+
"learning_rate": 6.857142857142858e-06,
|
| 86 |
+
"loss": 1.5941,
|
| 87 |
"step": 60
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 0.11,
|
| 91 |
+
"learning_rate": 7.428571428571429e-06,
|
| 92 |
+
"loss": 1.5832,
|
| 93 |
"step": 65
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.12,
|
| 97 |
+
"learning_rate": 8.000000000000001e-06,
|
| 98 |
+
"loss": 1.5369,
|
| 99 |
"step": 70
|
| 100 |
},
|
| 101 |
{
|
| 102 |
+
"epoch": 0.13,
|
| 103 |
+
"learning_rate": 8.571428571428571e-06,
|
| 104 |
+
"loss": 1.4847,
|
| 105 |
"step": 75
|
| 106 |
},
|
| 107 |
{
|
| 108 |
+
"epoch": 0.14,
|
| 109 |
+
"learning_rate": 9.142857142857144e-06,
|
| 110 |
+
"loss": 1.544,
|
| 111 |
"step": 80
|
| 112 |
},
|
| 113 |
{
|
| 114 |
+
"epoch": 0.15,
|
| 115 |
+
"learning_rate": 9.714285714285715e-06,
|
| 116 |
+
"loss": 1.5631,
|
| 117 |
"step": 85
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 0.15,
|
| 121 |
+
"learning_rate": 1.0285714285714285e-05,
|
| 122 |
+
"loss": 1.5124,
|
| 123 |
"step": 90
|
| 124 |
},
|
| 125 |
{
|
| 126 |
+
"epoch": 0.16,
|
| 127 |
+
"learning_rate": 1.0857142857142858e-05,
|
| 128 |
+
"loss": 1.4762,
|
| 129 |
"step": 95
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 0.17,
|
| 133 |
+
"learning_rate": 1.1428571428571429e-05,
|
| 134 |
+
"loss": 1.5304,
|
| 135 |
"step": 100
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.18,
|
| 139 |
+
"learning_rate": 1.2e-05,
|
| 140 |
+
"loss": 1.5102,
|
| 141 |
"step": 105
|
| 142 |
},
|
| 143 |
{
|
| 144 |
+
"epoch": 0.19,
|
| 145 |
+
"learning_rate": 1.2571428571428572e-05,
|
| 146 |
+
"loss": 1.477,
|
| 147 |
"step": 110
|
| 148 |
},
|
| 149 |
{
|
| 150 |
+
"epoch": 0.2,
|
| 151 |
+
"learning_rate": 1.3142857142857145e-05,
|
| 152 |
+
"loss": 1.5082,
|
| 153 |
"step": 115
|
| 154 |
},
|
| 155 |
{
|
| 156 |
+
"epoch": 0.21,
|
| 157 |
+
"learning_rate": 1.3714285714285716e-05,
|
| 158 |
+
"loss": 1.4684,
|
| 159 |
"step": 120
|
| 160 |
},
|
| 161 |
{
|
| 162 |
+
"epoch": 0.21,
|
| 163 |
+
"learning_rate": 1.4285714285714287e-05,
|
| 164 |
+
"loss": 1.5267,
|
| 165 |
"step": 125
|
| 166 |
},
|
| 167 |
{
|
| 168 |
+
"epoch": 0.22,
|
| 169 |
+
"learning_rate": 1.4857142857142858e-05,
|
| 170 |
+
"loss": 1.4454,
|
| 171 |
"step": 130
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 0.23,
|
| 175 |
+
"learning_rate": 1.542857142857143e-05,
|
| 176 |
+
"loss": 1.4974,
|
| 177 |
"step": 135
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 0.24,
|
| 181 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 182 |
+
"loss": 1.5087,
|
| 183 |
"step": 140
|
| 184 |
},
|
| 185 |
{
|
| 186 |
+
"epoch": 0.25,
|
| 187 |
+
"learning_rate": 1.6571428571428574e-05,
|
| 188 |
+
"loss": 1.5172,
|
| 189 |
"step": 145
|
| 190 |
},
|
| 191 |
{
|
| 192 |
+
"epoch": 0.26,
|
| 193 |
+
"learning_rate": 1.7142857142857142e-05,
|
| 194 |
+
"loss": 1.5034,
|
| 195 |
"step": 150
|
| 196 |
},
|
| 197 |
{
|
| 198 |
+
"epoch": 0.27,
|
| 199 |
+
"learning_rate": 1.7714285714285717e-05,
|
| 200 |
+
"loss": 1.4841,
|
| 201 |
"step": 155
|
| 202 |
},
|
| 203 |
{
|
| 204 |
+
"epoch": 0.27,
|
| 205 |
+
"learning_rate": 1.8285714285714288e-05,
|
| 206 |
+
"loss": 1.5138,
|
| 207 |
"step": 160
|
| 208 |
},
|
| 209 |
{
|
| 210 |
+
"epoch": 0.28,
|
| 211 |
+
"learning_rate": 1.885714285714286e-05,
|
| 212 |
+
"loss": 1.5664,
|
| 213 |
"step": 165
|
| 214 |
},
|
| 215 |
{
|
| 216 |
+
"epoch": 0.29,
|
| 217 |
+
"learning_rate": 1.942857142857143e-05,
|
| 218 |
+
"loss": 1.5134,
|
| 219 |
"step": 170
|
| 220 |
},
|
| 221 |
{
|
| 222 |
+
"epoch": 0.3,
|
| 223 |
+
"learning_rate": 2e-05,
|
| 224 |
+
"loss": 1.5214,
|
| 225 |
"step": 175
|
| 226 |
},
|
| 227 |
{
|
| 228 |
+
"epoch": 0.31,
|
| 229 |
+
"learning_rate": 1.9999500133801672e-05,
|
| 230 |
+
"loss": 1.502,
|
| 231 |
"step": 180
|
| 232 |
},
|
| 233 |
{
|
| 234 |
+
"epoch": 0.32,
|
| 235 |
+
"learning_rate": 1.9998000585179915e-05,
|
| 236 |
+
"loss": 1.4694,
|
| 237 |
"step": 185
|
| 238 |
},
|
| 239 |
{
|
| 240 |
+
"epoch": 0.33,
|
| 241 |
+
"learning_rate": 1.9995501504049474e-05,
|
| 242 |
+
"loss": 1.5153,
|
| 243 |
"step": 190
|
| 244 |
},
|
| 245 |
{
|
| 246 |
+
"epoch": 0.34,
|
| 247 |
+
"learning_rate": 1.9992003140251582e-05,
|
| 248 |
+
"loss": 1.5056,
|
| 249 |
"step": 195
|
| 250 |
},
|
| 251 |
{
|
| 252 |
+
"epoch": 0.34,
|
| 253 |
+
"learning_rate": 1.9987505843529004e-05,
|
| 254 |
+
"loss": 1.4324,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 255 |
"step": 200
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 0.35,
|
| 259 |
+
"learning_rate": 1.998201006349106e-05,
|
| 260 |
+
"loss": 1.4889,
|
| 261 |
"step": 205
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 0.36,
|
| 265 |
+
"learning_rate": 1.997551634956868e-05,
|
| 266 |
+
"loss": 1.5285,
|
| 267 |
"step": 210
|
| 268 |
},
|
| 269 |
{
|
| 270 |
+
"epoch": 0.37,
|
| 271 |
+
"learning_rate": 1.9968025350959497e-05,
|
| 272 |
+
"loss": 1.5191,
|
| 273 |
"step": 215
|
| 274 |
},
|
| 275 |
{
|
| 276 |
+
"epoch": 0.38,
|
| 277 |
+
"learning_rate": 1.9959537816562897e-05,
|
| 278 |
+
"loss": 1.5196,
|
| 279 |
"step": 220
|
| 280 |
},
|
| 281 |
{
|
| 282 |
+
"epoch": 0.39,
|
| 283 |
+
"learning_rate": 1.9950054594905196e-05,
|
| 284 |
+
"loss": 1.5304,
|
| 285 |
"step": 225
|
| 286 |
},
|
| 287 |
{
|
| 288 |
+
"epoch": 0.4,
|
| 289 |
+
"learning_rate": 1.9939576634054782e-05,
|
| 290 |
+
"loss": 1.4943,
|
| 291 |
"step": 230
|
| 292 |
},
|
| 293 |
{
|
| 294 |
+
"epoch": 0.4,
|
| 295 |
+
"learning_rate": 1.992810498152735e-05,
|
| 296 |
+
"loss": 1.5335,
|
| 297 |
"step": 235
|
| 298 |
},
|
| 299 |
{
|
| 300 |
+
"epoch": 0.41,
|
| 301 |
+
"learning_rate": 1.9915640784181163e-05,
|
| 302 |
+
"loss": 1.4391,
|
| 303 |
"step": 240
|
| 304 |
},
|
| 305 |
{
|
| 306 |
+
"epoch": 0.42,
|
| 307 |
+
"learning_rate": 1.990218528810242e-05,
|
| 308 |
+
"loss": 1.494,
|
| 309 |
"step": 245
|
| 310 |
},
|
| 311 |
{
|
| 312 |
+
"epoch": 0.43,
|
| 313 |
+
"learning_rate": 1.988773983848064e-05,
|
| 314 |
+
"loss": 1.4153,
|
| 315 |
"step": 250
|
| 316 |
},
|
| 317 |
{
|
| 318 |
+
"epoch": 0.44,
|
| 319 |
+
"learning_rate": 1.9872305879474234e-05,
|
| 320 |
+
"loss": 1.4412,
|
| 321 |
"step": 255
|
| 322 |
},
|
| 323 |
{
|
| 324 |
+
"epoch": 0.45,
|
| 325 |
+
"learning_rate": 1.9855884954066077e-05,
|
| 326 |
+
"loss": 1.4328,
|
| 327 |
"step": 260
|
| 328 |
},
|
| 329 |
{
|
| 330 |
+
"epoch": 0.46,
|
| 331 |
+
"learning_rate": 1.983847870390928e-05,
|
| 332 |
+
"loss": 1.4412,
|
| 333 |
"step": 265
|
| 334 |
},
|
| 335 |
{
|
| 336 |
+
"epoch": 0.46,
|
| 337 |
+
"learning_rate": 1.9820088869163066e-05,
|
| 338 |
+
"loss": 1.4981,
|
| 339 |
"step": 270
|
| 340 |
},
|
| 341 |
{
|
| 342 |
+
"epoch": 0.47,
|
| 343 |
+
"learning_rate": 1.9800717288318792e-05,
|
| 344 |
+
"loss": 1.4064,
|
| 345 |
"step": 275
|
| 346 |
},
|
| 347 |
{
|
| 348 |
+
"epoch": 0.48,
|
| 349 |
+
"learning_rate": 1.9780365898016147e-05,
|
| 350 |
+
"loss": 1.4006,
|
| 351 |
"step": 280
|
| 352 |
},
|
| 353 |
{
|
| 354 |
+
"epoch": 0.49,
|
| 355 |
+
"learning_rate": 1.9759036732849552e-05,
|
| 356 |
+
"loss": 1.4635,
|
| 357 |
"step": 285
|
| 358 |
},
|
| 359 |
{
|
| 360 |
+
"epoch": 0.5,
|
| 361 |
+
"learning_rate": 1.9736731925164755e-05,
|
| 362 |
+
"loss": 1.4569,
|
| 363 |
"step": 290
|
| 364 |
},
|
| 365 |
{
|
| 366 |
+
"epoch": 0.51,
|
| 367 |
+
"learning_rate": 1.971345370484563e-05,
|
| 368 |
+
"loss": 1.5035,
|
| 369 |
"step": 295
|
| 370 |
},
|
| 371 |
{
|
| 372 |
+
"epoch": 0.52,
|
| 373 |
+
"learning_rate": 1.9689204399091285e-05,
|
| 374 |
+
"loss": 1.3976,
|
| 375 |
"step": 300
|
| 376 |
},
|
| 377 |
{
|
| 378 |
+
"epoch": 0.52,
|
| 379 |
+
"learning_rate": 1.9663986432183374e-05,
|
| 380 |
+
"loss": 1.4406,
|
| 381 |
"step": 305
|
| 382 |
},
|
| 383 |
{
|
| 384 |
+
"epoch": 0.53,
|
| 385 |
+
"learning_rate": 1.963780232524374e-05,
|
| 386 |
+
"loss": 1.4725,
|
| 387 |
"step": 310
|
| 388 |
},
|
| 389 |
{
|
| 390 |
+
"epoch": 0.54,
|
| 391 |
+
"learning_rate": 1.961065469598239e-05,
|
| 392 |
+
"loss": 1.4824,
|
| 393 |
"step": 315
|
| 394 |
},
|
| 395 |
{
|
| 396 |
+
"epoch": 0.55,
|
| 397 |
+
"learning_rate": 1.958254625843577e-05,
|
| 398 |
+
"loss": 1.5003,
|
| 399 |
"step": 320
|
| 400 |
},
|
| 401 |
{
|
| 402 |
+
"epoch": 0.56,
|
| 403 |
+
"learning_rate": 1.9553479822695435e-05,
|
| 404 |
+
"loss": 1.4694,
|
| 405 |
"step": 325
|
| 406 |
},
|
| 407 |
{
|
| 408 |
+
"epoch": 0.57,
|
| 409 |
+
"learning_rate": 1.952345829462714e-05,
|
| 410 |
+
"loss": 1.481,
|
| 411 |
"step": 330
|
| 412 |
},
|
| 413 |
{
|
| 414 |
+
"epoch": 0.58,
|
| 415 |
+
"learning_rate": 1.9492484675580302e-05,
|
| 416 |
+
"loss": 1.4117,
|
| 417 |
"step": 335
|
| 418 |
},
|
| 419 |
{
|
| 420 |
+
"epoch": 0.58,
|
| 421 |
+
"learning_rate": 1.9460562062087963e-05,
|
| 422 |
+
"loss": 1.4311,
|
| 423 |
"step": 340
|
| 424 |
},
|
| 425 |
{
|
| 426 |
+
"epoch": 0.59,
|
| 427 |
+
"learning_rate": 1.942769364555721e-05,
|
| 428 |
+
"loss": 1.43,
|
| 429 |
"step": 345
|
| 430 |
},
|
| 431 |
{
|
| 432 |
+
"epoch": 0.6,
|
| 433 |
+
"learning_rate": 1.939388271195013e-05,
|
| 434 |
+
"loss": 1.5002,
|
| 435 |
"step": 350
|
| 436 |
},
|
| 437 |
{
|
| 438 |
+
"epoch": 0.61,
|
| 439 |
+
"learning_rate": 1.935913264145529e-05,
|
| 440 |
+
"loss": 1.4256,
|
| 441 |
"step": 355
|
| 442 |
},
|
| 443 |
{
|
| 444 |
+
"epoch": 0.62,
|
| 445 |
+
"learning_rate": 1.932344690814981e-05,
|
| 446 |
+
"loss": 1.4574,
|
| 447 |
"step": 360
|
| 448 |
},
|
| 449 |
{
|
| 450 |
+
"epoch": 0.63,
|
| 451 |
+
"learning_rate": 1.928682907965207e-05,
|
| 452 |
+
"loss": 1.4027,
|
| 453 |
"step": 365
|
| 454 |
},
|
| 455 |
{
|
| 456 |
+
"epoch": 0.64,
|
| 457 |
+
"learning_rate": 1.924928281676501e-05,
|
| 458 |
+
"loss": 1.3318,
|
| 459 |
"step": 370
|
| 460 |
},
|
| 461 |
{
|
| 462 |
+
"epoch": 0.64,
|
| 463 |
+
"learning_rate": 1.9210811873110164e-05,
|
| 464 |
+
"loss": 1.4903,
|
| 465 |
"step": 375
|
| 466 |
},
|
| 467 |
{
|
| 468 |
+
"epoch": 0.65,
|
| 469 |
+
"learning_rate": 1.9171420094752404e-05,
|
| 470 |
+
"loss": 1.4705,
|
| 471 |
"step": 380
|
| 472 |
},
|
| 473 |
{
|
| 474 |
+
"epoch": 0.66,
|
| 475 |
+
"learning_rate": 1.9131111419815428e-05,
|
| 476 |
+
"loss": 1.4515,
|
| 477 |
"step": 385
|
| 478 |
},
|
| 479 |
{
|
| 480 |
+
"epoch": 0.67,
|
| 481 |
+
"learning_rate": 1.9089889878088063e-05,
|
| 482 |
+
"loss": 1.3972,
|
| 483 |
"step": 390
|
| 484 |
},
|
| 485 |
{
|
| 486 |
+
"epoch": 0.68,
|
| 487 |
+
"learning_rate": 1.904775959062137e-05,
|
| 488 |
+
"loss": 1.4329,
|
| 489 |
"step": 395
|
| 490 |
},
|
| 491 |
{
|
| 492 |
+
"epoch": 0.69,
|
| 493 |
+
"learning_rate": 1.9004724769316682e-05,
|
| 494 |
+
"loss": 1.4771,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 495 |
"step": 400
|
| 496 |
},
|
| 497 |
{
|
| 498 |
+
"epoch": 0.7,
|
| 499 |
+
"learning_rate": 1.8960789716504502e-05,
|
| 500 |
+
"loss": 1.4283,
|
| 501 |
"step": 405
|
| 502 |
},
|
| 503 |
{
|
| 504 |
+
"epoch": 0.7,
|
| 505 |
+
"learning_rate": 1.8915958824514388e-05,
|
| 506 |
+
"loss": 1.4044,
|
| 507 |
"step": 410
|
| 508 |
},
|
| 509 |
{
|
| 510 |
+
"epoch": 0.71,
|
| 511 |
+
"learning_rate": 1.887023657523586e-05,
|
| 512 |
+
"loss": 1.4577,
|
| 513 |
"step": 415
|
| 514 |
},
|
| 515 |
{
|
| 516 |
+
"epoch": 0.72,
|
| 517 |
+
"learning_rate": 1.8823627539670296e-05,
|
| 518 |
+
"loss": 1.4222,
|
| 519 |
"step": 420
|
| 520 |
},
|
| 521 |
{
|
| 522 |
+
"epoch": 0.73,
|
| 523 |
+
"learning_rate": 1.877613637747398e-05,
|
| 524 |
+
"loss": 1.4151,
|
| 525 |
"step": 425
|
| 526 |
},
|
| 527 |
{
|
| 528 |
+
"epoch": 0.74,
|
| 529 |
+
"learning_rate": 1.8727767836492256e-05,
|
| 530 |
+
"loss": 1.425,
|
| 531 |
"step": 430
|
| 532 |
},
|
| 533 |
{
|
| 534 |
+
"epoch": 0.75,
|
| 535 |
+
"learning_rate": 1.8678526752284857e-05,
|
| 536 |
+
"loss": 1.3253,
|
| 537 |
"step": 435
|
| 538 |
},
|
| 539 |
{
|
| 540 |
+
"epoch": 0.76,
|
| 541 |
+
"learning_rate": 1.8628418047642505e-05,
|
| 542 |
+
"loss": 1.3933,
|
| 543 |
"step": 440
|
| 544 |
},
|
| 545 |
{
|
| 546 |
+
"epoch": 0.76,
|
| 547 |
+
"learning_rate": 1.8577446732094732e-05,
|
| 548 |
+
"loss": 1.4658,
|
| 549 |
"step": 445
|
| 550 |
},
|
| 551 |
{
|
| 552 |
+
"epoch": 0.77,
|
| 553 |
+
"learning_rate": 1.8525617901409082e-05,
|
| 554 |
+
"loss": 1.4172,
|
| 555 |
"step": 450
|
| 556 |
},
|
| 557 |
{
|
| 558 |
+
"epoch": 0.78,
|
| 559 |
+
"learning_rate": 1.8472936737081673e-05,
|
| 560 |
+
"loss": 1.4866,
|
| 561 |
"step": 455
|
| 562 |
},
|
| 563 |
{
|
| 564 |
+
"epoch": 0.79,
|
| 565 |
+
"learning_rate": 1.841940850581917e-05,
|
| 566 |
+
"loss": 1.4293,
|
| 567 |
"step": 460
|
| 568 |
},
|
| 569 |
{
|
| 570 |
+
"epoch": 0.8,
|
| 571 |
+
"learning_rate": 1.8365038559012263e-05,
|
| 572 |
+
"loss": 1.4085,
|
| 573 |
"step": 465
|
| 574 |
},
|
| 575 |
{
|
| 576 |
+
"epoch": 0.81,
|
| 577 |
+
"learning_rate": 1.8309832332200677e-05,
|
| 578 |
+
"loss": 1.421,
|
| 579 |
"step": 470
|
| 580 |
},
|
| 581 |
{
|
| 582 |
+
"epoch": 0.82,
|
| 583 |
+
"learning_rate": 1.8253795344529758e-05,
|
| 584 |
+
"loss": 1.3329,
|
| 585 |
"step": 475
|
| 586 |
},
|
| 587 |
{
|
| 588 |
+
"epoch": 0.82,
|
| 589 |
+
"learning_rate": 1.81969331981987e-05,
|
| 590 |
+
"loss": 1.3832,
|
| 591 |
"step": 480
|
| 592 |
},
|
| 593 |
{
|
| 594 |
+
"epoch": 0.83,
|
| 595 |
+
"learning_rate": 1.813925157790049e-05,
|
| 596 |
+
"loss": 1.2692,
|
| 597 |
"step": 485
|
| 598 |
},
|
| 599 |
{
|
| 600 |
+
"epoch": 0.84,
|
| 601 |
+
"learning_rate": 1.8080756250253575e-05,
|
| 602 |
+
"loss": 1.4182,
|
| 603 |
"step": 490
|
| 604 |
},
|
| 605 |
{
|
| 606 |
+
"epoch": 0.85,
|
| 607 |
+
"learning_rate": 1.802145306322537e-05,
|
| 608 |
+
"loss": 1.3419,
|
| 609 |
"step": 495
|
| 610 |
},
|
| 611 |
{
|
| 612 |
+
"epoch": 0.86,
|
| 613 |
+
"learning_rate": 1.79613479455476e-05,
|
| 614 |
+
"loss": 1.356,
|
| 615 |
"step": 500
|
| 616 |
},
|
| 617 |
{
|
| 618 |
+
"epoch": 0.87,
|
| 619 |
+
"learning_rate": 1.7900446906123604e-05,
|
| 620 |
+
"loss": 1.3546,
|
| 621 |
"step": 505
|
| 622 |
},
|
| 623 |
{
|
| 624 |
+
"epoch": 0.88,
|
| 625 |
+
"learning_rate": 1.7838756033427585e-05,
|
| 626 |
+
"loss": 1.3617,
|
| 627 |
"step": 510
|
| 628 |
},
|
| 629 |
{
|
| 630 |
+
"epoch": 0.88,
|
| 631 |
+
"learning_rate": 1.7776281494895956e-05,
|
| 632 |
+
"loss": 1.3494,
|
| 633 |
"step": 515
|
| 634 |
},
|
| 635 |
{
|
| 636 |
+
"epoch": 0.89,
|
| 637 |
+
"learning_rate": 1.7713029536310722e-05,
|
| 638 |
+
"loss": 1.382,
|
| 639 |
"step": 520
|
| 640 |
},
|
| 641 |
{
|
| 642 |
+
"epoch": 0.9,
|
| 643 |
+
"learning_rate": 1.76490064811751e-05,
|
| 644 |
+
"loss": 1.3565,
|
| 645 |
"step": 525
|
| 646 |
},
|
| 647 |
{
|
| 648 |
+
"epoch": 0.91,
|
| 649 |
+
"learning_rate": 1.7584218730081325e-05,
|
| 650 |
+
"loss": 1.3842,
|
| 651 |
"step": 530
|
| 652 |
},
|
| 653 |
{
|
| 654 |
+
"epoch": 0.92,
|
| 655 |
+
"learning_rate": 1.7518672760070764e-05,
|
| 656 |
+
"loss": 1.369,
|
| 657 |
"step": 535
|
| 658 |
},
|
| 659 |
{
|
| 660 |
+
"epoch": 0.93,
|
| 661 |
+
"learning_rate": 1.7452375123986386e-05,
|
| 662 |
+
"loss": 1.3536,
|
| 663 |
"step": 540
|
| 664 |
},
|
| 665 |
{
|
| 666 |
+
"epoch": 0.94,
|
| 667 |
+
"learning_rate": 1.7385332449817655e-05,
|
| 668 |
+
"loss": 1.3606,
|
| 669 |
"step": 545
|
| 670 |
},
|
| 671 |
{
|
| 672 |
+
"epoch": 0.95,
|
| 673 |
+
"learning_rate": 1.73175514400379e-05,
|
| 674 |
+
"loss": 1.3455,
|
| 675 |
"step": 550
|
| 676 |
},
|
| 677 |
{
|
| 678 |
+
"epoch": 0.95,
|
| 679 |
+
"learning_rate": 1.7249038870934263e-05,
|
| 680 |
+
"loss": 1.3637,
|
| 681 |
"step": 555
|
| 682 |
},
|
| 683 |
{
|
| 684 |
+
"epoch": 0.96,
|
| 685 |
+
"learning_rate": 1.717980159193023e-05,
|
| 686 |
+
"loss": 1.3794,
|
| 687 |
"step": 560
|
| 688 |
},
|
| 689 |
{
|
| 690 |
+
"epoch": 0.97,
|
| 691 |
+
"learning_rate": 1.7109846524900885e-05,
|
| 692 |
+
"loss": 1.3189,
|
| 693 |
"step": 565
|
| 694 |
},
|
| 695 |
{
|
| 696 |
+
"epoch": 0.98,
|
| 697 |
+
"learning_rate": 1.7039180663480917e-05,
|
| 698 |
+
"loss": 1.3252,
|
| 699 |
"step": 570
|
| 700 |
},
|
| 701 |
{
|
| 702 |
+
"epoch": 0.99,
|
| 703 |
+
"learning_rate": 1.6967811072365423e-05,
|
| 704 |
+
"loss": 1.342,
|
| 705 |
"step": 575
|
| 706 |
},
|
| 707 |
{
|
| 708 |
+
"epoch": 1.0,
|
| 709 |
+
"learning_rate": 1.689574488660364e-05,
|
| 710 |
+
"loss": 1.3497,
|
| 711 |
"step": 580
|
| 712 |
},
|
| 713 |
{
|
| 714 |
+
"epoch": 1.0,
|
| 715 |
+
"eval_loss": 1.4483790397644043,
|
| 716 |
+
"eval_runtime": 133.8837,
|
| 717 |
+
"eval_samples_per_second": 10.18,
|
| 718 |
+
"eval_steps_per_second": 0.112,
|
| 719 |
+
"step": 582
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 1.01,
|
| 723 |
+
"learning_rate": 1.682298931088563e-05,
|
| 724 |
+
"loss": 1.2855,
|
| 725 |
"step": 585
|
| 726 |
},
|
| 727 |
{
|
| 728 |
+
"epoch": 1.01,
|
| 729 |
+
"learning_rate": 1.6749551618822006e-05,
|
| 730 |
+
"loss": 1.2097,
|
| 731 |
"step": 590
|
| 732 |
},
|
| 733 |
{
|
| 734 |
+
"epoch": 1.02,
|
| 735 |
+
"learning_rate": 1.6675439152216747e-05,
|
| 736 |
+
"loss": 1.3221,
|
| 737 |
"step": 595
|
| 738 |
},
|
| 739 |
{
|
| 740 |
+
"epoch": 1.03,
|
| 741 |
+
"learning_rate": 1.6600659320333247e-05,
|
| 742 |
+
"loss": 1.229,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 743 |
"step": 600
|
| 744 |
},
|
| 745 |
{
|
| 746 |
+
"epoch": 1.04,
|
| 747 |
+
"learning_rate": 1.652521959915356e-05,
|
| 748 |
+
"loss": 1.1311,
|
| 749 |
"step": 605
|
| 750 |
},
|
| 751 |
{
|
| 752 |
+
"epoch": 1.05,
|
| 753 |
+
"learning_rate": 1.6449127530631005e-05,
|
| 754 |
+
"loss": 1.2322,
|
| 755 |
"step": 610
|
| 756 |
},
|
| 757 |
{
|
| 758 |
+
"epoch": 1.06,
|
| 759 |
+
"learning_rate": 1.6372390721936198e-05,
|
| 760 |
+
"loss": 1.2631,
|
| 761 |
"step": 615
|
| 762 |
},
|
| 763 |
{
|
| 764 |
+
"epoch": 1.07,
|
| 765 |
+
"learning_rate": 1.6295016844696494e-05,
|
| 766 |
+
"loss": 1.1868,
|
| 767 |
"step": 620
|
| 768 |
},
|
| 769 |
{
|
| 770 |
+
"epoch": 1.07,
|
| 771 |
+
"learning_rate": 1.621701363422907e-05,
|
| 772 |
+
"loss": 1.1462,
|
| 773 |
"step": 625
|
| 774 |
},
|
| 775 |
{
|
| 776 |
+
"epoch": 1.08,
|
| 777 |
+
"learning_rate": 1.6138388888767587e-05,
|
| 778 |
+
"loss": 1.0917,
|
| 779 |
"step": 630
|
| 780 |
},
|
| 781 |
{
|
| 782 |
+
"epoch": 1.09,
|
| 783 |
+
"learning_rate": 1.6059150468682558e-05,
|
| 784 |
+
"loss": 1.0819,
|
| 785 |
"step": 635
|
| 786 |
},
|
| 787 |
{
|
| 788 |
+
"epoch": 1.1,
|
| 789 |
+
"learning_rate": 1.5979306295695547e-05,
|
| 790 |
+
"loss": 1.1196,
|
| 791 |
"step": 640
|
| 792 |
},
|
| 793 |
{
|
| 794 |
+
"epoch": 1.11,
|
| 795 |
+
"learning_rate": 1.58988643520872e-05,
|
| 796 |
+
"loss": 1.0547,
|
| 797 |
"step": 645
|
| 798 |
},
|
| 799 |
{
|
| 800 |
+
"epoch": 1.12,
|
| 801 |
+
"learning_rate": 1.5817832679899222e-05,
|
| 802 |
+
"loss": 1.0307,
|
| 803 |
"step": 650
|
| 804 |
},
|
| 805 |
{
|
| 806 |
+
"epoch": 1.13,
|
| 807 |
+
"learning_rate": 1.5736219380130395e-05,
|
| 808 |
+
"loss": 0.9733,
|
| 809 |
"step": 655
|
| 810 |
},
|
| 811 |
{
|
| 812 |
+
"epoch": 1.13,
|
| 813 |
+
"learning_rate": 1.5654032611926698e-05,
|
| 814 |
+
"loss": 1.0215,
|
| 815 |
"step": 660
|
| 816 |
},
|
| 817 |
{
|
| 818 |
+
"epoch": 1.14,
|
| 819 |
+
"learning_rate": 1.5571280591765613e-05,
|
| 820 |
+
"loss": 0.9941,
|
| 821 |
"step": 665
|
| 822 |
},
|
| 823 |
{
|
| 824 |
+
"epoch": 1.15,
|
| 825 |
+
"learning_rate": 1.548797159263467e-05,
|
| 826 |
+
"loss": 0.9208,
|
| 827 |
"step": 670
|
| 828 |
},
|
| 829 |
{
|
| 830 |
+
"epoch": 1.16,
|
| 831 |
+
"learning_rate": 1.540411394320442e-05,
|
| 832 |
+
"loss": 0.9202,
|
| 833 |
"step": 675
|
| 834 |
},
|
| 835 |
{
|
| 836 |
+
"epoch": 1.17,
|
| 837 |
+
"learning_rate": 1.531971602699574e-05,
|
| 838 |
+
"loss": 0.9048,
|
| 839 |
"step": 680
|
| 840 |
},
|
| 841 |
{
|
| 842 |
+
"epoch": 1.18,
|
| 843 |
+
"learning_rate": 1.5234786281541736e-05,
|
| 844 |
+
"loss": 0.9502,
|
| 845 |
"step": 685
|
| 846 |
},
|
| 847 |
{
|
| 848 |
+
"epoch": 1.19,
|
| 849 |
+
"learning_rate": 1.5149333197544203e-05,
|
| 850 |
+
"loss": 0.8838,
|
| 851 |
"step": 690
|
| 852 |
},
|
| 853 |
{
|
| 854 |
+
"epoch": 1.19,
|
| 855 |
+
"learning_rate": 1.5063365318024791e-05,
|
| 856 |
+
"loss": 0.9124,
|
| 857 |
"step": 695
|
| 858 |
},
|
| 859 |
{
|
| 860 |
+
"epoch": 1.2,
|
| 861 |
+
"learning_rate": 1.497689123747092e-05,
|
| 862 |
+
"loss": 0.8446,
|
| 863 |
"step": 700
|
| 864 |
},
|
| 865 |
{
|
| 866 |
+
"epoch": 1.21,
|
| 867 |
+
"learning_rate": 1.4889919600976573e-05,
|
| 868 |
+
"loss": 0.8778,
|
| 869 |
"step": 705
|
| 870 |
},
|
| 871 |
{
|
| 872 |
+
"epoch": 1.22,
|
| 873 |
+
"learning_rate": 1.4802459103378008e-05,
|
| 874 |
+
"loss": 0.7858,
|
| 875 |
"step": 710
|
| 876 |
},
|
| 877 |
{
|
| 878 |
+
"epoch": 1.23,
|
| 879 |
+
"learning_rate": 1.4714518488384512e-05,
|
| 880 |
+
"loss": 0.8087,
|
| 881 |
"step": 715
|
| 882 |
},
|
| 883 |
{
|
| 884 |
+
"epoch": 1.24,
|
| 885 |
+
"learning_rate": 1.4626106547704262e-05,
|
| 886 |
+
"loss": 0.8072,
|
| 887 |
"step": 720
|
| 888 |
},
|
| 889 |
{
|
| 890 |
+
"epoch": 1.25,
|
| 891 |
+
"learning_rate": 1.4537232120165401e-05,
|
| 892 |
+
"loss": 0.7525,
|
| 893 |
"step": 725
|
| 894 |
},
|
| 895 |
{
|
| 896 |
+
"epoch": 1.25,
|
| 897 |
+
"learning_rate": 1.4447904090832366e-05,
|
| 898 |
+
"loss": 0.8278,
|
| 899 |
"step": 730
|
| 900 |
},
|
| 901 |
{
|
| 902 |
+
"epoch": 1.26,
|
| 903 |
+
"learning_rate": 1.4358131390117645e-05,
|
| 904 |
+
"loss": 0.8115,
|
| 905 |
"step": 735
|
| 906 |
},
|
| 907 |
{
|
| 908 |
+
"epoch": 1.27,
|
| 909 |
+
"learning_rate": 1.4267922992888961e-05,
|
| 910 |
+
"loss": 0.7743,
|
| 911 |
"step": 740
|
| 912 |
},
|
| 913 |
{
|
| 914 |
+
"epoch": 1.28,
|
| 915 |
+
"learning_rate": 1.4177287917572032e-05,
|
| 916 |
+
"loss": 0.8078,
|
| 917 |
"step": 745
|
| 918 |
},
|
| 919 |
{
|
| 920 |
+
"epoch": 1.29,
|
| 921 |
+
"learning_rate": 1.4086235225248966e-05,
|
| 922 |
+
"loss": 0.7173,
|
| 923 |
"step": 750
|
| 924 |
},
|
| 925 |
{
|
| 926 |
+
"epoch": 1.3,
|
| 927 |
+
"learning_rate": 1.3994774018752389e-05,
|
| 928 |
+
"loss": 0.7239,
|
| 929 |
"step": 755
|
| 930 |
},
|
| 931 |
{
|
| 932 |
+
"epoch": 1.31,
|
| 933 |
+
"learning_rate": 1.3902913441755422e-05,
|
| 934 |
+
"loss": 0.704,
|
| 935 |
"step": 760
|
| 936 |
},
|
| 937 |
{
|
| 938 |
+
"epoch": 1.31,
|
| 939 |
+
"learning_rate": 1.3810662677857548e-05,
|
| 940 |
+
"loss": 0.7421,
|
| 941 |
"step": 765
|
| 942 |
},
|
| 943 |
{
|
| 944 |
+
"epoch": 1.32,
|
| 945 |
+
"learning_rate": 1.3718030949666488e-05,
|
| 946 |
+
"loss": 0.7604,
|
| 947 |
"step": 770
|
| 948 |
},
|
| 949 |
{
|
| 950 |
+
"epoch": 1.33,
|
| 951 |
+
"learning_rate": 1.3625027517876216e-05,
|
| 952 |
+
"loss": 0.7587,
|
| 953 |
"step": 775
|
| 954 |
},
|
| 955 |
{
|
| 956 |
+
"epoch": 1.34,
|
| 957 |
+
"learning_rate": 1.3531661680341098e-05,
|
| 958 |
+
"loss": 0.7139,
|
| 959 |
"step": 780
|
| 960 |
},
|
| 961 |
{
|
| 962 |
+
"epoch": 1.35,
|
| 963 |
+
"learning_rate": 1.3437942771146388e-05,
|
| 964 |
+
"loss": 0.715,
|
| 965 |
"step": 785
|
| 966 |
},
|
| 967 |
{
|
| 968 |
+
"epoch": 1.36,
|
| 969 |
+
"learning_rate": 1.3343880159675062e-05,
|
| 970 |
+
"loss": 0.7455,
|
| 971 |
"step": 790
|
| 972 |
},
|
| 973 |
{
|
| 974 |
+
"epoch": 1.37,
|
| 975 |
+
"learning_rate": 1.3249483249671117e-05,
|
| 976 |
+
"loss": 0.7436,
|
| 977 |
"step": 795
|
| 978 |
},
|
| 979 |
{
|
| 980 |
+
"epoch": 1.37,
|
| 981 |
+
"learning_rate": 1.3154761478299461e-05,
|
| 982 |
+
"loss": 0.7172,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 983 |
"step": 800
|
| 984 |
},
|
| 985 |
{
|
| 986 |
+
"epoch": 1.38,
|
| 987 |
+
"learning_rate": 1.3059724315202444e-05,
|
| 988 |
+
"loss": 0.741,
|
| 989 |
"step": 805
|
| 990 |
},
|
| 991 |
{
|
| 992 |
+
"epoch": 1.39,
|
| 993 |
+
"learning_rate": 1.296438126155315e-05,
|
| 994 |
+
"loss": 0.6807,
|
| 995 |
"step": 810
|
| 996 |
},
|
| 997 |
{
|
| 998 |
+
"epoch": 1.4,
|
| 999 |
+
"learning_rate": 1.286874184910553e-05,
|
| 1000 |
+
"loss": 0.7379,
|
| 1001 |
"step": 815
|
| 1002 |
},
|
| 1003 |
{
|
| 1004 |
+
"epoch": 1.41,
|
| 1005 |
+
"learning_rate": 1.2772815639241491e-05,
|
| 1006 |
+
"loss": 0.7086,
|
| 1007 |
"step": 820
|
| 1008 |
},
|
| 1009 |
{
|
| 1010 |
+
"epoch": 1.42,
|
| 1011 |
+
"learning_rate": 1.2676612222014999e-05,
|
| 1012 |
+
"loss": 0.7675,
|
| 1013 |
"step": 825
|
| 1014 |
},
|
| 1015 |
{
|
| 1016 |
+
"epoch": 1.43,
|
| 1017 |
+
"learning_rate": 1.258014121519334e-05,
|
| 1018 |
+
"loss": 0.7193,
|
| 1019 |
"step": 830
|
| 1020 |
},
|
| 1021 |
{
|
| 1022 |
+
"epoch": 1.43,
|
| 1023 |
+
"learning_rate": 1.2483412263295602e-05,
|
| 1024 |
+
"loss": 0.6671,
|
| 1025 |
"step": 835
|
| 1026 |
},
|
| 1027 |
{
|
| 1028 |
+
"epoch": 1.44,
|
| 1029 |
+
"learning_rate": 1.2386435036628474e-05,
|
| 1030 |
+
"loss": 0.6785,
|
| 1031 |
"step": 840
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
+
"epoch": 1.45,
|
| 1035 |
+
"learning_rate": 1.228921923031948e-05,
|
| 1036 |
+
"loss": 0.6683,
|
| 1037 |
"step": 845
|
| 1038 |
},
|
| 1039 |
{
|
| 1040 |
+
"epoch": 1.46,
|
| 1041 |
+
"learning_rate": 1.2191774563347727e-05,
|
| 1042 |
+
"loss": 0.7199,
|
| 1043 |
"step": 850
|
| 1044 |
},
|
| 1045 |
{
|
| 1046 |
+
"epoch": 1.47,
|
| 1047 |
+
"learning_rate": 1.2094110777572257e-05,
|
| 1048 |
+
"loss": 0.7024,
|
| 1049 |
"step": 855
|
| 1050 |
},
|
| 1051 |
{
|
| 1052 |
+
"epoch": 1.48,
|
| 1053 |
+
"learning_rate": 1.1996237636758133e-05,
|
| 1054 |
+
"loss": 0.6914,
|
| 1055 |
"step": 860
|
| 1056 |
},
|
| 1057 |
{
|
| 1058 |
+
"epoch": 1.49,
|
| 1059 |
+
"learning_rate": 1.1898164925600316e-05,
|
| 1060 |
+
"loss": 0.6947,
|
| 1061 |
"step": 865
|
| 1062 |
},
|
| 1063 |
{
|
| 1064 |
+
"epoch": 1.49,
|
| 1065 |
+
"learning_rate": 1.1799902448745466e-05,
|
| 1066 |
+
"loss": 0.6846,
|
| 1067 |
"step": 870
|
| 1068 |
},
|
| 1069 |
{
|
| 1070 |
+
"epoch": 1.5,
|
| 1071 |
+
"learning_rate": 1.1701460029811733e-05,
|
| 1072 |
+
"loss": 0.6788,
|
| 1073 |
"step": 875
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
+
"epoch": 1.51,
|
| 1077 |
+
"learning_rate": 1.1602847510406657e-05,
|
| 1078 |
+
"loss": 0.6986,
|
| 1079 |
"step": 880
|
| 1080 |
},
|
| 1081 |
{
|
| 1082 |
+
"epoch": 1.52,
|
| 1083 |
+
"learning_rate": 1.150407474914327e-05,
|
| 1084 |
+
"loss": 0.7083,
|
| 1085 |
"step": 885
|
| 1086 |
},
|
| 1087 |
{
|
| 1088 |
+
"epoch": 1.53,
|
| 1089 |
+
"learning_rate": 1.1405151620654513e-05,
|
| 1090 |
+
"loss": 0.6188,
|
| 1091 |
"step": 890
|
| 1092 |
},
|
| 1093 |
{
|
| 1094 |
+
"epoch": 1.54,
|
| 1095 |
+
"learning_rate": 1.1306088014606018e-05,
|
| 1096 |
+
"loss": 0.7138,
|
| 1097 |
"step": 895
|
| 1098 |
},
|
| 1099 |
{
|
| 1100 |
+
"epoch": 1.55,
|
| 1101 |
+
"learning_rate": 1.1206893834707415e-05,
|
| 1102 |
+
"loss": 0.6972,
|
| 1103 |
"step": 900
|
| 1104 |
},
|
| 1105 |
{
|
| 1106 |
+
"epoch": 1.55,
|
| 1107 |
+
"learning_rate": 1.110757899772222e-05,
|
| 1108 |
+
"loss": 0.7069,
|
| 1109 |
"step": 905
|
| 1110 |
},
|
| 1111 |
{
|
| 1112 |
+
"epoch": 1.56,
|
| 1113 |
+
"learning_rate": 1.1008153432476436e-05,
|
| 1114 |
+
"loss": 0.7038,
|
| 1115 |
"step": 910
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
+
"epoch": 1.57,
|
| 1119 |
+
"learning_rate": 1.0908627078865928e-05,
|
| 1120 |
+
"loss": 0.6523,
|
| 1121 |
"step": 915
|
| 1122 |
},
|
| 1123 |
{
|
| 1124 |
+
"epoch": 1.58,
|
| 1125 |
+
"learning_rate": 1.0809009886862695e-05,
|
| 1126 |
+
"loss": 0.6411,
|
| 1127 |
"step": 920
|
| 1128 |
},
|
| 1129 |
{
|
| 1130 |
+
"epoch": 1.59,
|
| 1131 |
+
"learning_rate": 1.0709311815520151e-05,
|
| 1132 |
+
"loss": 0.6691,
|
| 1133 |
"step": 925
|
| 1134 |
},
|
| 1135 |
{
|
| 1136 |
+
"epoch": 1.6,
|
| 1137 |
+
"learning_rate": 1.0609542831977478e-05,
|
| 1138 |
+
"loss": 0.6952,
|
| 1139 |
"step": 930
|
| 1140 |
},
|
| 1141 |
{
|
| 1142 |
+
"epoch": 1.61,
|
| 1143 |
+
"learning_rate": 1.0509712910463175e-05,
|
| 1144 |
+
"loss": 0.7437,
|
| 1145 |
"step": 935
|
| 1146 |
},
|
| 1147 |
{
|
| 1148 |
+
"epoch": 1.62,
|
| 1149 |
+
"learning_rate": 1.0409832031297912e-05,
|
| 1150 |
+
"loss": 0.7274,
|
| 1151 |
"step": 940
|
| 1152 |
},
|
| 1153 |
{
|
| 1154 |
+
"epoch": 1.62,
|
| 1155 |
+
"learning_rate": 1.0309910179896762e-05,
|
| 1156 |
+
"loss": 0.6345,
|
| 1157 |
"step": 945
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
+
"epoch": 1.63,
|
| 1161 |
+
"learning_rate": 1.0209957345770924e-05,
|
| 1162 |
+
"loss": 0.6331,
|
| 1163 |
"step": 950
|
| 1164 |
},
|
| 1165 |
{
|
| 1166 |
+
"epoch": 1.64,
|
| 1167 |
+
"learning_rate": 1.0109983521529036e-05,
|
| 1168 |
+
"loss": 0.6683,
|
| 1169 |
"step": 955
|
| 1170 |
},
|
| 1171 |
{
|
| 1172 |
+
"epoch": 1.65,
|
| 1173 |
+
"learning_rate": 1.000999870187819e-05,
|
| 1174 |
+
"loss": 0.6999,
|
| 1175 |
"step": 960
|
| 1176 |
},
|
| 1177 |
{
|
| 1178 |
+
"epoch": 1.66,
|
| 1179 |
+
"learning_rate": 9.910012882624726e-06,
|
| 1180 |
+
"loss": 0.6881,
|
| 1181 |
"step": 965
|
| 1182 |
},
|
| 1183 |
{
|
| 1184 |
+
"epoch": 1.67,
|
| 1185 |
+
"learning_rate": 9.810036059674914e-06,
|
| 1186 |
+
"loss": 0.6735,
|
| 1187 |
"step": 970
|
| 1188 |
},
|
| 1189 |
{
|
| 1190 |
+
"epoch": 1.68,
|
| 1191 |
+
"learning_rate": 9.710078228035635e-06,
|
| 1192 |
+
"loss": 0.6506,
|
| 1193 |
"step": 975
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
+
"epoch": 1.68,
|
| 1197 |
+
"learning_rate": 9.610149380815148e-06,
|
| 1198 |
+
"loss": 0.6896,
|
| 1199 |
"step": 980
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
+
"epoch": 1.69,
|
| 1203 |
+
"learning_rate": 9.510259508224052e-06,
|
| 1204 |
+
"loss": 0.6672,
|
| 1205 |
"step": 985
|
| 1206 |
},
|
| 1207 |
{
|
| 1208 |
+
"epoch": 1.7,
|
| 1209 |
+
"learning_rate": 9.41041859657651e-06,
|
| 1210 |
+
"loss": 0.6851,
|
| 1211 |
"step": 990
|
| 1212 |
},
|
| 1213 |
{
|
| 1214 |
+
"epoch": 1.71,
|
| 1215 |
+
"learning_rate": 9.310636627291915e-06,
|
| 1216 |
+
"loss": 0.6618,
|
| 1217 |
"step": 995
|
| 1218 |
},
|
| 1219 |
{
|
| 1220 |
+
"epoch": 1.72,
|
| 1221 |
+
"learning_rate": 9.210923575897001e-06,
|
| 1222 |
+
"loss": 0.6249,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1223 |
"step": 1000
|
| 1224 |
},
|
| 1225 |
{
|
| 1226 |
+
"epoch": 1.73,
|
| 1227 |
+
"learning_rate": 9.111289411028545e-06,
|
| 1228 |
+
"loss": 0.631,
|
| 1229 |
"step": 1005
|
| 1230 |
},
|
| 1231 |
{
|
| 1232 |
+
"epoch": 1.74,
|
| 1233 |
+
"learning_rate": 9.0117440934368e-06,
|
| 1234 |
+
"loss": 0.6352,
|
| 1235 |
"step": 1010
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
+
"epoch": 1.74,
|
| 1239 |
+
"learning_rate": 8.912297574989649e-06,
|
| 1240 |
+
"loss": 0.6613,
|
| 1241 |
"step": 1015
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
+
"epoch": 1.75,
|
| 1245 |
+
"learning_rate": 8.812959797677718e-06,
|
| 1246 |
+
"loss": 0.6452,
|
| 1247 |
"step": 1020
|
| 1248 |
},
|
| 1249 |
{
|
| 1250 |
+
"epoch": 1.76,
|
| 1251 |
+
"learning_rate": 8.713740692620432e-06,
|
| 1252 |
+
"loss": 0.6671,
|
| 1253 |
"step": 1025
|
| 1254 |
},
|
| 1255 |
{
|
| 1256 |
+
"epoch": 1.77,
|
| 1257 |
+
"learning_rate": 8.614650179073152e-06,
|
| 1258 |
+
"loss": 0.6833,
|
| 1259 |
"step": 1030
|
| 1260 |
},
|
| 1261 |
{
|
| 1262 |
+
"epoch": 1.78,
|
| 1263 |
+
"learning_rate": 8.515698163435542e-06,
|
| 1264 |
+
"loss": 0.7135,
|
| 1265 |
"step": 1035
|
| 1266 |
},
|
| 1267 |
{
|
| 1268 |
+
"epoch": 1.79,
|
| 1269 |
+
"learning_rate": 8.41689453826118e-06,
|
| 1270 |
+
"loss": 0.6784,
|
| 1271 |
"step": 1040
|
| 1272 |
},
|
| 1273 |
{
|
| 1274 |
+
"epoch": 1.8,
|
| 1275 |
+
"learning_rate": 8.318249181268556e-06,
|
| 1276 |
+
"loss": 0.6886,
|
| 1277 |
"step": 1045
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
+
"epoch": 1.8,
|
| 1281 |
+
"learning_rate": 8.219771954353597e-06,
|
| 1282 |
+
"loss": 0.6221,
|
| 1283 |
"step": 1050
|
| 1284 |
},
|
| 1285 |
{
|
| 1286 |
+
"epoch": 1.81,
|
| 1287 |
+
"learning_rate": 8.121472702603703e-06,
|
| 1288 |
+
"loss": 0.677,
|
| 1289 |
"step": 1055
|
| 1290 |
},
|
| 1291 |
{
|
| 1292 |
+
"epoch": 1.82,
|
| 1293 |
+
"learning_rate": 8.023361253313533e-06,
|
| 1294 |
+
"loss": 0.631,
|
| 1295 |
"step": 1060
|
| 1296 |
},
|
| 1297 |
{
|
| 1298 |
+
"epoch": 1.83,
|
| 1299 |
+
"learning_rate": 7.925447415002523e-06,
|
| 1300 |
+
"loss": 0.5897,
|
| 1301 |
"step": 1065
|
| 1302 |
},
|
| 1303 |
{
|
| 1304 |
+
"epoch": 1.84,
|
| 1305 |
+
"learning_rate": 7.827740976434287e-06,
|
| 1306 |
+
"loss": 0.6003,
|
| 1307 |
"step": 1070
|
| 1308 |
},
|
| 1309 |
{
|
| 1310 |
+
"epoch": 1.85,
|
| 1311 |
+
"learning_rate": 7.730251705638034e-06,
|
| 1312 |
+
"loss": 0.6457,
|
| 1313 |
"step": 1075
|
| 1314 |
},
|
| 1315 |
{
|
| 1316 |
+
"epoch": 1.86,
|
| 1317 |
+
"learning_rate": 7.632989348931993e-06,
|
| 1318 |
+
"loss": 0.6322,
|
| 1319 |
"step": 1080
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
+
"epoch": 1.86,
|
| 1323 |
+
"learning_rate": 7.535963629949071e-06,
|
| 1324 |
+
"loss": 0.5849,
|
| 1325 |
"step": 1085
|
| 1326 |
},
|
| 1327 |
{
|
| 1328 |
+
"epoch": 1.87,
|
| 1329 |
+
"learning_rate": 7.439184248664713e-06,
|
| 1330 |
+
"loss": 0.6439,
|
| 1331 |
"step": 1090
|
| 1332 |
},
|
| 1333 |
{
|
| 1334 |
+
"epoch": 1.88,
|
| 1335 |
+
"learning_rate": 7.34266088042721e-06,
|
| 1336 |
+
"loss": 0.641,
|
| 1337 |
"step": 1095
|
| 1338 |
},
|
| 1339 |
{
|
| 1340 |
+
"epoch": 1.89,
|
| 1341 |
+
"learning_rate": 7.246403174990379e-06,
|
| 1342 |
+
"loss": 0.6496,
|
| 1343 |
"step": 1100
|
| 1344 |
},
|
| 1345 |
{
|
| 1346 |
+
"epoch": 1.9,
|
| 1347 |
+
"learning_rate": 7.15042075554888e-06,
|
| 1348 |
+
"loss": 0.6245,
|
| 1349 |
"step": 1105
|
| 1350 |
},
|
| 1351 |
{
|
| 1352 |
+
"epoch": 1.91,
|
| 1353 |
+
"learning_rate": 7.054723217776138e-06,
|
| 1354 |
+
"loss": 0.6483,
|
| 1355 |
"step": 1110
|
| 1356 |
},
|
| 1357 |
{
|
| 1358 |
+
"epoch": 1.92,
|
| 1359 |
+
"learning_rate": 6.959320128865026e-06,
|
| 1360 |
+
"loss": 0.6532,
|
| 1361 |
"step": 1115
|
| 1362 |
},
|
| 1363 |
{
|
| 1364 |
+
"epoch": 1.92,
|
| 1365 |
+
"learning_rate": 6.86422102657142e-06,
|
| 1366 |
+
"loss": 0.6116,
|
| 1367 |
"step": 1120
|
| 1368 |
},
|
| 1369 |
{
|
| 1370 |
+
"epoch": 1.93,
|
| 1371 |
+
"learning_rate": 6.76943541826067e-06,
|
| 1372 |
+
"loss": 0.5523,
|
| 1373 |
"step": 1125
|
| 1374 |
},
|
| 1375 |
{
|
| 1376 |
+
"epoch": 1.94,
|
| 1377 |
+
"learning_rate": 6.6749727799571016e-06,
|
| 1378 |
+
"loss": 0.6321,
|
| 1379 |
"step": 1130
|
| 1380 |
},
|
| 1381 |
{
|
| 1382 |
+
"epoch": 1.95,
|
| 1383 |
+
"learning_rate": 6.580842555396706e-06,
|
| 1384 |
+
"loss": 0.6502,
|
| 1385 |
"step": 1135
|
| 1386 |
},
|
| 1387 |
{
|
| 1388 |
+
"epoch": 1.96,
|
| 1389 |
+
"learning_rate": 6.487054155082973e-06,
|
| 1390 |
+
"loss": 0.6254,
|
| 1391 |
"step": 1140
|
| 1392 |
},
|
| 1393 |
{
|
| 1394 |
+
"epoch": 1.97,
|
| 1395 |
+
"learning_rate": 6.393616955346129e-06,
|
| 1396 |
+
"loss": 0.6547,
|
| 1397 |
"step": 1145
|
| 1398 |
},
|
| 1399 |
{
|
| 1400 |
+
"epoch": 1.98,
|
| 1401 |
+
"learning_rate": 6.3005402974057415e-06,
|
| 1402 |
+
"loss": 0.6172,
|
| 1403 |
"step": 1150
|
| 1404 |
},
|
| 1405 |
{
|
| 1406 |
+
"epoch": 1.98,
|
| 1407 |
+
"learning_rate": 6.207833486436832e-06,
|
| 1408 |
+
"loss": 0.6557,
|
| 1409 |
"step": 1155
|
| 1410 |
},
|
| 1411 |
{
|
| 1412 |
+
"epoch": 1.99,
|
| 1413 |
+
"learning_rate": 6.115505790639642e-06,
|
| 1414 |
+
"loss": 0.6709,
|
| 1415 |
"step": 1160
|
| 1416 |
},
|
| 1417 |
{
|
| 1418 |
+
"epoch": 2.0,
|
| 1419 |
+
"eval_loss": 1.5766805410385132,
|
| 1420 |
+
"eval_runtime": 133.3882,
|
| 1421 |
+
"eval_samples_per_second": 10.218,
|
| 1422 |
+
"eval_steps_per_second": 0.112,
|
| 1423 |
+
"step": 1164
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 2.0,
|
| 1427 |
+
"learning_rate": 6.023566440313023e-06,
|
| 1428 |
+
"loss": 0.559,
|
| 1429 |
"step": 1165
|
| 1430 |
},
|
| 1431 |
{
|
| 1432 |
+
"epoch": 2.01,
|
| 1433 |
+
"learning_rate": 5.932024626931689e-06,
|
| 1434 |
+
"loss": 0.5316,
|
| 1435 |
"step": 1170
|
| 1436 |
},
|
| 1437 |
{
|
| 1438 |
+
"epoch": 2.02,
|
| 1439 |
+
"learning_rate": 5.840889502227285e-06,
|
| 1440 |
+
"loss": 0.6037,
|
| 1441 |
"step": 1175
|
| 1442 |
},
|
| 1443 |
{
|
| 1444 |
+
"epoch": 2.03,
|
| 1445 |
+
"learning_rate": 5.750170177273471e-06,
|
| 1446 |
+
"loss": 0.5708,
|
| 1447 |
"step": 1180
|
| 1448 |
},
|
| 1449 |
{
|
| 1450 |
+
"epoch": 2.04,
|
| 1451 |
+
"learning_rate": 5.659875721575065e-06,
|
| 1452 |
+
"loss": 0.5343,
|
| 1453 |
"step": 1185
|
| 1454 |
},
|
| 1455 |
{
|
| 1456 |
+
"epoch": 2.04,
|
| 1457 |
+
"learning_rate": 5.570015162161331e-06,
|
| 1458 |
+
"loss": 0.5318,
|
| 1459 |
"step": 1190
|
| 1460 |
},
|
| 1461 |
{
|
| 1462 |
+
"epoch": 2.05,
|
| 1463 |
+
"learning_rate": 5.480597482683507e-06,
|
| 1464 |
+
"loss": 0.5536,
|
| 1465 |
"step": 1195
|
| 1466 |
},
|
| 1467 |
{
|
| 1468 |
+
"epoch": 2.06,
|
| 1469 |
+
"learning_rate": 5.391631622516697e-06,
|
| 1470 |
+
"loss": 0.5613,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1471 |
"step": 1200
|
| 1472 |
},
|
| 1473 |
{
|
| 1474 |
+
"epoch": 2.07,
|
| 1475 |
+
"learning_rate": 5.3031264758661595e-06,
|
| 1476 |
+
"loss": 0.5205,
|
| 1477 |
"step": 1205
|
| 1478 |
},
|
| 1479 |
{
|
| 1480 |
+
"epoch": 2.08,
|
| 1481 |
+
"learning_rate": 5.2150908908781335e-06,
|
| 1482 |
+
"loss": 0.5141,
|
| 1483 |
"step": 1210
|
| 1484 |
},
|
| 1485 |
{
|
| 1486 |
+
"epoch": 2.09,
|
| 1487 |
+
"learning_rate": 5.1275336687552554e-06,
|
| 1488 |
+
"loss": 0.4753,
|
| 1489 |
"step": 1215
|
| 1490 |
},
|
| 1491 |
{
|
| 1492 |
+
"epoch": 2.1,
|
| 1493 |
+
"learning_rate": 5.040463562876678e-06,
|
| 1494 |
+
"loss": 0.5046,
|
| 1495 |
"step": 1220
|
| 1496 |
},
|
| 1497 |
{
|
| 1498 |
+
"epoch": 2.1,
|
| 1499 |
+
"learning_rate": 4.9538892779229645e-06,
|
| 1500 |
+
"loss": 0.509,
|
| 1501 |
"step": 1225
|
| 1502 |
},
|
| 1503 |
{
|
| 1504 |
+
"epoch": 2.11,
|
| 1505 |
+
"learning_rate": 4.867819469005851e-06,
|
| 1506 |
+
"loss": 0.4434,
|
| 1507 |
"step": 1230
|
| 1508 |
},
|
| 1509 |
{
|
| 1510 |
+
"epoch": 2.12,
|
| 1511 |
+
"learning_rate": 4.782262740802975e-06,
|
| 1512 |
+
"loss": 0.4493,
|
| 1513 |
"step": 1235
|
| 1514 |
},
|
| 1515 |
{
|
| 1516 |
+
"epoch": 2.13,
|
| 1517 |
+
"learning_rate": 4.697227646697627e-06,
|
| 1518 |
+
"loss": 0.4474,
|
| 1519 |
"step": 1240
|
| 1520 |
},
|
| 1521 |
{
|
| 1522 |
+
"epoch": 2.14,
|
| 1523 |
+
"learning_rate": 4.612722687923658e-06,
|
| 1524 |
+
"loss": 0.4455,
|
| 1525 |
"step": 1245
|
| 1526 |
},
|
| 1527 |
{
|
| 1528 |
+
"epoch": 2.15,
|
| 1529 |
+
"learning_rate": 4.5287563127155545e-06,
|
| 1530 |
+
"loss": 0.43,
|
| 1531 |
"step": 1250
|
| 1532 |
},
|
| 1533 |
{
|
| 1534 |
+
"epoch": 2.16,
|
| 1535 |
+
"learning_rate": 4.445336915463874e-06,
|
| 1536 |
+
"loss": 0.4147,
|
| 1537 |
"step": 1255
|
| 1538 |
},
|
| 1539 |
{
|
| 1540 |
+
"epoch": 2.16,
|
| 1541 |
+
"learning_rate": 4.362472835876011e-06,
|
| 1542 |
+
"loss": 0.3961,
|
| 1543 |
"step": 1260
|
| 1544 |
},
|
| 1545 |
{
|
| 1546 |
+
"epoch": 2.17,
|
| 1547 |
+
"learning_rate": 4.280172358142448e-06,
|
| 1548 |
+
"loss": 0.4416,
|
| 1549 |
"step": 1265
|
| 1550 |
},
|
| 1551 |
{
|
| 1552 |
+
"epoch": 2.18,
|
| 1553 |
+
"learning_rate": 4.198443710108583e-06,
|
| 1554 |
+
"loss": 0.3929,
|
| 1555 |
"step": 1270
|
| 1556 |
},
|
| 1557 |
{
|
| 1558 |
+
"epoch": 2.19,
|
| 1559 |
+
"learning_rate": 4.117295062452122e-06,
|
| 1560 |
+
"loss": 0.4044,
|
| 1561 |
"step": 1275
|
| 1562 |
},
|
| 1563 |
{
|
| 1564 |
+
"epoch": 2.2,
|
| 1565 |
+
"learning_rate": 4.036734527866266e-06,
|
| 1566 |
+
"loss": 0.4289,
|
| 1567 |
"step": 1280
|
| 1568 |
},
|
| 1569 |
{
|
| 1570 |
+
"epoch": 2.21,
|
| 1571 |
+
"learning_rate": 3.956770160248658e-06,
|
| 1572 |
+
"loss": 0.383,
|
| 1573 |
"step": 1285
|
| 1574 |
},
|
| 1575 |
{
|
| 1576 |
+
"epoch": 2.22,
|
| 1577 |
+
"learning_rate": 3.8774099538961705e-06,
|
| 1578 |
+
"loss": 0.391,
|
| 1579 |
"step": 1290
|
| 1580 |
},
|
| 1581 |
{
|
| 1582 |
+
"epoch": 2.23,
|
| 1583 |
+
"learning_rate": 3.79866184270575e-06,
|
| 1584 |
+
"loss": 0.3592,
|
| 1585 |
"step": 1295
|
| 1586 |
},
|
| 1587 |
{
|
| 1588 |
+
"epoch": 2.23,
|
| 1589 |
+
"learning_rate": 3.720533699381177e-06,
|
| 1590 |
+
"loss": 0.3803,
|
| 1591 |
"step": 1300
|
| 1592 |
},
|
| 1593 |
{
|
| 1594 |
+
"epoch": 2.24,
|
| 1595 |
+
"learning_rate": 3.64303333464605e-06,
|
| 1596 |
+
"loss": 0.3274,
|
| 1597 |
"step": 1305
|
| 1598 |
},
|
| 1599 |
{
|
| 1600 |
+
"epoch": 2.25,
|
| 1601 |
+
"learning_rate": 3.5661684964629174e-06,
|
| 1602 |
+
"loss": 0.3652,
|
| 1603 |
"step": 1310
|
| 1604 |
},
|
| 1605 |
{
|
| 1606 |
+
"epoch": 2.26,
|
| 1607 |
+
"learning_rate": 3.4899468692586537e-06,
|
| 1608 |
+
"loss": 0.3681,
|
| 1609 |
"step": 1315
|
| 1610 |
},
|
| 1611 |
{
|
| 1612 |
+
"epoch": 2.27,
|
| 1613 |
+
"learning_rate": 3.414376073156275e-06,
|
| 1614 |
+
"loss": 0.3589,
|
| 1615 |
"step": 1320
|
| 1616 |
},
|
| 1617 |
{
|
| 1618 |
+
"epoch": 2.28,
|
| 1619 |
+
"learning_rate": 3.3394636632130893e-06,
|
| 1620 |
+
"loss": 0.3431,
|
| 1621 |
"step": 1325
|
| 1622 |
},
|
| 1623 |
{
|
| 1624 |
+
"epoch": 2.29,
|
| 1625 |
+
"learning_rate": 3.2652171286654032e-06,
|
| 1626 |
+
"loss": 0.3458,
|
| 1627 |
"step": 1330
|
| 1628 |
},
|
| 1629 |
{
|
| 1630 |
+
"epoch": 2.29,
|
| 1631 |
+
"learning_rate": 3.191643892179821e-06,
|
| 1632 |
+
"loss": 0.3163,
|
| 1633 |
"step": 1335
|
| 1634 |
},
|
| 1635 |
{
|
| 1636 |
+
"epoch": 2.3,
|
| 1637 |
+
"learning_rate": 3.1187513091111345e-06,
|
| 1638 |
+
"loss": 0.2961,
|
| 1639 |
"step": 1340
|
| 1640 |
},
|
| 1641 |
{
|
| 1642 |
+
"epoch": 2.31,
|
| 1643 |
+
"learning_rate": 3.0465466667670286e-06,
|
| 1644 |
+
"loss": 0.3045,
|
| 1645 |
"step": 1345
|
| 1646 |
},
|
| 1647 |
{
|
| 1648 |
+
"epoch": 2.32,
|
| 1649 |
+
"learning_rate": 2.975037183679519e-06,
|
| 1650 |
+
"loss": 0.3366,
|
| 1651 |
"step": 1350
|
| 1652 |
},
|
| 1653 |
{
|
| 1654 |
+
"epoch": 2.33,
|
| 1655 |
+
"learning_rate": 2.9042300088832877e-06,
|
| 1656 |
+
"loss": 0.3511,
|
| 1657 |
"step": 1355
|
| 1658 |
},
|
| 1659 |
{
|
| 1660 |
+
"epoch": 2.34,
|
| 1661 |
+
"learning_rate": 2.834132221200998e-06,
|
| 1662 |
+
"loss": 0.3254,
|
| 1663 |
"step": 1360
|
| 1664 |
},
|
| 1665 |
{
|
| 1666 |
+
"epoch": 2.35,
|
| 1667 |
+
"learning_rate": 2.7647508285355774e-06,
|
| 1668 |
+
"loss": 0.3208,
|
| 1669 |
"step": 1365
|
| 1670 |
},
|
| 1671 |
{
|
| 1672 |
+
"epoch": 2.35,
|
| 1673 |
+
"learning_rate": 2.6960927671696213e-06,
|
| 1674 |
+
"loss": 0.3209,
|
| 1675 |
"step": 1370
|
| 1676 |
},
|
| 1677 |
{
|
| 1678 |
+
"epoch": 2.36,
|
| 1679 |
+
"learning_rate": 2.6281649010719577e-06,
|
| 1680 |
+
"loss": 0.3605,
|
| 1681 |
"step": 1375
|
| 1682 |
},
|
| 1683 |
{
|
| 1684 |
+
"epoch": 2.37,
|
| 1685 |
+
"learning_rate": 2.5609740212114154e-06,
|
| 1686 |
+
"loss": 0.3405,
|
| 1687 |
"step": 1380
|
| 1688 |
},
|
| 1689 |
{
|
| 1690 |
+
"epoch": 2.38,
|
| 1691 |
+
"learning_rate": 2.494526844877938e-06,
|
| 1692 |
+
"loss": 0.3352,
|
| 1693 |
"step": 1385
|
| 1694 |
},
|
| 1695 |
{
|
| 1696 |
+
"epoch": 2.39,
|
| 1697 |
+
"learning_rate": 2.4288300150110077e-06,
|
| 1698 |
+
"loss": 0.307,
|
| 1699 |
"step": 1390
|
| 1700 |
},
|
| 1701 |
{
|
| 1702 |
+
"epoch": 2.4,
|
| 1703 |
+
"learning_rate": 2.3638900995355418e-06,
|
| 1704 |
+
"loss": 0.3171,
|
| 1705 |
"step": 1395
|
| 1706 |
},
|
| 1707 |
{
|
| 1708 |
+
"epoch": 2.41,
|
| 1709 |
+
"learning_rate": 2.299713590705275e-06,
|
| 1710 |
+
"loss": 0.3189,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1711 |
"step": 1400
|
| 1712 |
},
|
| 1713 |
{
|
| 1714 |
+
"epoch": 2.41,
|
| 1715 |
+
"learning_rate": 2.2363069044537044e-06,
|
| 1716 |
+
"loss": 0.3225,
|
| 1717 |
"step": 1405
|
| 1718 |
},
|
| 1719 |
{
|
| 1720 |
+
"epoch": 2.42,
|
| 1721 |
+
"learning_rate": 2.173676379752672e-06,
|
| 1722 |
+
"loss": 0.3645,
|
| 1723 |
"step": 1410
|
| 1724 |
},
|
| 1725 |
{
|
| 1726 |
+
"epoch": 2.43,
|
| 1727 |
+
"learning_rate": 2.1118282779786325e-06,
|
| 1728 |
+
"loss": 0.3216,
|
| 1729 |
"step": 1415
|
| 1730 |
},
|
| 1731 |
{
|
| 1732 |
+
"epoch": 2.44,
|
| 1733 |
+
"learning_rate": 2.0507687822866897e-06,
|
| 1734 |
+
"loss": 0.3094,
|
| 1735 |
"step": 1420
|
| 1736 |
},
|
| 1737 |
{
|
| 1738 |
+
"epoch": 2.45,
|
| 1739 |
+
"learning_rate": 1.9905039969924388e-06,
|
| 1740 |
+
"loss": 0.3119,
|
| 1741 |
"step": 1425
|
| 1742 |
},
|
| 1743 |
{
|
| 1744 |
+
"epoch": 2.46,
|
| 1745 |
+
"learning_rate": 1.9310399469617048e-06,
|
| 1746 |
+
"loss": 0.3193,
|
| 1747 |
"step": 1430
|
| 1748 |
},
|
| 1749 |
{
|
| 1750 |
+
"epoch": 2.47,
|
| 1751 |
+
"learning_rate": 1.8723825770082116e-06,
|
| 1752 |
+
"loss": 0.3117,
|
| 1753 |
"step": 1435
|
| 1754 |
},
|
| 1755 |
{
|
| 1756 |
+
"epoch": 2.47,
|
| 1757 |
+
"learning_rate": 1.8145377512992646e-06,
|
| 1758 |
+
"loss": 0.3199,
|
| 1759 |
"step": 1440
|
| 1760 |
},
|
| 1761 |
{
|
| 1762 |
+
"epoch": 2.48,
|
| 1763 |
+
"learning_rate": 1.7575112527694871e-06,
|
| 1764 |
+
"loss": 0.2935,
|
| 1765 |
"step": 1445
|
| 1766 |
},
|
| 1767 |
{
|
| 1768 |
+
"epoch": 2.49,
|
| 1769 |
+
"learning_rate": 1.7013087825426855e-06,
|
| 1770 |
+
"loss": 0.3258,
|
| 1771 |
"step": 1450
|
| 1772 |
},
|
| 1773 |
{
|
| 1774 |
+
"epoch": 2.5,
|
| 1775 |
+
"learning_rate": 1.645935959361884e-06,
|
| 1776 |
+
"loss": 0.3007,
|
| 1777 |
"step": 1455
|
| 1778 |
},
|
| 1779 |
{
|
| 1780 |
+
"epoch": 2.51,
|
| 1781 |
+
"learning_rate": 1.591398319027605e-06,
|
| 1782 |
+
"loss": 0.3276,
|
| 1783 |
"step": 1460
|
| 1784 |
},
|
| 1785 |
{
|
| 1786 |
+
"epoch": 2.52,
|
| 1787 |
+
"learning_rate": 1.5377013138444386e-06,
|
| 1788 |
+
"loss": 0.3204,
|
| 1789 |
"step": 1465
|
| 1790 |
},
|
| 1791 |
{
|
| 1792 |
+
"epoch": 2.53,
|
| 1793 |
+
"learning_rate": 1.4848503120759527e-06,
|
| 1794 |
+
"loss": 0.3157,
|
| 1795 |
"step": 1470
|
| 1796 |
},
|
| 1797 |
{
|
| 1798 |
+
"epoch": 2.53,
|
| 1799 |
+
"learning_rate": 1.432850597408012e-06,
|
| 1800 |
+
"loss": 0.2953,
|
| 1801 |
"step": 1475
|
| 1802 |
},
|
| 1803 |
{
|
| 1804 |
+
"epoch": 2.54,
|
| 1805 |
+
"learning_rate": 1.3817073684205562e-06,
|
| 1806 |
+
"loss": 0.2959,
|
| 1807 |
"step": 1480
|
| 1808 |
},
|
| 1809 |
{
|
| 1810 |
+
"epoch": 2.55,
|
| 1811 |
+
"learning_rate": 1.331425738067872e-06,
|
| 1812 |
+
"loss": 0.3428,
|
| 1813 |
"step": 1485
|
| 1814 |
},
|
| 1815 |
{
|
| 1816 |
+
"epoch": 2.56,
|
| 1817 |
+
"learning_rate": 1.2820107331674426e-06,
|
| 1818 |
+
"loss": 0.3364,
|
| 1819 |
"step": 1490
|
| 1820 |
},
|
| 1821 |
{
|
| 1822 |
+
"epoch": 2.57,
|
| 1823 |
+
"learning_rate": 1.2334672938973958e-06,
|
| 1824 |
+
"loss": 0.3162,
|
| 1825 |
"step": 1495
|
| 1826 |
},
|
| 1827 |
{
|
| 1828 |
+
"epoch": 2.58,
|
| 1829 |
+
"learning_rate": 1.1858002733026196e-06,
|
| 1830 |
+
"loss": 0.2815,
|
| 1831 |
"step": 1500
|
| 1832 |
},
|
| 1833 |
{
|
| 1834 |
+
"epoch": 2.59,
|
| 1835 |
+
"learning_rate": 1.139014436809589e-06,
|
| 1836 |
+
"loss": 0.3064,
|
| 1837 |
"step": 1505
|
| 1838 |
},
|
| 1839 |
{
|
| 1840 |
+
"epoch": 2.59,
|
| 1841 |
+
"learning_rate": 1.093114461749948e-06,
|
| 1842 |
+
"loss": 0.3043,
|
| 1843 |
"step": 1510
|
| 1844 |
},
|
| 1845 |
{
|
| 1846 |
+
"epoch": 2.6,
|
| 1847 |
+
"learning_rate": 1.048104936892902e-06,
|
| 1848 |
+
"loss": 0.3775,
|
| 1849 |
"step": 1515
|
| 1850 |
},
|
| 1851 |
{
|
| 1852 |
+
"epoch": 2.61,
|
| 1853 |
+
"learning_rate": 1.003990361986472e-06,
|
| 1854 |
+
"loss": 0.3461,
|
| 1855 |
"step": 1520
|
| 1856 |
},
|
| 1857 |
{
|
| 1858 |
+
"epoch": 2.62,
|
| 1859 |
+
"learning_rate": 9.607751473076232e-07,
|
| 1860 |
+
"loss": 0.3095,
|
| 1861 |
"step": 1525
|
| 1862 |
},
|
| 1863 |
{
|
| 1864 |
+
"epoch": 2.63,
|
| 1865 |
+
"learning_rate": 9.18463613221372e-07,
|
| 1866 |
+
"loss": 0.2902,
|
| 1867 |
"step": 1530
|
| 1868 |
},
|
| 1869 |
{
|
| 1870 |
+
"epoch": 2.64,
|
| 1871 |
+
"learning_rate": 8.770599897488563e-07,
|
| 1872 |
+
"loss": 0.3263,
|
| 1873 |
"step": 1535
|
| 1874 |
},
|
| 1875 |
{
|
| 1876 |
+
"epoch": 2.65,
|
| 1877 |
+
"learning_rate": 8.365684161444454e-07,
|
| 1878 |
+
"loss": 0.3094,
|
| 1879 |
"step": 1540
|
| 1880 |
},
|
| 1881 |
{
|
| 1882 |
+
"epoch": 2.65,
|
| 1883 |
+
"learning_rate": 7.969929404819377e-07,
|
| 1884 |
+
"loss": 0.3099,
|
| 1885 |
"step": 1545
|
| 1886 |
},
|
| 1887 |
{
|
| 1888 |
+
"epoch": 2.66,
|
| 1889 |
+
"learning_rate": 7.583375192498422e-07,
|
| 1890 |
+
"loss": 0.3441,
|
| 1891 |
"step": 1550
|
| 1892 |
},
|
| 1893 |
{
|
| 1894 |
+
"epoch": 2.67,
|
| 1895 |
+
"learning_rate": 7.206060169558482e-07,
|
| 1896 |
+
"loss": 0.3085,
|
| 1897 |
"step": 1555
|
| 1898 |
},
|
| 1899 |
{
|
| 1900 |
+
"epoch": 2.68,
|
| 1901 |
+
"learning_rate": 6.838022057404848e-07,
|
| 1902 |
+
"loss": 0.2788,
|
| 1903 |
"step": 1560
|
| 1904 |
},
|
| 1905 |
{
|
| 1906 |
+
"epoch": 2.69,
|
| 1907 |
+
"learning_rate": 6.479297649999839e-07,
|
| 1908 |
+
"loss": 0.3381,
|
| 1909 |
"step": 1565
|
| 1910 |
},
|
| 1911 |
{
|
| 1912 |
+
"epoch": 2.7,
|
| 1913 |
+
"learning_rate": 6.129922810184663e-07,
|
| 1914 |
+
"loss": 0.3003,
|
| 1915 |
"step": 1570
|
| 1916 |
},
|
| 1917 |
{
|
| 1918 |
+
"epoch": 2.71,
|
| 1919 |
+
"learning_rate": 5.789932466093884e-07,
|
| 1920 |
+
"loss": 0.3267,
|
| 1921 |
"step": 1575
|
| 1922 |
},
|
| 1923 |
{
|
| 1924 |
+
"epoch": 2.71,
|
| 1925 |
+
"learning_rate": 5.459360607663644e-07,
|
| 1926 |
+
"loss": 0.3103,
|
| 1927 |
"step": 1580
|
| 1928 |
},
|
| 1929 |
{
|
| 1930 |
+
"epoch": 2.72,
|
| 1931 |
+
"learning_rate": 5.138240283233631e-07,
|
| 1932 |
+
"loss": 0.2711,
|
| 1933 |
"step": 1585
|
| 1934 |
},
|
| 1935 |
{
|
| 1936 |
+
"epoch": 2.73,
|
| 1937 |
+
"learning_rate": 4.826603596242918e-07,
|
| 1938 |
+
"loss": 0.3007,
|
| 1939 |
"step": 1590
|
| 1940 |
},
|
| 1941 |
{
|
| 1942 |
+
"epoch": 2.74,
|
| 1943 |
+
"learning_rate": 4.5244817020207775e-07,
|
| 1944 |
+
"loss": 0.3087,
|
| 1945 |
"step": 1595
|
| 1946 |
},
|
| 1947 |
{
|
| 1948 |
+
"epoch": 2.75,
|
| 1949 |
+
"learning_rate": 4.231904804671727e-07,
|
| 1950 |
+
"loss": 0.3132,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1951 |
"step": 1600
|
| 1952 |
},
|
| 1953 |
{
|
| 1954 |
+
"epoch": 2.76,
|
| 1955 |
+
"learning_rate": 3.9489021540560114e-07,
|
| 1956 |
+
"loss": 0.3119,
|
| 1957 |
"step": 1605
|
| 1958 |
},
|
| 1959 |
{
|
| 1960 |
+
"epoch": 2.77,
|
| 1961 |
+
"learning_rate": 3.6755020428654974e-07,
|
| 1962 |
+
"loss": 0.2992,
|
| 1963 |
"step": 1610
|
| 1964 |
},
|
| 1965 |
{
|
| 1966 |
+
"epoch": 2.77,
|
| 1967 |
+
"learning_rate": 3.411731803794982e-07,
|
| 1968 |
+
"loss": 0.3365,
|
| 1969 |
"step": 1615
|
| 1970 |
},
|
| 1971 |
{
|
| 1972 |
+
"epoch": 2.78,
|
| 1973 |
+
"learning_rate": 3.1576178068098205e-07,
|
| 1974 |
+
"loss": 0.3505,
|
| 1975 |
"step": 1620
|
| 1976 |
},
|
| 1977 |
{
|
| 1978 |
+
"epoch": 2.79,
|
| 1979 |
+
"learning_rate": 2.913185456509537e-07,
|
| 1980 |
+
"loss": 0.3634,
|
| 1981 |
"step": 1625
|
| 1982 |
},
|
| 1983 |
{
|
| 1984 |
+
"epoch": 2.8,
|
| 1985 |
+
"learning_rate": 2.678459189588045e-07,
|
| 1986 |
+
"loss": 0.3027,
|
| 1987 |
"step": 1630
|
| 1988 |
},
|
| 1989 |
{
|
| 1990 |
+
"epoch": 2.81,
|
| 1991 |
+
"learning_rate": 2.453462472390711e-07,
|
| 1992 |
+
"loss": 0.3239,
|
| 1993 |
"step": 1635
|
| 1994 |
},
|
| 1995 |
{
|
| 1996 |
+
"epoch": 2.82,
|
| 1997 |
+
"learning_rate": 2.238217798568254e-07,
|
| 1998 |
+
"loss": 0.3222,
|
| 1999 |
"step": 1640
|
| 2000 |
},
|
| 2001 |
{
|
| 2002 |
+
"epoch": 2.83,
|
| 2003 |
+
"learning_rate": 2.0327466868280354e-07,
|
| 2004 |
+
"loss": 0.3058,
|
| 2005 |
"step": 1645
|
| 2006 |
},
|
| 2007 |
{
|
| 2008 |
+
"epoch": 2.84,
|
| 2009 |
+
"learning_rate": 1.8370696787827658e-07,
|
| 2010 |
+
"loss": 0.3013,
|
| 2011 |
"step": 1650
|
| 2012 |
},
|
| 2013 |
{
|
| 2014 |
+
"epoch": 2.84,
|
| 2015 |
+
"learning_rate": 1.6512063368968488e-07,
|
| 2016 |
+
"loss": 0.2995,
|
| 2017 |
"step": 1655
|
| 2018 |
},
|
| 2019 |
{
|
| 2020 |
+
"epoch": 2.85,
|
| 2021 |
+
"learning_rate": 1.4751752425307243e-07,
|
| 2022 |
+
"loss": 0.325,
|
| 2023 |
"step": 1660
|
| 2024 |
},
|
| 2025 |
{
|
| 2026 |
+
"epoch": 2.86,
|
| 2027 |
+
"learning_rate": 1.3089939940831854e-07,
|
| 2028 |
+
"loss": 0.3107,
|
| 2029 |
"step": 1665
|
| 2030 |
},
|
| 2031 |
{
|
| 2032 |
+
"epoch": 2.87,
|
| 2033 |
+
"learning_rate": 1.1526792052319769e-07,
|
| 2034 |
+
"loss": 0.3109,
|
| 2035 |
"step": 1670
|
| 2036 |
},
|
| 2037 |
{
|
| 2038 |
+
"epoch": 2.88,
|
| 2039 |
+
"learning_rate": 1.0062465032729896e-07,
|
| 2040 |
+
"loss": 0.3114,
|
| 2041 |
"step": 1675
|
| 2042 |
},
|
| 2043 |
{
|
| 2044 |
+
"epoch": 2.89,
|
| 2045 |
+
"learning_rate": 8.697105275577989e-08,
|
| 2046 |
+
"loss": 0.3613,
|
| 2047 |
"step": 1680
|
| 2048 |
},
|
| 2049 |
{
|
| 2050 |
+
"epoch": 2.9,
|
| 2051 |
+
"learning_rate": 7.430849280302465e-08,
|
| 2052 |
+
"loss": 0.3221,
|
| 2053 |
"step": 1685
|
| 2054 |
},
|
| 2055 |
{
|
| 2056 |
+
"epoch": 2.9,
|
| 2057 |
+
"learning_rate": 6.263823638617328e-08,
|
| 2058 |
+
"loss": 0.286,
|
| 2059 |
"step": 1690
|
| 2060 |
},
|
| 2061 |
{
|
| 2062 |
+
"epoch": 2.91,
|
| 2063 |
+
"learning_rate": 5.196145021856836e-08,
|
| 2064 |
+
"loss": 0.3689,
|
| 2065 |
"step": 1695
|
| 2066 |
},
|
| 2067 |
{
|
| 2068 |
+
"epoch": 2.92,
|
| 2069 |
+
"learning_rate": 4.2279201693110615e-08,
|
| 2070 |
+
"loss": 0.3032,
|
| 2071 |
"step": 1700
|
| 2072 |
},
|
| 2073 |
{
|
| 2074 |
+
"epoch": 2.93,
|
| 2075 |
+
"learning_rate": 3.359245877555428e-08,
|
| 2076 |
+
"loss": 0.3079,
|
| 2077 |
"step": 1705
|
| 2078 |
},
|
| 2079 |
{
|
| 2080 |
+
"epoch": 2.94,
|
| 2081 |
+
"learning_rate": 2.5902089907730022e-08,
|
| 2082 |
+
"loss": 0.3242,
|
| 2083 |
"step": 1710
|
| 2084 |
},
|
| 2085 |
{
|
| 2086 |
+
"epoch": 2.95,
|
| 2087 |
+
"learning_rate": 1.920886392072885e-08,
|
| 2088 |
+
"loss": 0.3382,
|
| 2089 |
"step": 1715
|
| 2090 |
},
|
| 2091 |
{
|
| 2092 |
+
"epoch": 2.96,
|
| 2093 |
+
"learning_rate": 1.3513449958034718e-08,
|
| 2094 |
+
"loss": 0.329,
|
| 2095 |
"step": 1720
|
| 2096 |
},
|
| 2097 |
{
|
| 2098 |
+
"epoch": 2.96,
|
| 2099 |
+
"learning_rate": 8.816417408634704e-09,
|
| 2100 |
+
"loss": 0.3636,
|
| 2101 |
"step": 1725
|
| 2102 |
},
|
| 2103 |
{
|
| 2104 |
+
"epoch": 2.97,
|
| 2105 |
+
"learning_rate": 5.11823585008675e-09,
|
| 2106 |
+
"loss": 0.3181,
|
| 2107 |
"step": 1730
|
| 2108 |
},
|
| 2109 |
{
|
| 2110 |
+
"epoch": 2.98,
|
| 2111 |
+
"learning_rate": 2.4192750015850088e-09,
|
| 2112 |
+
"loss": 0.3656,
|
| 2113 |
"step": 1735
|
| 2114 |
},
|
| 2115 |
{
|
| 2116 |
+
"epoch": 2.99,
|
| 2117 |
+
"learning_rate": 7.198046869882902e-10,
|
| 2118 |
+
"loss": 0.3502,
|
| 2119 |
"step": 1740
|
| 2120 |
},
|
| 2121 |
{
|
| 2122 |
+
"epoch": 3.0,
|
| 2123 |
+
"learning_rate": 1.999480784942165e-11,
|
| 2124 |
+
"loss": 0.3156,
|
| 2125 |
"step": 1745
|
| 2126 |
},
|
| 2127 |
{
|
| 2128 |
+
"epoch": 3.0,
|
| 2129 |
+
"eval_loss": 1.7387598752975464,
|
| 2130 |
+
"eval_runtime": 133.3136,
|
| 2131 |
+
"eval_samples_per_second": 10.224,
|
| 2132 |
+
"eval_steps_per_second": 0.113,
|
| 2133 |
+
"step": 1746
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2134 |
},
|
| 2135 |
{
|
| 2136 |
"epoch": 3.0,
|
| 2137 |
+
"step": 1746,
|
| 2138 |
+
"total_flos": 1230956927057920.0,
|
| 2139 |
+
"train_loss": 0.872046828952715,
|
| 2140 |
+
"train_runtime": 30104.3205,
|
| 2141 |
+
"train_samples_per_second": 2.783,
|
| 2142 |
+
"train_steps_per_second": 0.058
|
| 2143 |
}
|
| 2144 |
],
|
| 2145 |
"logging_steps": 5,
|
| 2146 |
+
"max_steps": 1746,
|
| 2147 |
"num_train_epochs": 3,
|
| 2148 |
+
"save_steps": 500,
|
| 2149 |
+
"total_flos": 1230956927057920.0,
|
| 2150 |
"trial_name": null,
|
| 2151 |
"trial_params": null
|
| 2152 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5b0290edc7be3949c10c128c426498fb54dddfcd02410971074d0790c2b6ac2b
|
| 3 |
+
size 5816
|