Commit
•
0722a99
1
Parent(s):
26e4e72
🍻 cheers
Browse files- README.md +10 -6
- all_results.json +14 -14
- config.json +1 -1
- eval_results.json +10 -10
- train_results.json +6 -6
- trainer_state.json +730 -14
README.md
CHANGED
@@ -1,7 +1,11 @@
|
|
1 |
---
|
|
|
|
|
2 |
license: apache-2.0
|
3 |
base_model: facebook/dinov2-large
|
4 |
tags:
|
|
|
|
|
5 |
- generated_from_trainer
|
6 |
metrics:
|
7 |
- accuracy
|
@@ -15,13 +19,13 @@ should probably proofread and complete it, then remove this comment. -->
|
|
15 |
|
16 |
# dinov2-large-2024_01_05-kornia_img-size518_batch-size32_epochs70_freeze
|
17 |
|
18 |
-
This model is a fine-tuned version of [facebook/dinov2-large](https://huggingface.co/facebook/dinov2-large) on the
|
19 |
It achieves the following results on the evaluation set:
|
20 |
-
- Loss: 0.
|
21 |
-
- F1 Micro: 0.
|
22 |
-
- F1 Macro: 0.
|
23 |
-
- Roc Auc: 0.
|
24 |
-
- Accuracy: 0.
|
25 |
- Learning Rate: 0.0000
|
26 |
|
27 |
## Model description
|
|
|
1 |
---
|
2 |
+
language:
|
3 |
+
- eng
|
4 |
license: apache-2.0
|
5 |
base_model: facebook/dinov2-large
|
6 |
tags:
|
7 |
+
- multilabel-image-classification
|
8 |
+
- multilabel
|
9 |
- generated_from_trainer
|
10 |
metrics:
|
11 |
- accuracy
|
|
|
19 |
|
20 |
# dinov2-large-2024_01_05-kornia_img-size518_batch-size32_epochs70_freeze
|
21 |
|
22 |
+
This model is a fine-tuned version of [facebook/dinov2-large](https://huggingface.co/facebook/dinov2-large) on the multilabel_complete_dataset dataset.
|
23 |
It achieves the following results on the evaluation set:
|
24 |
+
- Loss: 0.0825
|
25 |
+
- F1 Micro: 0.8570
|
26 |
+
- F1 Macro: 0.7430
|
27 |
+
- Roc Auc: 0.9080
|
28 |
+
- Accuracy: 0.5739
|
29 |
- Learning Rate: 0.0000
|
30 |
|
31 |
## Model description
|
all_results.json
CHANGED
@@ -1,16 +1,16 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"eval_accuracy": 0.
|
4 |
-
"eval_f1_macro": 0.
|
5 |
-
"eval_f1_micro": 0.
|
6 |
-
"eval_loss": 0.
|
7 |
-
"eval_roc_auc": 0.
|
8 |
-
"eval_runtime":
|
9 |
-
"eval_samples_per_second": 4.
|
10 |
-
"eval_steps_per_second": 0.
|
11 |
-
"learning_rate":
|
12 |
-
"train_loss": 0.
|
13 |
-
"train_runtime":
|
14 |
-
"train_samples_per_second":
|
15 |
-
"train_steps_per_second": 0.
|
16 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 114.0,
|
3 |
+
"eval_accuracy": 0.5738795518207283,
|
4 |
+
"eval_f1_macro": 0.7429818572746157,
|
5 |
+
"eval_f1_micro": 0.8570132153593103,
|
6 |
+
"eval_loss": 0.08252906054258347,
|
7 |
+
"eval_roc_auc": 0.9080498428032193,
|
8 |
+
"eval_runtime": 681.2636,
|
9 |
+
"eval_samples_per_second": 4.192,
|
10 |
+
"eval_steps_per_second": 0.132,
|
11 |
+
"learning_rate": 1.0000000000000002e-07,
|
12 |
+
"train_loss": 0.033337813025782426,
|
13 |
+
"train_runtime": 131818.0427,
|
14 |
+
"train_samples_per_second": 9.311,
|
15 |
+
"train_steps_per_second": 0.291
|
16 |
}
|
config.json
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "facebook/dinov2-
|
3 |
"apply_layernorm": true,
|
4 |
"architectures": [
|
5 |
"NewheadDinov2ForImageClassification"
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "facebook/dinov2-large2024_01_08",
|
3 |
"apply_layernorm": true,
|
4 |
"architectures": [
|
5 |
"NewheadDinov2ForImageClassification"
|
eval_results.json
CHANGED
@@ -1,12 +1,12 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"eval_accuracy": 0.
|
4 |
-
"eval_f1_macro": 0.
|
5 |
-
"eval_f1_micro": 0.
|
6 |
-
"eval_loss": 0.
|
7 |
-
"eval_roc_auc": 0.
|
8 |
-
"eval_runtime":
|
9 |
-
"eval_samples_per_second": 4.
|
10 |
-
"eval_steps_per_second": 0.
|
11 |
-
"learning_rate":
|
12 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 114.0,
|
3 |
+
"eval_accuracy": 0.5738795518207283,
|
4 |
+
"eval_f1_macro": 0.7429818572746157,
|
5 |
+
"eval_f1_micro": 0.8570132153593103,
|
6 |
+
"eval_loss": 0.08252906054258347,
|
7 |
+
"eval_roc_auc": 0.9080498428032193,
|
8 |
+
"eval_runtime": 681.2636,
|
9 |
+
"eval_samples_per_second": 4.192,
|
10 |
+
"eval_steps_per_second": 0.132,
|
11 |
+
"learning_rate": 1.0000000000000002e-07
|
12 |
}
|
train_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"learning_rate":
|
4 |
-
"train_loss": 0.
|
5 |
-
"train_runtime":
|
6 |
-
"train_samples_per_second":
|
7 |
-
"train_steps_per_second": 0.
|
8 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 114.0,
|
3 |
+
"learning_rate": 1.0000000000000002e-07,
|
4 |
+
"train_loss": 0.033337813025782426,
|
5 |
+
"train_runtime": 131818.0427,
|
6 |
+
"train_samples_per_second": 9.311,
|
7 |
+
"train_steps_per_second": 0.291
|
8 |
}
|
trainer_state.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
-
"best_metric": 0.
|
3 |
-
"best_model_checkpoint": "/home1/datawork/mcontini/models/multilabel/huggingface/dinov2-large-2024_01_05-kornia_img-size518_batch-size32_epochs70_freeze/checkpoint-
|
4 |
-
"epoch":
|
5 |
"eval_steps": 500,
|
6 |
-
"global_step":
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
@@ -1147,21 +1147,737 @@
|
|
1147 |
"step": 19180
|
1148 |
},
|
1149 |
{
|
1150 |
-
"epoch":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1151 |
"learning_rate": 0.0001,
|
1152 |
-
"step":
|
1153 |
-
|
1154 |
-
|
1155 |
-
"
|
1156 |
-
"
|
1157 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1158 |
}
|
1159 |
],
|
1160 |
"logging_steps": 500,
|
1161 |
-
"max_steps":
|
1162 |
-
"num_train_epochs":
|
1163 |
"save_steps": 500,
|
1164 |
-
"total_flos":
|
1165 |
"trial_name": null,
|
1166 |
"trial_params": null
|
1167 |
}
|
|
|
1 |
{
|
2 |
+
"best_metric": 0.08124219626188278,
|
3 |
+
"best_model_checkpoint": "/home1/datawork/mcontini/models/multilabel/huggingface/dinov2-large-2024_01_05-kornia_img-size518_batch-size32_epochs70_freeze/checkpoint-28496",
|
4 |
+
"epoch": 114.0,
|
5 |
"eval_steps": 500,
|
6 |
+
"global_step": 31236,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
|
|
1147 |
"step": 19180
|
1148 |
},
|
1149 |
{
|
1150 |
+
"epoch": 71.0,
|
1151 |
+
"eval_accuracy": 0.557920446615492,
|
1152 |
+
"eval_f1_macro": 0.74841480686329,
|
1153 |
+
"eval_f1_micro": 0.8557041347283614,
|
1154 |
+
"eval_loss": 0.08348394185304642,
|
1155 |
+
"eval_roc_auc": 0.9102149107350765,
|
1156 |
+
"eval_runtime": 688.9991,
|
1157 |
+
"eval_samples_per_second": 4.16,
|
1158 |
+
"eval_steps_per_second": 0.131,
|
1159 |
+
"learning_rate": 0.0001,
|
1160 |
+
"step": 19454
|
1161 |
+
},
|
1162 |
+
{
|
1163 |
+
"epoch": 71.17,
|
1164 |
+
"learning_rate": 0.0001,
|
1165 |
+
"loss": 0.0907,
|
1166 |
+
"step": 19500
|
1167 |
+
},
|
1168 |
+
{
|
1169 |
+
"epoch": 72.0,
|
1170 |
+
"eval_accuracy": 0.5610607117934403,
|
1171 |
+
"eval_f1_macro": 0.7445560004670461,
|
1172 |
+
"eval_f1_micro": 0.8531667140017041,
|
1173 |
+
"eval_loss": 0.08317266404628754,
|
1174 |
+
"eval_roc_auc": 0.9036835353705291,
|
1175 |
+
"eval_runtime": 686.0467,
|
1176 |
+
"eval_samples_per_second": 4.178,
|
1177 |
+
"eval_steps_per_second": 0.131,
|
1178 |
+
"learning_rate": 0.0001,
|
1179 |
+
"step": 19728
|
1180 |
+
},
|
1181 |
+
{
|
1182 |
+
"epoch": 72.99,
|
1183 |
+
"learning_rate": 0.0001,
|
1184 |
+
"loss": 0.0905,
|
1185 |
+
"step": 20000
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 73.0,
|
1189 |
+
"eval_accuracy": 0.5575715282623867,
|
1190 |
+
"eval_f1_macro": 0.7511661878734329,
|
1191 |
+
"eval_f1_micro": 0.8557848235754626,
|
1192 |
+
"eval_loss": 0.0826636552810669,
|
1193 |
+
"eval_roc_auc": 0.9104516775814524,
|
1194 |
+
"eval_runtime": 696.2462,
|
1195 |
+
"eval_samples_per_second": 4.116,
|
1196 |
+
"eval_steps_per_second": 0.129,
|
1197 |
+
"learning_rate": 0.0001,
|
1198 |
+
"step": 20002
|
1199 |
+
},
|
1200 |
+
{
|
1201 |
+
"epoch": 74.0,
|
1202 |
+
"eval_accuracy": 0.5589672016748081,
|
1203 |
+
"eval_f1_macro": 0.7519078199055363,
|
1204 |
+
"eval_f1_micro": 0.8547880690737834,
|
1205 |
+
"eval_loss": 0.08345863223075867,
|
1206 |
+
"eval_roc_auc": 0.9089978296204139,
|
1207 |
+
"eval_runtime": 684.5044,
|
1208 |
+
"eval_samples_per_second": 4.187,
|
1209 |
+
"eval_steps_per_second": 0.131,
|
1210 |
+
"learning_rate": 0.0001,
|
1211 |
+
"step": 20276
|
1212 |
+
},
|
1213 |
+
{
|
1214 |
+
"epoch": 74.82,
|
1215 |
+
"learning_rate": 0.0001,
|
1216 |
+
"loss": 0.0896,
|
1217 |
+
"step": 20500
|
1218 |
+
},
|
1219 |
+
{
|
1220 |
+
"epoch": 75.0,
|
1221 |
+
"eval_accuracy": 0.5565247732030705,
|
1222 |
+
"eval_f1_macro": 0.7427640048071817,
|
1223 |
+
"eval_f1_micro": 0.8535384963511908,
|
1224 |
+
"eval_loss": 0.08291995525360107,
|
1225 |
+
"eval_roc_auc": 0.9052642407433178,
|
1226 |
+
"eval_runtime": 693.0369,
|
1227 |
+
"eval_samples_per_second": 4.135,
|
1228 |
+
"eval_steps_per_second": 0.13,
|
1229 |
+
"learning_rate": 0.0001,
|
1230 |
+
"step": 20550
|
1231 |
+
},
|
1232 |
+
{
|
1233 |
+
"epoch": 76.0,
|
1234 |
+
"eval_accuracy": 0.5642009769713887,
|
1235 |
+
"eval_f1_macro": 0.7449358839801685,
|
1236 |
+
"eval_f1_micro": 0.8560516708789666,
|
1237 |
+
"eval_loss": 0.08278567343950272,
|
1238 |
+
"eval_roc_auc": 0.9091009598641902,
|
1239 |
+
"eval_runtime": 689.5027,
|
1240 |
+
"eval_samples_per_second": 4.157,
|
1241 |
+
"eval_steps_per_second": 0.131,
|
1242 |
+
"learning_rate": 0.0001,
|
1243 |
+
"step": 20824
|
1244 |
+
},
|
1245 |
+
{
|
1246 |
+
"epoch": 76.64,
|
1247 |
+
"learning_rate": 0.0001,
|
1248 |
+
"loss": 0.089,
|
1249 |
+
"step": 21000
|
1250 |
+
},
|
1251 |
+
{
|
1252 |
+
"epoch": 77.0,
|
1253 |
+
"eval_accuracy": 0.5603628750872296,
|
1254 |
+
"eval_f1_macro": 0.7507123347472292,
|
1255 |
+
"eval_f1_micro": 0.8567585743106928,
|
1256 |
+
"eval_loss": 0.08268015086650848,
|
1257 |
+
"eval_roc_auc": 0.9102413784270108,
|
1258 |
+
"eval_runtime": 691.3385,
|
1259 |
+
"eval_samples_per_second": 4.146,
|
1260 |
+
"eval_steps_per_second": 0.13,
|
1261 |
+
"learning_rate": 0.0001,
|
1262 |
+
"step": 21098
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 78.0,
|
1266 |
+
"eval_accuracy": 0.557920446615492,
|
1267 |
+
"eval_f1_macro": 0.7435504236075495,
|
1268 |
+
"eval_f1_micro": 0.8528667079277555,
|
1269 |
+
"eval_loss": 0.0833190307021141,
|
1270 |
+
"eval_roc_auc": 0.9067310989812636,
|
1271 |
+
"eval_runtime": 698.1968,
|
1272 |
+
"eval_samples_per_second": 4.105,
|
1273 |
+
"eval_steps_per_second": 0.129,
|
1274 |
+
"learning_rate": 0.0001,
|
1275 |
+
"step": 21372
|
1276 |
+
},
|
1277 |
+
{
|
1278 |
+
"epoch": 78.47,
|
1279 |
+
"learning_rate": 0.0001,
|
1280 |
+
"loss": 0.0892,
|
1281 |
+
"step": 21500
|
1282 |
+
},
|
1283 |
+
{
|
1284 |
+
"epoch": 79.0,
|
1285 |
+
"eval_accuracy": 0.5589672016748081,
|
1286 |
+
"eval_f1_macro": 0.7501683098189544,
|
1287 |
+
"eval_f1_micro": 0.8540393754243042,
|
1288 |
+
"eval_loss": 0.08301213383674622,
|
1289 |
+
"eval_roc_auc": 0.905519599881194,
|
1290 |
+
"eval_runtime": 693.3725,
|
1291 |
+
"eval_samples_per_second": 4.133,
|
1292 |
+
"eval_steps_per_second": 0.13,
|
1293 |
"learning_rate": 0.0001,
|
1294 |
+
"step": 21646
|
1295 |
+
},
|
1296 |
+
{
|
1297 |
+
"epoch": 80.0,
|
1298 |
+
"eval_accuracy": 0.5600139567341242,
|
1299 |
+
"eval_f1_macro": 0.7461247141006309,
|
1300 |
+
"eval_f1_micro": 0.8548240635641317,
|
1301 |
+
"eval_loss": 0.08266153186559677,
|
1302 |
+
"eval_roc_auc": 0.9048757931076342,
|
1303 |
+
"eval_runtime": 692.1533,
|
1304 |
+
"eval_samples_per_second": 4.141,
|
1305 |
+
"eval_steps_per_second": 0.13,
|
1306 |
+
"learning_rate": 1e-05,
|
1307 |
+
"step": 21920
|
1308 |
+
},
|
1309 |
+
{
|
1310 |
+
"epoch": 80.29,
|
1311 |
+
"learning_rate": 1e-05,
|
1312 |
+
"loss": 0.0879,
|
1313 |
+
"step": 22000
|
1314 |
+
},
|
1315 |
+
{
|
1316 |
+
"epoch": 81.0,
|
1317 |
+
"eval_accuracy": 0.560711793440335,
|
1318 |
+
"eval_f1_macro": 0.7542778186740821,
|
1319 |
+
"eval_f1_micro": 0.8576379185065117,
|
1320 |
+
"eval_loss": 0.08230110257863998,
|
1321 |
+
"eval_roc_auc": 0.9116332184274908,
|
1322 |
+
"eval_runtime": 691.9752,
|
1323 |
+
"eval_samples_per_second": 4.142,
|
1324 |
+
"eval_steps_per_second": 0.13,
|
1325 |
+
"learning_rate": 1e-05,
|
1326 |
+
"step": 22194
|
1327 |
+
},
|
1328 |
+
{
|
1329 |
+
"epoch": 82.0,
|
1330 |
+
"eval_accuracy": 0.5631542219120725,
|
1331 |
+
"eval_f1_macro": 0.7536111063550082,
|
1332 |
+
"eval_f1_micro": 0.8576225654801881,
|
1333 |
+
"eval_loss": 0.08222728967666626,
|
1334 |
+
"eval_roc_auc": 0.9111816666216643,
|
1335 |
+
"eval_runtime": 688.2687,
|
1336 |
+
"eval_samples_per_second": 4.164,
|
1337 |
+
"eval_steps_per_second": 0.131,
|
1338 |
+
"learning_rate": 1e-05,
|
1339 |
+
"step": 22468
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 82.12,
|
1343 |
+
"learning_rate": 1e-05,
|
1344 |
+
"loss": 0.0867,
|
1345 |
+
"step": 22500
|
1346 |
+
},
|
1347 |
+
{
|
1348 |
+
"epoch": 83.0,
|
1349 |
+
"eval_accuracy": 0.5624563852058618,
|
1350 |
+
"eval_f1_macro": 0.7519980936807484,
|
1351 |
+
"eval_f1_micro": 0.8553623024191264,
|
1352 |
+
"eval_loss": 0.08224428445100784,
|
1353 |
+
"eval_roc_auc": 0.9057647902327958,
|
1354 |
+
"eval_runtime": 675.1087,
|
1355 |
+
"eval_samples_per_second": 4.245,
|
1356 |
+
"eval_steps_per_second": 0.133,
|
1357 |
+
"learning_rate": 1e-05,
|
1358 |
+
"step": 22742
|
1359 |
+
},
|
1360 |
+
{
|
1361 |
+
"epoch": 83.94,
|
1362 |
+
"learning_rate": 1e-05,
|
1363 |
+
"loss": 0.0864,
|
1364 |
+
"step": 23000
|
1365 |
+
},
|
1366 |
+
{
|
1367 |
+
"epoch": 84.0,
|
1368 |
+
"eval_accuracy": 0.5638520586182834,
|
1369 |
+
"eval_f1_macro": 0.751088987860079,
|
1370 |
+
"eval_f1_micro": 0.8550642905481616,
|
1371 |
+
"eval_loss": 0.08214889466762543,
|
1372 |
+
"eval_roc_auc": 0.9071555340236915,
|
1373 |
+
"eval_runtime": 684.6287,
|
1374 |
+
"eval_samples_per_second": 4.186,
|
1375 |
+
"eval_steps_per_second": 0.131,
|
1376 |
+
"learning_rate": 1e-05,
|
1377 |
+
"step": 23016
|
1378 |
+
},
|
1379 |
+
{
|
1380 |
+
"epoch": 85.0,
|
1381 |
+
"eval_accuracy": 0.5617585484996511,
|
1382 |
+
"eval_f1_macro": 0.7532850236770345,
|
1383 |
+
"eval_f1_micro": 0.8560117633751838,
|
1384 |
+
"eval_loss": 0.08197268098592758,
|
1385 |
+
"eval_roc_auc": 0.9067157953185159,
|
1386 |
+
"eval_runtime": 685.9962,
|
1387 |
+
"eval_samples_per_second": 4.178,
|
1388 |
+
"eval_steps_per_second": 0.131,
|
1389 |
+
"learning_rate": 1e-05,
|
1390 |
+
"step": 23290
|
1391 |
+
},
|
1392 |
+
{
|
1393 |
+
"epoch": 85.77,
|
1394 |
+
"learning_rate": 1e-05,
|
1395 |
+
"loss": 0.0865,
|
1396 |
+
"step": 23500
|
1397 |
+
},
|
1398 |
+
{
|
1399 |
+
"epoch": 86.0,
|
1400 |
+
"eval_accuracy": 0.5600139567341242,
|
1401 |
+
"eval_f1_macro": 0.7495768803790436,
|
1402 |
+
"eval_f1_micro": 0.8553152949167893,
|
1403 |
+
"eval_loss": 0.0821395218372345,
|
1404 |
+
"eval_roc_auc": 0.9060269285615214,
|
1405 |
+
"eval_runtime": 676.0683,
|
1406 |
+
"eval_samples_per_second": 4.239,
|
1407 |
+
"eval_steps_per_second": 0.133,
|
1408 |
+
"learning_rate": 1e-05,
|
1409 |
+
"step": 23564
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 87.0,
|
1413 |
+
"eval_accuracy": 0.5586182833217027,
|
1414 |
+
"eval_f1_macro": 0.7518865385404515,
|
1415 |
+
"eval_f1_micro": 0.8558878188883257,
|
1416 |
+
"eval_loss": 0.08169202506542206,
|
1417 |
+
"eval_roc_auc": 0.9080884949227034,
|
1418 |
+
"eval_runtime": 683.932,
|
1419 |
+
"eval_samples_per_second": 4.19,
|
1420 |
+
"eval_steps_per_second": 0.132,
|
1421 |
+
"learning_rate": 1e-05,
|
1422 |
+
"step": 23838
|
1423 |
+
},
|
1424 |
+
{
|
1425 |
+
"epoch": 87.59,
|
1426 |
+
"learning_rate": 1e-05,
|
1427 |
+
"loss": 0.0868,
|
1428 |
+
"step": 24000
|
1429 |
+
},
|
1430 |
+
{
|
1431 |
+
"epoch": 88.0,
|
1432 |
+
"eval_accuracy": 0.5621074668527565,
|
1433 |
+
"eval_f1_macro": 0.7526313975322289,
|
1434 |
+
"eval_f1_micro": 0.8558239036198839,
|
1435 |
+
"eval_loss": 0.08171343803405762,
|
1436 |
+
"eval_roc_auc": 0.9081685470699702,
|
1437 |
+
"eval_runtime": 687.6304,
|
1438 |
+
"eval_samples_per_second": 4.168,
|
1439 |
+
"eval_steps_per_second": 0.131,
|
1440 |
+
"learning_rate": 1e-05,
|
1441 |
+
"step": 24112
|
1442 |
+
},
|
1443 |
+
{
|
1444 |
+
"epoch": 89.0,
|
1445 |
+
"eval_accuracy": 0.5638520586182834,
|
1446 |
+
"eval_f1_macro": 0.7535945449960157,
|
1447 |
+
"eval_f1_micro": 0.8569978572234127,
|
1448 |
+
"eval_loss": 0.08180436491966248,
|
1449 |
+
"eval_roc_auc": 0.9082570227224631,
|
1450 |
+
"eval_runtime": 685.8987,
|
1451 |
+
"eval_samples_per_second": 4.178,
|
1452 |
+
"eval_steps_per_second": 0.131,
|
1453 |
+
"learning_rate": 1e-05,
|
1454 |
+
"step": 24386
|
1455 |
+
},
|
1456 |
+
{
|
1457 |
+
"epoch": 89.42,
|
1458 |
+
"learning_rate": 1e-05,
|
1459 |
+
"loss": 0.0857,
|
1460 |
+
"step": 24500
|
1461 |
+
},
|
1462 |
+
{
|
1463 |
+
"epoch": 90.0,
|
1464 |
+
"eval_accuracy": 0.5617585484996511,
|
1465 |
+
"eval_f1_macro": 0.7521712775972677,
|
1466 |
+
"eval_f1_micro": 0.8557914296976182,
|
1467 |
+
"eval_loss": 0.08184907585382462,
|
1468 |
+
"eval_roc_auc": 0.9080738403314199,
|
1469 |
+
"eval_runtime": 2199.2107,
|
1470 |
+
"eval_samples_per_second": 1.303,
|
1471 |
+
"eval_steps_per_second": 0.041,
|
1472 |
+
"learning_rate": 1e-05,
|
1473 |
+
"step": 24660
|
1474 |
+
},
|
1475 |
+
{
|
1476 |
+
"epoch": 91.0,
|
1477 |
+
"eval_accuracy": 0.5631542219120725,
|
1478 |
+
"eval_f1_macro": 0.7496381590553177,
|
1479 |
+
"eval_f1_micro": 0.856868937514099,
|
1480 |
+
"eval_loss": 0.08175913989543915,
|
1481 |
+
"eval_roc_auc": 0.9081476613926293,
|
1482 |
+
"eval_runtime": 685.7142,
|
1483 |
+
"eval_samples_per_second": 4.18,
|
1484 |
+
"eval_steps_per_second": 0.131,
|
1485 |
+
"learning_rate": 1e-05,
|
1486 |
+
"step": 24934
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 91.24,
|
1490 |
+
"learning_rate": 1e-05,
|
1491 |
+
"loss": 0.0862,
|
1492 |
+
"step": 25000
|
1493 |
+
},
|
1494 |
+
{
|
1495 |
+
"epoch": 92.0,
|
1496 |
+
"eval_accuracy": 0.5648988136775994,
|
1497 |
+
"eval_f1_macro": 0.7551884126903309,
|
1498 |
+
"eval_f1_micro": 0.8566131025957974,
|
1499 |
+
"eval_loss": 0.08205177634954453,
|
1500 |
+
"eval_roc_auc": 0.9092762668547993,
|
1501 |
+
"eval_runtime": 688.0975,
|
1502 |
+
"eval_samples_per_second": 4.165,
|
1503 |
+
"eval_steps_per_second": 0.131,
|
1504 |
+
"learning_rate": 1e-05,
|
1505 |
+
"step": 25208
|
1506 |
+
},
|
1507 |
+
{
|
1508 |
+
"epoch": 93.0,
|
1509 |
+
"eval_accuracy": 0.5628053035589672,
|
1510 |
+
"eval_f1_macro": 0.7579638727848569,
|
1511 |
+
"eval_f1_micro": 0.8588642195693407,
|
1512 |
+
"eval_loss": 0.08153587579727173,
|
1513 |
+
"eval_roc_auc": 0.9129889700545543,
|
1514 |
+
"eval_runtime": 686.4355,
|
1515 |
+
"eval_samples_per_second": 4.175,
|
1516 |
+
"eval_steps_per_second": 0.131,
|
1517 |
+
"learning_rate": 1e-05,
|
1518 |
+
"step": 25482
|
1519 |
+
},
|
1520 |
+
{
|
1521 |
+
"epoch": 93.07,
|
1522 |
+
"learning_rate": 1e-05,
|
1523 |
+
"loss": 0.0851,
|
1524 |
+
"step": 25500
|
1525 |
+
},
|
1526 |
+
{
|
1527 |
+
"epoch": 94.0,
|
1528 |
+
"eval_accuracy": 0.5600139567341242,
|
1529 |
+
"eval_f1_macro": 0.7565755790887538,
|
1530 |
+
"eval_f1_micro": 0.8570630932439977,
|
1531 |
+
"eval_loss": 0.0816139355301857,
|
1532 |
+
"eval_roc_auc": 0.911680023691974,
|
1533 |
+
"eval_runtime": 686.9681,
|
1534 |
+
"eval_samples_per_second": 4.172,
|
1535 |
+
"eval_steps_per_second": 0.131,
|
1536 |
+
"learning_rate": 1e-05,
|
1537 |
+
"step": 25756
|
1538 |
+
},
|
1539 |
+
{
|
1540 |
+
"epoch": 94.89,
|
1541 |
+
"learning_rate": 1e-05,
|
1542 |
+
"loss": 0.0854,
|
1543 |
+
"step": 26000
|
1544 |
+
},
|
1545 |
+
{
|
1546 |
+
"epoch": 95.0,
|
1547 |
+
"eval_accuracy": 0.5631542219120725,
|
1548 |
+
"eval_f1_macro": 0.7553215815461111,
|
1549 |
+
"eval_f1_micro": 0.8563582357226924,
|
1550 |
+
"eval_loss": 0.08153104782104492,
|
1551 |
+
"eval_roc_auc": 0.9100006738804182,
|
1552 |
+
"eval_runtime": 792.6719,
|
1553 |
+
"eval_samples_per_second": 3.616,
|
1554 |
+
"eval_steps_per_second": 0.114,
|
1555 |
+
"learning_rate": 1e-05,
|
1556 |
+
"step": 26030
|
1557 |
+
},
|
1558 |
+
{
|
1559 |
+
"epoch": 96.0,
|
1560 |
+
"eval_accuracy": 0.5621074668527565,
|
1561 |
+
"eval_f1_macro": 0.7584737150271473,
|
1562 |
+
"eval_f1_micro": 0.8576208800401539,
|
1563 |
+
"eval_loss": 0.081536203622818,
|
1564 |
+
"eval_roc_auc": 0.9123942619313182,
|
1565 |
+
"eval_runtime": 689.1211,
|
1566 |
+
"eval_samples_per_second": 4.159,
|
1567 |
+
"eval_steps_per_second": 0.131,
|
1568 |
+
"learning_rate": 1e-05,
|
1569 |
+
"step": 26304
|
1570 |
+
},
|
1571 |
+
{
|
1572 |
+
"epoch": 96.72,
|
1573 |
+
"learning_rate": 1e-05,
|
1574 |
+
"loss": 0.0854,
|
1575 |
+
"step": 26500
|
1576 |
+
},
|
1577 |
+
{
|
1578 |
+
"epoch": 97.0,
|
1579 |
+
"eval_accuracy": 0.5628053035589672,
|
1580 |
+
"eval_f1_macro": 0.7578675241138624,
|
1581 |
+
"eval_f1_micro": 0.8575591168889387,
|
1582 |
+
"eval_loss": 0.0817214846611023,
|
1583 |
+
"eval_roc_auc": 0.9107227875201963,
|
1584 |
+
"eval_runtime": 791.8215,
|
1585 |
+
"eval_samples_per_second": 3.62,
|
1586 |
+
"eval_steps_per_second": 0.114,
|
1587 |
+
"learning_rate": 1e-05,
|
1588 |
+
"step": 26578
|
1589 |
+
},
|
1590 |
+
{
|
1591 |
+
"epoch": 98.0,
|
1592 |
+
"eval_accuracy": 0.5638520586182834,
|
1593 |
+
"eval_f1_macro": 0.7526862577512524,
|
1594 |
+
"eval_f1_micro": 0.8570947651910453,
|
1595 |
+
"eval_loss": 0.08161807805299759,
|
1596 |
+
"eval_roc_auc": 0.9100232038721354,
|
1597 |
+
"eval_runtime": 689.7672,
|
1598 |
+
"eval_samples_per_second": 4.155,
|
1599 |
+
"eval_steps_per_second": 0.13,
|
1600 |
+
"learning_rate": 1e-05,
|
1601 |
+
"step": 26852
|
1602 |
+
},
|
1603 |
+
{
|
1604 |
+
"epoch": 98.54,
|
1605 |
+
"learning_rate": 1e-05,
|
1606 |
+
"loss": 0.0855,
|
1607 |
+
"step": 27000
|
1608 |
+
},
|
1609 |
+
{
|
1610 |
+
"epoch": 99.0,
|
1611 |
+
"eval_accuracy": 0.5642009769713887,
|
1612 |
+
"eval_f1_macro": 0.7556285126609241,
|
1613 |
+
"eval_f1_micro": 0.8578035986237239,
|
1614 |
+
"eval_loss": 0.08184286206960678,
|
1615 |
+
"eval_roc_auc": 0.9086036990034648,
|
1616 |
+
"eval_runtime": 685.3364,
|
1617 |
+
"eval_samples_per_second": 4.182,
|
1618 |
+
"eval_steps_per_second": 0.131,
|
1619 |
+
"learning_rate": 1e-05,
|
1620 |
+
"step": 27126
|
1621 |
+
},
|
1622 |
+
{
|
1623 |
+
"epoch": 100.0,
|
1624 |
+
"eval_accuracy": 0.5631542219120725,
|
1625 |
+
"eval_f1_macro": 0.753319281703831,
|
1626 |
+
"eval_f1_micro": 0.857062242537103,
|
1627 |
+
"eval_loss": 0.08161789923906326,
|
1628 |
+
"eval_roc_auc": 0.9080422377630124,
|
1629 |
+
"eval_runtime": 680.1579,
|
1630 |
+
"eval_samples_per_second": 4.214,
|
1631 |
+
"eval_steps_per_second": 0.132,
|
1632 |
+
"learning_rate": 1.0000000000000002e-06,
|
1633 |
+
"step": 27400
|
1634 |
+
},
|
1635 |
+
{
|
1636 |
+
"epoch": 100.36,
|
1637 |
+
"learning_rate": 1.0000000000000002e-06,
|
1638 |
+
"loss": 0.0837,
|
1639 |
+
"step": 27500
|
1640 |
+
},
|
1641 |
+
{
|
1642 |
+
"epoch": 101.0,
|
1643 |
+
"eval_accuracy": 0.5645498953244941,
|
1644 |
+
"eval_f1_macro": 0.7553159980269647,
|
1645 |
+
"eval_f1_micro": 0.8575286968264686,
|
1646 |
+
"eval_loss": 0.08142262697219849,
|
1647 |
+
"eval_roc_auc": 0.9092807526598085,
|
1648 |
+
"eval_runtime": 693.9432,
|
1649 |
+
"eval_samples_per_second": 4.13,
|
1650 |
+
"eval_steps_per_second": 0.13,
|
1651 |
+
"learning_rate": 1.0000000000000002e-06,
|
1652 |
+
"step": 27674
|
1653 |
+
},
|
1654 |
+
{
|
1655 |
+
"epoch": 102.0,
|
1656 |
+
"eval_accuracy": 0.5652477320307048,
|
1657 |
+
"eval_f1_macro": 0.7559444477928223,
|
1658 |
+
"eval_f1_micro": 0.8571749382438806,
|
1659 |
+
"eval_loss": 0.08143333345651627,
|
1660 |
+
"eval_roc_auc": 0.9098557722819599,
|
1661 |
+
"eval_runtime": 695.0171,
|
1662 |
+
"eval_samples_per_second": 4.124,
|
1663 |
+
"eval_steps_per_second": 0.129,
|
1664 |
+
"learning_rate": 1.0000000000000002e-06,
|
1665 |
+
"step": 27948
|
1666 |
+
},
|
1667 |
+
{
|
1668 |
+
"epoch": 102.19,
|
1669 |
+
"learning_rate": 1.0000000000000002e-06,
|
1670 |
+
"loss": 0.085,
|
1671 |
+
"step": 28000
|
1672 |
+
},
|
1673 |
+
{
|
1674 |
+
"epoch": 103.0,
|
1675 |
+
"eval_accuracy": 0.5645498953244941,
|
1676 |
+
"eval_f1_macro": 0.7565514506209431,
|
1677 |
+
"eval_f1_micro": 0.8570462613399448,
|
1678 |
+
"eval_loss": 0.08157742768526077,
|
1679 |
+
"eval_roc_auc": 0.9085338156424723,
|
1680 |
+
"eval_runtime": 710.2061,
|
1681 |
+
"eval_samples_per_second": 4.035,
|
1682 |
+
"eval_steps_per_second": 0.127,
|
1683 |
+
"learning_rate": 1.0000000000000002e-06,
|
1684 |
+
"step": 28222
|
1685 |
+
},
|
1686 |
+
{
|
1687 |
+
"epoch": 104.0,
|
1688 |
+
"eval_accuracy": 0.5645498953244941,
|
1689 |
+
"eval_f1_macro": 0.7572978759334409,
|
1690 |
+
"eval_f1_micro": 0.857623884617543,
|
1691 |
+
"eval_loss": 0.08124219626188278,
|
1692 |
+
"eval_roc_auc": 0.9102385369363781,
|
1693 |
+
"eval_runtime": 703.8703,
|
1694 |
+
"eval_samples_per_second": 4.072,
|
1695 |
+
"eval_steps_per_second": 0.128,
|
1696 |
+
"learning_rate": 1.0000000000000002e-06,
|
1697 |
+
"step": 28496
|
1698 |
+
},
|
1699 |
+
{
|
1700 |
+
"epoch": 104.01,
|
1701 |
+
"learning_rate": 1.0000000000000002e-06,
|
1702 |
+
"loss": 0.0844,
|
1703 |
+
"step": 28500
|
1704 |
+
},
|
1705 |
+
{
|
1706 |
+
"epoch": 105.0,
|
1707 |
+
"eval_accuracy": 0.5603628750872296,
|
1708 |
+
"eval_f1_macro": 0.7588882568234135,
|
1709 |
+
"eval_f1_micro": 0.8571747004736696,
|
1710 |
+
"eval_loss": 0.08170615136623383,
|
1711 |
+
"eval_roc_auc": 0.9124156957134514,
|
1712 |
+
"eval_runtime": 693.4706,
|
1713 |
+
"eval_samples_per_second": 4.133,
|
1714 |
+
"eval_steps_per_second": 0.13,
|
1715 |
+
"learning_rate": 1.0000000000000002e-06,
|
1716 |
+
"step": 28770
|
1717 |
+
},
|
1718 |
+
{
|
1719 |
+
"epoch": 105.84,
|
1720 |
+
"learning_rate": 1.0000000000000002e-06,
|
1721 |
+
"loss": 0.0845,
|
1722 |
+
"step": 29000
|
1723 |
+
},
|
1724 |
+
{
|
1725 |
+
"epoch": 106.0,
|
1726 |
+
"eval_accuracy": 0.5628053035589672,
|
1727 |
+
"eval_f1_macro": 0.7514055460147794,
|
1728 |
+
"eval_f1_micro": 0.8562891131532955,
|
1729 |
+
"eval_loss": 0.08144387602806091,
|
1730 |
+
"eval_roc_auc": 0.9079250010327448,
|
1731 |
+
"eval_runtime": 693.924,
|
1732 |
+
"eval_samples_per_second": 4.13,
|
1733 |
+
"eval_steps_per_second": 0.13,
|
1734 |
+
"learning_rate": 1.0000000000000002e-06,
|
1735 |
+
"step": 29044
|
1736 |
+
},
|
1737 |
+
{
|
1738 |
+
"epoch": 107.0,
|
1739 |
+
"eval_accuracy": 0.5635031402651779,
|
1740 |
+
"eval_f1_macro": 0.7490061784836257,
|
1741 |
+
"eval_f1_micro": 0.8557823129251702,
|
1742 |
+
"eval_loss": 0.08168599754571915,
|
1743 |
+
"eval_roc_auc": 0.9057833825242961,
|
1744 |
+
"eval_runtime": 697.3665,
|
1745 |
+
"eval_samples_per_second": 4.11,
|
1746 |
+
"eval_steps_per_second": 0.129,
|
1747 |
+
"learning_rate": 1.0000000000000002e-06,
|
1748 |
+
"step": 29318
|
1749 |
+
},
|
1750 |
+
{
|
1751 |
+
"epoch": 107.66,
|
1752 |
+
"learning_rate": 1.0000000000000002e-06,
|
1753 |
+
"loss": 0.0854,
|
1754 |
+
"step": 29500
|
1755 |
+
},
|
1756 |
+
{
|
1757 |
+
"epoch": 108.0,
|
1758 |
+
"eval_accuracy": 0.5642009769713887,
|
1759 |
+
"eval_f1_macro": 0.7569164155873785,
|
1760 |
+
"eval_f1_micro": 0.8569341425039335,
|
1761 |
+
"eval_loss": 0.08160943537950516,
|
1762 |
+
"eval_roc_auc": 0.9094149373671999,
|
1763 |
+
"eval_runtime": 692.7895,
|
1764 |
+
"eval_samples_per_second": 4.137,
|
1765 |
+
"eval_steps_per_second": 0.13,
|
1766 |
+
"learning_rate": 1.0000000000000002e-06,
|
1767 |
+
"step": 29592
|
1768 |
+
},
|
1769 |
+
{
|
1770 |
+
"epoch": 109.0,
|
1771 |
+
"eval_accuracy": 0.5652477320307048,
|
1772 |
+
"eval_f1_macro": 0.7558234898859766,
|
1773 |
+
"eval_f1_micro": 0.8573669467787115,
|
1774 |
+
"eval_loss": 0.08135490119457245,
|
1775 |
+
"eval_roc_auc": 0.9106934783376293,
|
1776 |
+
"eval_runtime": 695.7903,
|
1777 |
+
"eval_samples_per_second": 4.119,
|
1778 |
+
"eval_steps_per_second": 0.129,
|
1779 |
+
"learning_rate": 1.0000000000000002e-06,
|
1780 |
+
"step": 29866
|
1781 |
+
},
|
1782 |
+
{
|
1783 |
+
"epoch": 109.49,
|
1784 |
+
"learning_rate": 1.0000000000000002e-06,
|
1785 |
+
"loss": 0.0854,
|
1786 |
+
"step": 30000
|
1787 |
+
},
|
1788 |
+
{
|
1789 |
+
"epoch": 110.0,
|
1790 |
+
"eval_accuracy": 0.5638520586182834,
|
1791 |
+
"eval_f1_macro": 0.7564667466516518,
|
1792 |
+
"eval_f1_micro": 0.8577973962116557,
|
1793 |
+
"eval_loss": 0.08132224529981613,
|
1794 |
+
"eval_roc_auc": 0.9118372864958749,
|
1795 |
+
"eval_runtime": 682.0266,
|
1796 |
+
"eval_samples_per_second": 4.202,
|
1797 |
+
"eval_steps_per_second": 0.132,
|
1798 |
+
"learning_rate": 1.0000000000000002e-06,
|
1799 |
+
"step": 30140
|
1800 |
+
},
|
1801 |
+
{
|
1802 |
+
"epoch": 111.0,
|
1803 |
+
"eval_accuracy": 0.5638520586182834,
|
1804 |
+
"eval_f1_macro": 0.7579481466413214,
|
1805 |
+
"eval_f1_micro": 0.8576381122791322,
|
1806 |
+
"eval_loss": 0.0814104825258255,
|
1807 |
+
"eval_roc_auc": 0.911498485615307,
|
1808 |
+
"eval_runtime": 686.5544,
|
1809 |
+
"eval_samples_per_second": 4.174,
|
1810 |
+
"eval_steps_per_second": 0.131,
|
1811 |
+
"learning_rate": 1.0000000000000002e-07,
|
1812 |
+
"step": 30414
|
1813 |
+
},
|
1814 |
+
{
|
1815 |
+
"epoch": 111.31,
|
1816 |
+
"learning_rate": 1.0000000000000002e-07,
|
1817 |
+
"loss": 0.0851,
|
1818 |
+
"step": 30500
|
1819 |
+
},
|
1820 |
+
{
|
1821 |
+
"epoch": 112.0,
|
1822 |
+
"eval_accuracy": 0.5631542219120725,
|
1823 |
+
"eval_f1_macro": 0.7575801018117111,
|
1824 |
+
"eval_f1_micro": 0.8580721134974485,
|
1825 |
+
"eval_loss": 0.08170080929994583,
|
1826 |
+
"eval_roc_auc": 0.91075603440298,
|
1827 |
+
"eval_runtime": 746.778,
|
1828 |
+
"eval_samples_per_second": 3.838,
|
1829 |
+
"eval_steps_per_second": 0.121,
|
1830 |
+
"learning_rate": 1.0000000000000002e-07,
|
1831 |
+
"step": 30688
|
1832 |
+
},
|
1833 |
+
{
|
1834 |
+
"epoch": 113.0,
|
1835 |
+
"eval_accuracy": 0.5614096301465457,
|
1836 |
+
"eval_f1_macro": 0.7562692232145891,
|
1837 |
+
"eval_f1_micro": 0.8583059164668487,
|
1838 |
+
"eval_loss": 0.081505186855793,
|
1839 |
+
"eval_roc_auc": 0.912813663063945,
|
1840 |
+
"eval_runtime": 687.4479,
|
1841 |
+
"eval_samples_per_second": 4.169,
|
1842 |
+
"eval_steps_per_second": 0.131,
|
1843 |
+
"learning_rate": 1.0000000000000002e-07,
|
1844 |
+
"step": 30962
|
1845 |
+
},
|
1846 |
+
{
|
1847 |
+
"epoch": 113.14,
|
1848 |
+
"learning_rate": 1.0000000000000002e-07,
|
1849 |
+
"loss": 0.0848,
|
1850 |
+
"step": 31000
|
1851 |
+
},
|
1852 |
+
{
|
1853 |
+
"epoch": 114.0,
|
1854 |
+
"eval_accuracy": 0.5655966503838102,
|
1855 |
+
"eval_f1_macro": 0.7560403631861836,
|
1856 |
+
"eval_f1_micro": 0.8563654914408797,
|
1857 |
+
"eval_loss": 0.08186182379722595,
|
1858 |
+
"eval_roc_auc": 0.9061407756963643,
|
1859 |
+
"eval_runtime": 731.2304,
|
1860 |
+
"eval_samples_per_second": 3.919,
|
1861 |
+
"eval_steps_per_second": 0.123,
|
1862 |
+
"learning_rate": 1.0000000000000002e-07,
|
1863 |
+
"step": 31236
|
1864 |
+
},
|
1865 |
+
{
|
1866 |
+
"epoch": 114.0,
|
1867 |
+
"learning_rate": 1.0000000000000002e-07,
|
1868 |
+
"step": 31236,
|
1869 |
+
"total_flos": 1.4819663325231928e+21,
|
1870 |
+
"train_loss": 0.033337813025782426,
|
1871 |
+
"train_runtime": 131818.0427,
|
1872 |
+
"train_samples_per_second": 9.311,
|
1873 |
+
"train_steps_per_second": 0.291
|
1874 |
}
|
1875 |
],
|
1876 |
"logging_steps": 500,
|
1877 |
+
"max_steps": 38360,
|
1878 |
+
"num_train_epochs": 140,
|
1879 |
"save_steps": 500,
|
1880 |
+
"total_flos": 1.4819663325231928e+21,
|
1881 |
"trial_name": null,
|
1882 |
"trial_params": null
|
1883 |
}
|