Commit
•
0035cbd
1
Parent(s):
f35a4f9
Update README.md
Browse files
README.md
CHANGED
|
@@ -153,10 +153,10 @@ Total train batch size (w. parallel, distributed & accumulation) = 4
|
|
| 153 |
Gradient Accumulation steps = 2
|
| 154 |
Total optimization steps = 19570
|
| 155 |
|
| 156 |
-
Step Training Loss
|
| 157 |
-
300
|
| 158 |
-
600
|
| 159 |
-
900
|
| 160 |
1200 0.077800 0.126345 0.815400 0.865376 0.839645 0.967849
|
| 161 |
1500 0.074100 0.148207 0.779274 0.895914 0.833533 0.960184
|
| 162 |
1800 0.059500 0.116634 0.830829 0.868172 0.849090 0.969342
|
|
@@ -179,7 +179,7 @@ Step Training Loss Validation Loss Precision Recall F1 Accuracy
|
|
| 179 |
6900 0.016100 0.143160 0.789938 0.904946 0.843540 0.968245
|
| 180 |
7200 0.017000 0.145755 0.823274 0.897634 0.858848 0.969037
|
| 181 |
7500 0.012100 0.159342 0.825694 0.883226 0.853491 0.967468
|
| 182 |
-
7800 0.013800 0.194886
|
| 183 |
8100 0.008000 0.140271 0.829914 0.896129 0.861752 0.971567
|
| 184 |
8400 0.010300 0.143318 0.826844 0.908817 0.865895 0.973578
|
| 185 |
8700 0.015000 0.143392 0.847336 0.889247 0.867786 0.973365
|
|
@@ -187,40 +187,40 @@ Step Training Loss Validation Loss Precision Recall F1 Accuracy
|
|
| 187 |
9300 0.011800 0.138747 0.827133 0.894194 0.859357 0.971673
|
| 188 |
9600 0.008500 0.159490 0.837030 0.909032 0.871546 0.970028
|
| 189 |
9900 0.010700 0.159249 0.846692 0.910968 0.877655 0.970546
|
| 190 |
-
10200 0.008100
|
| 191 |
-
10500 0.004800
|
| 192 |
-
10800 0.010700
|
| 193 |
-
11100 0.003800
|
| 194 |
-
11400 0.009700
|
| 195 |
-
11700 0.008900
|
| 196 |
-
12000 0.006400
|
| 197 |
-
12300 0.007100
|
| 198 |
-
12600 0.015800
|
| 199 |
-
12900 0.006600
|
| 200 |
-
13200 0.006800
|
| 201 |
-
13500 0.003200
|
| 202 |
-
13800 0.003600
|
| 203 |
-
14100 0.003500
|
| 204 |
-
14400 0.003300
|
| 205 |
-
14700 0.002500
|
| 206 |
-
15000 0.003400
|
| 207 |
-
15300 0.006000
|
| 208 |
-
15600 0.002400
|
| 209 |
-
15900 0.004100
|
| 210 |
-
16200 0.002600
|
| 211 |
-
16500 0.002100
|
| 212 |
-
16800 0.002900
|
| 213 |
-
17100 0.001600
|
| 214 |
-
17400 0.003900
|
| 215 |
-
17700 0.002700
|
| 216 |
-
18000 0.001300
|
| 217 |
|
| 218 |
-
18300 0.000800
|
| 219 |
|
| 220 |
-
18600 0.002700
|
| 221 |
-
18900 0.001600
|
| 222 |
-
19200 0.002300
|
| 223 |
-
19500 0.001800
|
| 224 |
````
|
| 225 |
|
| 226 |
### Validation metrics by Named Entity
|
|
|
|
| 153 |
Gradient Accumulation steps = 2
|
| 154 |
Total optimization steps = 19570
|
| 155 |
|
| 156 |
+
Step Training Loss Validation Loss Precision Recall F1 Accuracy
|
| 157 |
+
300 0.127600 0.178613 0.722909 0.741720 0.732194 0.948802
|
| 158 |
+
600 0.088200 0.136965 0.733636 0.867742 0.795074 0.963079
|
| 159 |
+
900 0.078000 0.128858 0.791912 0.838065 0.814335 0.965243
|
| 160 |
1200 0.077800 0.126345 0.815400 0.865376 0.839645 0.967849
|
| 161 |
1500 0.074100 0.148207 0.779274 0.895914 0.833533 0.960184
|
| 162 |
1800 0.059500 0.116634 0.830829 0.868172 0.849090 0.969342
|
|
|
|
| 179 |
6900 0.016100 0.143160 0.789938 0.904946 0.843540 0.968245
|
| 180 |
7200 0.017000 0.145755 0.823274 0.897634 0.858848 0.969037
|
| 181 |
7500 0.012100 0.159342 0.825694 0.883226 0.853491 0.967468
|
| 182 |
+
7800 0.013800 0.194886 0.861237 0.859570 0.860403 0.964771
|
| 183 |
8100 0.008000 0.140271 0.829914 0.896129 0.861752 0.971567
|
| 184 |
8400 0.010300 0.143318 0.826844 0.908817 0.865895 0.973578
|
| 185 |
8700 0.015000 0.143392 0.847336 0.889247 0.867786 0.973365
|
|
|
|
| 187 |
9300 0.011800 0.138747 0.827133 0.894194 0.859357 0.971673
|
| 188 |
9600 0.008500 0.159490 0.837030 0.909032 0.871546 0.970028
|
| 189 |
9900 0.010700 0.159249 0.846692 0.910968 0.877655 0.970546
|
| 190 |
+
10200 0.008100 0.170069 0.848288 0.900645 0.873683 0.969113
|
| 191 |
+
10500 0.004800 0.183795 0.860317 0.899355 0.879403 0.969570
|
| 192 |
+
10800 0.010700 0.157024 0.837838 0.906667 0.870894 0.971094
|
| 193 |
+
11100 0.003800 0.164286 0.845312 0.880215 0.862410 0.970744
|
| 194 |
+
11400 0.009700 0.204025 0.884294 0.887527 0.885907 0.968854
|
| 195 |
+
11700 0.008900 0.162819 0.829415 0.887742 0.857588 0.970530
|
| 196 |
+
12000 0.006400 0.164296 0.852666 0.901075 0.876202 0.971414
|
| 197 |
+
12300 0.007100 0.143367 0.852959 0.895699 0.873807 0.973669
|
| 198 |
+
12600 0.015800 0.153383 0.859224 0.900430 0.879345 0.972679
|
| 199 |
+
12900 0.006600 0.173447 0.869954 0.899140 0.884306 0.970927
|
| 200 |
+
13200 0.006800 0.163234 0.856849 0.897204 0.876563 0.971795
|
| 201 |
+
13500 0.003200 0.167164 0.850867 0.907957 0.878485 0.971231
|
| 202 |
+
13800 0.003600 0.148950 0.867801 0.910538 0.888656 0.976961
|
| 203 |
+
14100 0.003500 0.155691 0.847621 0.907957 0.876752 0.974127
|
| 204 |
+
14400 0.003300 0.157672 0.846553 0.911183 0.877680 0.974584
|
| 205 |
+
14700 0.002500 0.169965 0.847804 0.917634 0.881338 0.973045
|
| 206 |
+
15000 0.003400 0.177099 0.842199 0.912473 0.875929 0.971155
|
| 207 |
+
15300 0.006000 0.164151 0.848928 0.911183 0.878954 0.973258
|
| 208 |
+
15600 0.002400 0.174305 0.847437 0.906667 0.876052 0.971765
|
| 209 |
+
15900 0.004100 0.174561 0.852929 0.907957 0.879583 0.972907
|
| 210 |
+
16200 0.002600 0.172626 0.843263 0.907097 0.874016 0.972100
|
| 211 |
+
16500 0.002100 0.185302 0.841108 0.907312 0.872957 0.970485
|
| 212 |
+
16800 0.002900 0.175638 0.840557 0.909247 0.873554 0.971704
|
| 213 |
+
17100 0.001600 0.178750 0.857056 0.906452 0.881062 0.971765
|
| 214 |
+
17400 0.003900 0.188910 0.853619 0.907957 0.879950 0.970835
|
| 215 |
+
17700 0.002700 0.180822 0.864699 0.907097 0.885390 0.972283
|
| 216 |
+
18000 0.001300 0.179974 0.868150 0.906237 0.886785 0.973060
|
| 217 |
|
| 218 |
+
18300 0.000800 0.188032 0.881022 0.904516 0.892615 0.972572
|
| 219 |
|
| 220 |
+
18600 0.002700 0.183266 0.868601 0.901290 0.884644 0.972298
|
| 221 |
+
18900 0.001600 0.180301 0.862041 0.903011 0.882050 0.972344
|
| 222 |
+
19200 0.002300 0.183432 0.855370 0.904301 0.879155 0.971109
|
| 223 |
+
19500 0.001800 0.183381 0.854501 0.904301 0.878696 0.971186
|
| 224 |
````
|
| 225 |
|
| 226 |
### Validation metrics by Named Entity
|