F1 0.7753 @ 2026-07-07T02:33:14
Browse files- README.md +18 -18
- best_model.pt +1 -1
- eval_metrics.json +785 -239
- model_meta.json +34 -34
- training_report.png +2 -2
README.md
CHANGED
|
@@ -10,18 +10,18 @@ a variable-size selection instead of a fixed top-k.
|
|
| 10 |
|
| 11 |
|
| 12 |
|
| 13 |
-
## Test metrics (threshold 0.
|
| 14 |
|
| 15 |
-
micro-P **0.
|
| 16 |
-
micro-F1 **0.
|
| 17 |
-
best epoch
|
| 18 |
|
| 19 |
| bench | groups | P | R | F1 |
|
| 20 |
|---|---|---|---|---|
|
| 21 |
-
| gold_easy |
|
| 22 |
-
| gold_medium | 2 |
|
| 23 |
-
| gold_hard | 5 |
|
| 24 |
-
| llm_held_out | 29 | 0.
|
| 25 |
|
| 26 |
`gold_*` tiers are hand-curated deterministic labels (easy = section how-to,
|
| 27 |
medium = single-doc, hard = table/matrix incl. reverse lookups); rephrasing
|
|
@@ -32,17 +32,17 @@ types. `llm_held_out` is strict generalization on unseen questions.
|
|
| 32 |
|
| 33 |
```json
|
| 34 |
{
|
| 35 |
-
"train_groups":
|
| 36 |
-
"gold_train_groups":
|
| 37 |
-
"test_groups":
|
| 38 |
-
"gold_test_groups":
|
| 39 |
-
"train_candidates":
|
| 40 |
-
"train_pass":
|
| 41 |
-
"train_fail":
|
| 42 |
-
"pass_ratio": 0.
|
| 43 |
-
"avg_candidates_per_train_group": 23.
|
| 44 |
"avg_pass_per_train_group": 4.7
|
| 45 |
}
|
| 46 |
```
|
| 47 |
|
| 48 |
-
Trained 2026-07-
|
|
|
|
| 10 |
|
| 11 |
|
| 12 |
|
| 13 |
+
## Test metrics (threshold 0.942)
|
| 14 |
|
| 15 |
+
micro-P **0.9345** · micro-R **0.6624** ·
|
| 16 |
+
micro-F1 **0.7753** · exact-set **0.2** ·
|
| 17 |
+
best epoch 53
|
| 18 |
|
| 19 |
| bench | groups | P | R | F1 |
|
| 20 |
|---|---|---|---|---|
|
| 21 |
+
| gold_easy | 9 | 0.8571 | 0.2069 | 0.3333 |
|
| 22 |
+
| gold_medium | 2 | 1.0 | 0.9211 | 0.9589 |
|
| 23 |
+
| gold_hard | 5 | 1.0 | 0.9667 | 0.9831 |
|
| 24 |
+
| llm_held_out | 29 | 0.7436 | 0.3625 | 0.4874 |
|
| 25 |
|
| 26 |
`gold_*` tiers are hand-curated deterministic labels (easy = section how-to,
|
| 27 |
medium = single-doc, hard = table/matrix incl. reverse lookups); rephrasing
|
|
|
|
| 32 |
|
| 33 |
```json
|
| 34 |
{
|
| 35 |
+
"train_groups": 1184,
|
| 36 |
+
"gold_train_groups": 48,
|
| 37 |
+
"test_groups": 45,
|
| 38 |
+
"gold_test_groups": 16,
|
| 39 |
+
"train_candidates": 27544,
|
| 40 |
+
"train_pass": 5567,
|
| 41 |
+
"train_fail": 21977,
|
| 42 |
+
"pass_ratio": 0.202,
|
| 43 |
+
"avg_candidates_per_train_group": 23.3,
|
| 44 |
"avg_pass_per_train_group": 4.7
|
| 45 |
}
|
| 46 |
```
|
| 47 |
|
| 48 |
+
Trained 2026-07-07T02:33:14 · source: https://github.com/s3777091/recursive_models
|
best_model.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 7633333
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:afdf752939b655c9329345be330056a9899c51e1413d6e14cd740cafb722d104
|
| 3 |
size 7633333
|
eval_metrics.json
CHANGED
|
@@ -1,308 +1,854 @@
|
|
| 1 |
{
|
| 2 |
-
"threshold": 0.
|
| 3 |
-
"micro_precision": 0.
|
| 4 |
-
"micro_recall": 0.
|
| 5 |
-
"micro_f1": 0.
|
| 6 |
-
"macro_f1": 0.
|
| 7 |
-
"exact_set_match": 0.
|
| 8 |
-
"avg_selected":
|
| 9 |
-
"avg_true_pass": 5.
|
| 10 |
-
"num_groups":
|
| 11 |
-
"best_epoch":
|
| 12 |
"history": [
|
| 13 |
{
|
| 14 |
"epoch": 1,
|
| 15 |
-
"loss":
|
| 16 |
-
"lr": 0.
|
| 17 |
"threshold": 0.805,
|
| 18 |
-
"micro_precision": 0.
|
| 19 |
-
"micro_recall": 0.
|
| 20 |
-
"micro_f1": 0.
|
| 21 |
-
"macro_f1": 0.
|
| 22 |
-
"exact_set_match": 0.
|
| 23 |
-
"avg_selected":
|
| 24 |
-
"avg_true_pass": 5.
|
| 25 |
-
"num_groups":
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"epoch": 2,
|
| 29 |
-
"loss":
|
| 30 |
-
"lr": 0.
|
| 31 |
-
"threshold": 0.
|
| 32 |
-
"micro_precision": 0.
|
| 33 |
-
"micro_recall": 0.
|
| 34 |
-
"micro_f1": 0.
|
| 35 |
-
"macro_f1": 0.
|
| 36 |
-
"exact_set_match": 0.
|
| 37 |
-
"avg_selected":
|
| 38 |
-
"avg_true_pass": 5.
|
| 39 |
-
"num_groups":
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 3,
|
| 43 |
-
"loss":
|
| 44 |
-
"lr": 0.
|
| 45 |
-
"threshold": 0.
|
| 46 |
-
"micro_precision": 0.
|
| 47 |
-
"micro_recall": 0.
|
| 48 |
-
"micro_f1": 0.
|
| 49 |
-
"macro_f1": 0.
|
| 50 |
-
"exact_set_match": 0.
|
| 51 |
-
"avg_selected": 1.
|
| 52 |
-
"avg_true_pass": 5.
|
| 53 |
-
"num_groups":
|
| 54 |
},
|
| 55 |
{
|
| 56 |
"epoch": 4,
|
| 57 |
-
"loss":
|
| 58 |
-
"lr": 0.
|
| 59 |
-
"threshold": 0.
|
| 60 |
-
"micro_precision": 0.
|
| 61 |
-
"micro_recall": 0.
|
| 62 |
-
"micro_f1": 0.
|
| 63 |
-
"macro_f1": 0.
|
| 64 |
-
"exact_set_match": 0.
|
| 65 |
-
"avg_selected":
|
| 66 |
-
"avg_true_pass": 5.
|
| 67 |
-
"num_groups":
|
| 68 |
},
|
| 69 |
{
|
| 70 |
"epoch": 5,
|
| 71 |
-
"loss":
|
| 72 |
-
"lr": 0.
|
| 73 |
-
"threshold": 0.
|
| 74 |
-
"micro_precision": 0.
|
| 75 |
-
"micro_recall": 0.
|
| 76 |
-
"micro_f1": 0.
|
| 77 |
-
"macro_f1": 0.
|
| 78 |
-
"exact_set_match": 0.
|
| 79 |
-
"avg_selected":
|
| 80 |
-
"avg_true_pass": 5.
|
| 81 |
-
"num_groups":
|
| 82 |
},
|
| 83 |
{
|
| 84 |
"epoch": 6,
|
| 85 |
-
"loss":
|
| 86 |
-
"lr": 0.
|
| 87 |
-
"threshold": 0.
|
| 88 |
-
"micro_precision": 0.
|
| 89 |
-
"micro_recall": 0.
|
| 90 |
-
"micro_f1": 0.
|
| 91 |
-
"macro_f1": 0.
|
| 92 |
-
"exact_set_match": 0.
|
| 93 |
-
"avg_selected":
|
| 94 |
-
"avg_true_pass": 5.
|
| 95 |
-
"num_groups":
|
| 96 |
},
|
| 97 |
{
|
| 98 |
"epoch": 7,
|
| 99 |
-
"loss":
|
| 100 |
-
"lr": 0.
|
| 101 |
-
"threshold": 0.
|
| 102 |
-
"micro_precision": 0.
|
| 103 |
-
"micro_recall": 0.
|
| 104 |
-
"micro_f1": 0.
|
| 105 |
-
"macro_f1": 0.
|
| 106 |
-
"exact_set_match": 0.
|
| 107 |
-
"avg_selected":
|
| 108 |
-
"avg_true_pass": 5.
|
| 109 |
-
"num_groups":
|
| 110 |
},
|
| 111 |
{
|
| 112 |
"epoch": 8,
|
| 113 |
-
"loss":
|
| 114 |
-
"lr": 0.
|
| 115 |
-
"threshold": 0.
|
| 116 |
-
"micro_precision": 0.
|
| 117 |
-
"micro_recall": 0.
|
| 118 |
-
"micro_f1": 0.
|
| 119 |
-
"macro_f1": 0.
|
| 120 |
-
"exact_set_match": 0.
|
| 121 |
-
"avg_selected":
|
| 122 |
-
"avg_true_pass": 5.
|
| 123 |
-
"num_groups":
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"epoch": 9,
|
| 127 |
-
"loss":
|
| 128 |
-
"lr": 0.
|
| 129 |
-
"threshold": 0.
|
| 130 |
-
"micro_precision": 0.
|
| 131 |
-
"micro_recall": 0.
|
| 132 |
-
"micro_f1": 0.
|
| 133 |
-
"macro_f1": 0.
|
| 134 |
-
"exact_set_match": 0.
|
| 135 |
-
"avg_selected":
|
| 136 |
-
"avg_true_pass": 5.
|
| 137 |
-
"num_groups":
|
| 138 |
},
|
| 139 |
{
|
| 140 |
"epoch": 10,
|
| 141 |
-
"loss":
|
| 142 |
-
"lr": 0.
|
| 143 |
-
"threshold": 0.
|
| 144 |
-
"micro_precision": 0.
|
| 145 |
-
"micro_recall": 0.
|
| 146 |
-
"micro_f1": 0.
|
| 147 |
-
"macro_f1": 0.
|
| 148 |
-
"exact_set_match": 0.
|
| 149 |
-
"avg_selected":
|
| 150 |
-
"avg_true_pass": 5.
|
| 151 |
-
"num_groups":
|
| 152 |
},
|
| 153 |
{
|
| 154 |
"epoch": 11,
|
| 155 |
-
"loss":
|
| 156 |
-
"lr": 0.
|
| 157 |
-
"threshold": 0.
|
| 158 |
-
"micro_precision": 0.
|
| 159 |
-
"micro_recall": 0.
|
| 160 |
-
"micro_f1": 0.
|
| 161 |
-
"macro_f1": 0.
|
| 162 |
-
"exact_set_match": 0.
|
| 163 |
-
"avg_selected":
|
| 164 |
-
"avg_true_pass": 5.
|
| 165 |
-
"num_groups":
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 12,
|
| 169 |
-
"loss":
|
| 170 |
-
"lr": 0.
|
| 171 |
-
"threshold": 0.
|
| 172 |
-
"micro_precision": 0.
|
| 173 |
-
"micro_recall": 0.
|
| 174 |
-
"micro_f1": 0.
|
| 175 |
-
"macro_f1": 0.
|
| 176 |
-
"exact_set_match": 0.
|
| 177 |
-
"avg_selected":
|
| 178 |
-
"avg_true_pass": 5.
|
| 179 |
-
"num_groups":
|
| 180 |
},
|
| 181 |
{
|
| 182 |
"epoch": 13,
|
| 183 |
-
"loss":
|
| 184 |
-
"lr": 0.
|
| 185 |
-
"threshold": 0.
|
| 186 |
-
"micro_precision": 0.
|
| 187 |
-
"micro_recall": 0.
|
| 188 |
-
"micro_f1": 0.
|
| 189 |
-
"macro_f1": 0.
|
| 190 |
-
"exact_set_match": 0.
|
| 191 |
-
"avg_selected":
|
| 192 |
-
"avg_true_pass": 5.
|
| 193 |
-
"num_groups":
|
| 194 |
},
|
| 195 |
{
|
| 196 |
"epoch": 14,
|
| 197 |
-
"loss":
|
| 198 |
-
"lr": 0.
|
| 199 |
-
"threshold": 0.
|
| 200 |
-
"micro_precision": 0.
|
| 201 |
-
"micro_recall": 0.
|
| 202 |
-
"micro_f1": 0.
|
| 203 |
-
"macro_f1": 0.
|
| 204 |
-
"exact_set_match": 0.
|
| 205 |
-
"avg_selected":
|
| 206 |
-
"avg_true_pass": 5.
|
| 207 |
-
"num_groups":
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 15,
|
| 211 |
-
"loss":
|
| 212 |
-
"lr": 0.
|
| 213 |
-
"threshold": 0.
|
| 214 |
-
"micro_precision": 0.
|
| 215 |
-
"micro_recall": 0.
|
| 216 |
-
"micro_f1": 0.
|
| 217 |
-
"macro_f1": 0.
|
| 218 |
-
"exact_set_match": 0.
|
| 219 |
-
"avg_selected":
|
| 220 |
-
"avg_true_pass": 5.
|
| 221 |
-
"num_groups":
|
| 222 |
},
|
| 223 |
{
|
| 224 |
"epoch": 16,
|
| 225 |
-
"loss":
|
| 226 |
-
"lr": 0.
|
| 227 |
-
"threshold": 0.
|
| 228 |
-
"micro_precision": 0.
|
| 229 |
-
"micro_recall": 0.
|
| 230 |
-
"micro_f1": 0.
|
| 231 |
-
"macro_f1": 0.
|
| 232 |
-
"exact_set_match": 0.
|
| 233 |
-
"avg_selected":
|
| 234 |
-
"avg_true_pass": 5.
|
| 235 |
-
"num_groups":
|
| 236 |
},
|
| 237 |
{
|
| 238 |
"epoch": 17,
|
| 239 |
-
"loss":
|
| 240 |
-
"lr": 0.
|
| 241 |
-
"threshold": 0.
|
| 242 |
-
"micro_precision": 0.
|
| 243 |
-
"micro_recall": 0.
|
| 244 |
-
"micro_f1": 0.
|
| 245 |
-
"macro_f1": 0.
|
| 246 |
-
"exact_set_match": 0.
|
| 247 |
-
"avg_selected":
|
| 248 |
-
"avg_true_pass": 5.
|
| 249 |
-
"num_groups":
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 18,
|
| 253 |
-
"loss":
|
| 254 |
-
"lr": 0.
|
| 255 |
-
"threshold": 0.
|
| 256 |
-
"micro_precision": 0.
|
| 257 |
-
"micro_recall": 0.
|
| 258 |
-
"micro_f1": 0.
|
| 259 |
-
"macro_f1": 0.
|
| 260 |
-
"exact_set_match": 0.
|
| 261 |
-
"avg_selected":
|
| 262 |
-
"avg_true_pass": 5.
|
| 263 |
-
"num_groups":
|
| 264 |
},
|
| 265 |
{
|
| 266 |
"epoch": 19,
|
| 267 |
-
"loss":
|
| 268 |
-
"lr": 0.
|
| 269 |
-
"threshold": 0.
|
| 270 |
-
"micro_precision": 0.
|
| 271 |
-
"micro_recall": 0.
|
| 272 |
-
"micro_f1": 0.
|
| 273 |
-
"macro_f1": 0.
|
| 274 |
-
"exact_set_match": 0.
|
| 275 |
-
"avg_selected":
|
| 276 |
-
"avg_true_pass": 5.
|
| 277 |
-
"num_groups":
|
| 278 |
},
|
| 279 |
{
|
| 280 |
"epoch": 20,
|
| 281 |
-
"loss":
|
| 282 |
-
"lr": 0.
|
| 283 |
-
"threshold": 0.
|
| 284 |
-
"micro_precision": 0.
|
| 285 |
-
"micro_recall": 0.
|
| 286 |
-
"micro_f1": 0.
|
| 287 |
-
"macro_f1": 0.
|
| 288 |
-
"exact_set_match": 0.
|
| 289 |
-
"avg_selected":
|
| 290 |
-
"avg_true_pass": 5.
|
| 291 |
-
"num_groups":
|
| 292 |
},
|
| 293 |
{
|
| 294 |
"epoch": 21,
|
| 295 |
-
"loss":
|
| 296 |
-
"lr": 0.
|
| 297 |
-
"threshold": 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 298 |
"micro_precision": 0.8889,
|
| 299 |
-
"micro_recall": 0.
|
| 300 |
-
"micro_f1": 0.
|
| 301 |
-
"macro_f1": 0.
|
| 302 |
-
"exact_set_match": 0.
|
| 303 |
-
"avg_selected":
|
| 304 |
-
"avg_true_pass": 5.
|
| 305 |
-
"num_groups":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 306 |
}
|
| 307 |
]
|
| 308 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"threshold": 0.942,
|
| 3 |
+
"micro_precision": 0.9345,
|
| 4 |
+
"micro_recall": 0.6624,
|
| 5 |
+
"micro_f1": 0.7753,
|
| 6 |
+
"macro_f1": 0.502,
|
| 7 |
+
"exact_set_match": 0.2,
|
| 8 |
+
"avg_selected": 3.73,
|
| 9 |
+
"avg_true_pass": 5.27,
|
| 10 |
+
"num_groups": 45,
|
| 11 |
+
"best_epoch": 53,
|
| 12 |
"history": [
|
| 13 |
{
|
| 14 |
"epoch": 1,
|
| 15 |
+
"loss": 146.576,
|
| 16 |
+
"lr": 0.000113,
|
| 17 |
"threshold": 0.805,
|
| 18 |
+
"micro_precision": 0.5294,
|
| 19 |
+
"micro_recall": 0.1519,
|
| 20 |
+
"micro_f1": 0.2361,
|
| 21 |
+
"macro_f1": 0.3534,
|
| 22 |
+
"exact_set_match": 0.0889,
|
| 23 |
+
"avg_selected": 1.51,
|
| 24 |
+
"avg_true_pass": 5.27,
|
| 25 |
+
"num_groups": 45
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"epoch": 2,
|
| 29 |
+
"loss": 114.155,
|
| 30 |
+
"lr": 0.000113,
|
| 31 |
+
"threshold": 0.712,
|
| 32 |
+
"micro_precision": 0.4571,
|
| 33 |
+
"micro_recall": 0.135,
|
| 34 |
+
"micro_f1": 0.2085,
|
| 35 |
+
"macro_f1": 0.2734,
|
| 36 |
+
"exact_set_match": 0.0889,
|
| 37 |
+
"avg_selected": 1.56,
|
| 38 |
+
"avg_true_pass": 5.27,
|
| 39 |
+
"num_groups": 45
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 3,
|
| 43 |
+
"loss": 109.051,
|
| 44 |
+
"lr": 0.000113,
|
| 45 |
+
"threshold": 0.793,
|
| 46 |
+
"micro_precision": 0.5,
|
| 47 |
+
"micro_recall": 0.1688,
|
| 48 |
+
"micro_f1": 0.2524,
|
| 49 |
+
"macro_f1": 0.2893,
|
| 50 |
+
"exact_set_match": 0.0667,
|
| 51 |
+
"avg_selected": 1.78,
|
| 52 |
+
"avg_true_pass": 5.27,
|
| 53 |
+
"num_groups": 45
|
| 54 |
},
|
| 55 |
{
|
| 56 |
"epoch": 4,
|
| 57 |
+
"loss": 104.063,
|
| 58 |
+
"lr": 0.000113,
|
| 59 |
+
"threshold": 0.85,
|
| 60 |
+
"micro_precision": 0.5402,
|
| 61 |
+
"micro_recall": 0.1983,
|
| 62 |
+
"micro_f1": 0.2901,
|
| 63 |
+
"macro_f1": 0.3327,
|
| 64 |
+
"exact_set_match": 0.0667,
|
| 65 |
+
"avg_selected": 1.93,
|
| 66 |
+
"avg_true_pass": 5.27,
|
| 67 |
+
"num_groups": 45
|
| 68 |
},
|
| 69 |
{
|
| 70 |
"epoch": 5,
|
| 71 |
+
"loss": 103.158,
|
| 72 |
+
"lr": 0.000113,
|
| 73 |
+
"threshold": 0.834,
|
| 74 |
+
"micro_precision": 0.5647,
|
| 75 |
+
"micro_recall": 0.2025,
|
| 76 |
+
"micro_f1": 0.2981,
|
| 77 |
+
"macro_f1": 0.3194,
|
| 78 |
+
"exact_set_match": 0.0889,
|
| 79 |
+
"avg_selected": 1.89,
|
| 80 |
+
"avg_true_pass": 5.27,
|
| 81 |
+
"num_groups": 45
|
| 82 |
},
|
| 83 |
{
|
| 84 |
"epoch": 6,
|
| 85 |
+
"loss": 100.432,
|
| 86 |
+
"lr": 0.000113,
|
| 87 |
+
"threshold": 0.881,
|
| 88 |
+
"micro_precision": 0.5488,
|
| 89 |
+
"micro_recall": 0.1899,
|
| 90 |
+
"micro_f1": 0.2821,
|
| 91 |
+
"macro_f1": 0.2903,
|
| 92 |
+
"exact_set_match": 0.0667,
|
| 93 |
+
"avg_selected": 1.82,
|
| 94 |
+
"avg_true_pass": 5.27,
|
| 95 |
+
"num_groups": 45
|
| 96 |
},
|
| 97 |
{
|
| 98 |
"epoch": 7,
|
| 99 |
+
"loss": 99.541,
|
| 100 |
+
"lr": 0.000113,
|
| 101 |
+
"threshold": 0.894,
|
| 102 |
+
"micro_precision": 0.5357,
|
| 103 |
+
"micro_recall": 0.1899,
|
| 104 |
+
"micro_f1": 0.2804,
|
| 105 |
+
"macro_f1": 0.2836,
|
| 106 |
+
"exact_set_match": 0.0667,
|
| 107 |
+
"avg_selected": 1.87,
|
| 108 |
+
"avg_true_pass": 5.27,
|
| 109 |
+
"num_groups": 45
|
| 110 |
},
|
| 111 |
{
|
| 112 |
"epoch": 8,
|
| 113 |
+
"loss": 99.107,
|
| 114 |
+
"lr": 0.000113,
|
| 115 |
+
"threshold": 0.846,
|
| 116 |
+
"micro_precision": 0.5385,
|
| 117 |
+
"micro_recall": 0.2068,
|
| 118 |
+
"micro_f1": 0.2988,
|
| 119 |
+
"macro_f1": 0.3142,
|
| 120 |
+
"exact_set_match": 0.0667,
|
| 121 |
+
"avg_selected": 2.02,
|
| 122 |
+
"avg_true_pass": 5.27,
|
| 123 |
+
"num_groups": 45
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"epoch": 9,
|
| 127 |
+
"loss": 97.153,
|
| 128 |
+
"lr": 0.000113,
|
| 129 |
+
"threshold": 0.814,
|
| 130 |
+
"micro_precision": 0.5584,
|
| 131 |
+
"micro_recall": 0.1814,
|
| 132 |
+
"micro_f1": 0.2739,
|
| 133 |
+
"macro_f1": 0.2812,
|
| 134 |
+
"exact_set_match": 0.0667,
|
| 135 |
+
"avg_selected": 1.71,
|
| 136 |
+
"avg_true_pass": 5.27,
|
| 137 |
+
"num_groups": 45
|
| 138 |
},
|
| 139 |
{
|
| 140 |
"epoch": 10,
|
| 141 |
+
"loss": 99.22,
|
| 142 |
+
"lr": 0.000113,
|
| 143 |
+
"threshold": 0.904,
|
| 144 |
+
"micro_precision": 0.5532,
|
| 145 |
+
"micro_recall": 0.2194,
|
| 146 |
+
"micro_f1": 0.3142,
|
| 147 |
+
"macro_f1": 0.2945,
|
| 148 |
+
"exact_set_match": 0.0667,
|
| 149 |
+
"avg_selected": 2.09,
|
| 150 |
+
"avg_true_pass": 5.27,
|
| 151 |
+
"num_groups": 45
|
| 152 |
},
|
| 153 |
{
|
| 154 |
"epoch": 11,
|
| 155 |
+
"loss": 95.665,
|
| 156 |
+
"lr": 0.000113,
|
| 157 |
+
"threshold": 0.832,
|
| 158 |
+
"micro_precision": 0.6082,
|
| 159 |
+
"micro_recall": 0.2489,
|
| 160 |
+
"micro_f1": 0.3533,
|
| 161 |
+
"macro_f1": 0.3182,
|
| 162 |
+
"exact_set_match": 0.0667,
|
| 163 |
+
"avg_selected": 2.16,
|
| 164 |
+
"avg_true_pass": 5.27,
|
| 165 |
+
"num_groups": 45
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 12,
|
| 169 |
+
"loss": 94.533,
|
| 170 |
+
"lr": 0.000113,
|
| 171 |
+
"threshold": 0.856,
|
| 172 |
+
"micro_precision": 0.5897,
|
| 173 |
+
"micro_recall": 0.1941,
|
| 174 |
+
"micro_f1": 0.2921,
|
| 175 |
+
"macro_f1": 0.2846,
|
| 176 |
+
"exact_set_match": 0.0667,
|
| 177 |
+
"avg_selected": 1.73,
|
| 178 |
+
"avg_true_pass": 5.27,
|
| 179 |
+
"num_groups": 45
|
| 180 |
},
|
| 181 |
{
|
| 182 |
"epoch": 13,
|
| 183 |
+
"loss": 94.128,
|
| 184 |
+
"lr": 0.000113,
|
| 185 |
+
"threshold": 0.834,
|
| 186 |
+
"micro_precision": 0.5444,
|
| 187 |
+
"micro_recall": 0.2068,
|
| 188 |
+
"micro_f1": 0.2997,
|
| 189 |
+
"macro_f1": 0.3167,
|
| 190 |
+
"exact_set_match": 0.0667,
|
| 191 |
+
"avg_selected": 2.0,
|
| 192 |
+
"avg_true_pass": 5.27,
|
| 193 |
+
"num_groups": 45
|
| 194 |
},
|
| 195 |
{
|
| 196 |
"epoch": 14,
|
| 197 |
+
"loss": 93.346,
|
| 198 |
+
"lr": 0.000113,
|
| 199 |
+
"threshold": 0.912,
|
| 200 |
+
"micro_precision": 0.6364,
|
| 201 |
+
"micro_recall": 0.2658,
|
| 202 |
+
"micro_f1": 0.375,
|
| 203 |
+
"macro_f1": 0.3274,
|
| 204 |
+
"exact_set_match": 0.0667,
|
| 205 |
+
"avg_selected": 2.2,
|
| 206 |
+
"avg_true_pass": 5.27,
|
| 207 |
+
"num_groups": 45
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 15,
|
| 211 |
+
"loss": 92.342,
|
| 212 |
+
"lr": 0.000113,
|
| 213 |
+
"threshold": 0.85,
|
| 214 |
+
"micro_precision": 0.6476,
|
| 215 |
+
"micro_recall": 0.2869,
|
| 216 |
+
"micro_f1": 0.3977,
|
| 217 |
+
"macro_f1": 0.3642,
|
| 218 |
+
"exact_set_match": 0.0889,
|
| 219 |
+
"avg_selected": 2.33,
|
| 220 |
+
"avg_true_pass": 5.27,
|
| 221 |
+
"num_groups": 45
|
| 222 |
},
|
| 223 |
{
|
| 224 |
"epoch": 16,
|
| 225 |
+
"loss": 91.0,
|
| 226 |
+
"lr": 0.000113,
|
| 227 |
+
"threshold": 0.854,
|
| 228 |
+
"micro_precision": 0.6349,
|
| 229 |
+
"micro_recall": 0.3376,
|
| 230 |
+
"micro_f1": 0.4408,
|
| 231 |
+
"macro_f1": 0.3765,
|
| 232 |
+
"exact_set_match": 0.0889,
|
| 233 |
+
"avg_selected": 2.8,
|
| 234 |
+
"avg_true_pass": 5.27,
|
| 235 |
+
"num_groups": 45
|
| 236 |
},
|
| 237 |
{
|
| 238 |
"epoch": 17,
|
| 239 |
+
"loss": 90.537,
|
| 240 |
+
"lr": 0.000113,
|
| 241 |
+
"threshold": 0.898,
|
| 242 |
+
"micro_precision": 0.6667,
|
| 243 |
+
"micro_recall": 0.3629,
|
| 244 |
+
"micro_f1": 0.4699,
|
| 245 |
+
"macro_f1": 0.3987,
|
| 246 |
+
"exact_set_match": 0.0889,
|
| 247 |
+
"avg_selected": 2.87,
|
| 248 |
+
"avg_true_pass": 5.27,
|
| 249 |
+
"num_groups": 45
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 18,
|
| 253 |
+
"loss": 89.641,
|
| 254 |
+
"lr": 0.000113,
|
| 255 |
+
"threshold": 0.885,
|
| 256 |
+
"micro_precision": 0.7,
|
| 257 |
+
"micro_recall": 0.3544,
|
| 258 |
+
"micro_f1": 0.4706,
|
| 259 |
+
"macro_f1": 0.3771,
|
| 260 |
+
"exact_set_match": 0.0889,
|
| 261 |
+
"avg_selected": 2.67,
|
| 262 |
+
"avg_true_pass": 5.27,
|
| 263 |
+
"num_groups": 45
|
| 264 |
},
|
| 265 |
{
|
| 266 |
"epoch": 19,
|
| 267 |
+
"loss": 90.677,
|
| 268 |
+
"lr": 0.000113,
|
| 269 |
+
"threshold": 0.86,
|
| 270 |
+
"micro_precision": 0.6562,
|
| 271 |
+
"micro_recall": 0.2658,
|
| 272 |
+
"micro_f1": 0.3784,
|
| 273 |
+
"macro_f1": 0.3735,
|
| 274 |
+
"exact_set_match": 0.0889,
|
| 275 |
+
"avg_selected": 2.13,
|
| 276 |
+
"avg_true_pass": 5.27,
|
| 277 |
+
"num_groups": 45
|
| 278 |
},
|
| 279 |
{
|
| 280 |
"epoch": 20,
|
| 281 |
+
"loss": 89.7,
|
| 282 |
+
"lr": 0.000113,
|
| 283 |
+
"threshold": 0.854,
|
| 284 |
+
"micro_precision": 0.6915,
|
| 285 |
+
"micro_recall": 0.2743,
|
| 286 |
+
"micro_f1": 0.3927,
|
| 287 |
+
"macro_f1": 0.3232,
|
| 288 |
+
"exact_set_match": 0.0667,
|
| 289 |
+
"avg_selected": 2.09,
|
| 290 |
+
"avg_true_pass": 5.27,
|
| 291 |
+
"num_groups": 45
|
| 292 |
},
|
| 293 |
{
|
| 294 |
"epoch": 21,
|
| 295 |
+
"loss": 88.301,
|
| 296 |
+
"lr": 0.000113,
|
| 297 |
+
"threshold": 0.866,
|
| 298 |
+
"micro_precision": 0.7361,
|
| 299 |
+
"micro_recall": 0.4473,
|
| 300 |
+
"micro_f1": 0.5564,
|
| 301 |
+
"macro_f1": 0.4062,
|
| 302 |
+
"exact_set_match": 0.0889,
|
| 303 |
+
"avg_selected": 3.2,
|
| 304 |
+
"avg_true_pass": 5.27,
|
| 305 |
+
"num_groups": 45
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 22,
|
| 309 |
+
"loss": 88.067,
|
| 310 |
+
"lr": 0.000113,
|
| 311 |
+
"threshold": 0.874,
|
| 312 |
+
"micro_precision": 0.7208,
|
| 313 |
+
"micro_recall": 0.4684,
|
| 314 |
+
"micro_f1": 0.5678,
|
| 315 |
+
"macro_f1": 0.4109,
|
| 316 |
+
"exact_set_match": 0.0889,
|
| 317 |
+
"avg_selected": 3.42,
|
| 318 |
+
"avg_true_pass": 5.27,
|
| 319 |
+
"num_groups": 45
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 23,
|
| 323 |
+
"loss": 86.773,
|
| 324 |
+
"lr": 0.000113,
|
| 325 |
+
"threshold": 0.872,
|
| 326 |
+
"micro_precision": 0.7561,
|
| 327 |
+
"micro_recall": 0.5232,
|
| 328 |
+
"micro_f1": 0.6185,
|
| 329 |
+
"macro_f1": 0.4446,
|
| 330 |
+
"exact_set_match": 0.0889,
|
| 331 |
+
"avg_selected": 3.64,
|
| 332 |
+
"avg_true_pass": 5.27,
|
| 333 |
+
"num_groups": 45
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 24,
|
| 337 |
+
"loss": 86.989,
|
| 338 |
+
"lr": 0.000113,
|
| 339 |
+
"threshold": 0.787,
|
| 340 |
+
"micro_precision": 0.7717,
|
| 341 |
+
"micro_recall": 0.4135,
|
| 342 |
+
"micro_f1": 0.5385,
|
| 343 |
+
"macro_f1": 0.4115,
|
| 344 |
+
"exact_set_match": 0.0889,
|
| 345 |
+
"avg_selected": 2.82,
|
| 346 |
+
"avg_true_pass": 5.27,
|
| 347 |
+
"num_groups": 45
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 25,
|
| 351 |
+
"loss": 85.558,
|
| 352 |
+
"lr": 0.000113,
|
| 353 |
+
"threshold": 0.908,
|
| 354 |
+
"micro_precision": 0.7628,
|
| 355 |
+
"micro_recall": 0.5021,
|
| 356 |
+
"micro_f1": 0.6056,
|
| 357 |
+
"macro_f1": 0.4208,
|
| 358 |
+
"exact_set_match": 0.0889,
|
| 359 |
+
"avg_selected": 3.47,
|
| 360 |
+
"avg_true_pass": 5.27,
|
| 361 |
+
"num_groups": 45
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 26,
|
| 365 |
+
"loss": 85.471,
|
| 366 |
+
"lr": 0.000113,
|
| 367 |
+
"threshold": 0.861,
|
| 368 |
+
"micro_precision": 0.8041,
|
| 369 |
+
"micro_recall": 0.5021,
|
| 370 |
+
"micro_f1": 0.6182,
|
| 371 |
+
"macro_f1": 0.4293,
|
| 372 |
+
"exact_set_match": 0.0889,
|
| 373 |
+
"avg_selected": 3.29,
|
| 374 |
+
"avg_true_pass": 5.27,
|
| 375 |
+
"num_groups": 45
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 27,
|
| 379 |
+
"loss": 84.254,
|
| 380 |
+
"lr": 0.000113,
|
| 381 |
+
"threshold": 0.833,
|
| 382 |
+
"micro_precision": 0.8095,
|
| 383 |
+
"micro_recall": 0.4304,
|
| 384 |
+
"micro_f1": 0.562,
|
| 385 |
+
"macro_f1": 0.4137,
|
| 386 |
+
"exact_set_match": 0.1111,
|
| 387 |
+
"avg_selected": 2.8,
|
| 388 |
+
"avg_true_pass": 5.27,
|
| 389 |
+
"num_groups": 45
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 28,
|
| 393 |
+
"loss": 82.08,
|
| 394 |
+
"lr": 5.65e-05,
|
| 395 |
+
"threshold": 0.875,
|
| 396 |
+
"micro_precision": 0.8497,
|
| 397 |
+
"micro_recall": 0.6203,
|
| 398 |
+
"micro_f1": 0.7171,
|
| 399 |
+
"macro_f1": 0.4292,
|
| 400 |
+
"exact_set_match": 0.1556,
|
| 401 |
+
"avg_selected": 3.84,
|
| 402 |
+
"avg_true_pass": 5.27,
|
| 403 |
+
"num_groups": 45
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 29,
|
| 407 |
+
"loss": 81.402,
|
| 408 |
+
"lr": 5.65e-05,
|
| 409 |
+
"threshold": 0.867,
|
| 410 |
+
"micro_precision": 0.8645,
|
| 411 |
+
"micro_recall": 0.5654,
|
| 412 |
+
"micro_f1": 0.6837,
|
| 413 |
+
"macro_f1": 0.4127,
|
| 414 |
+
"exact_set_match": 0.1333,
|
| 415 |
+
"avg_selected": 3.44,
|
| 416 |
+
"avg_true_pass": 5.27,
|
| 417 |
+
"num_groups": 45
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 30,
|
| 421 |
+
"loss": 80.185,
|
| 422 |
+
"lr": 5.65e-05,
|
| 423 |
+
"threshold": 0.889,
|
| 424 |
+
"micro_precision": 0.8621,
|
| 425 |
+
"micro_recall": 0.6329,
|
| 426 |
+
"micro_f1": 0.7299,
|
| 427 |
+
"macro_f1": 0.4593,
|
| 428 |
+
"exact_set_match": 0.1333,
|
| 429 |
+
"avg_selected": 3.87,
|
| 430 |
+
"avg_true_pass": 5.27,
|
| 431 |
+
"num_groups": 45
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 31,
|
| 435 |
+
"loss": 79.499,
|
| 436 |
+
"lr": 5.65e-05,
|
| 437 |
+
"threshold": 0.867,
|
| 438 |
+
"micro_precision": 0.8542,
|
| 439 |
+
"micro_recall": 0.519,
|
| 440 |
+
"micro_f1": 0.6457,
|
| 441 |
+
"macro_f1": 0.3873,
|
| 442 |
+
"exact_set_match": 0.1111,
|
| 443 |
+
"avg_selected": 3.2,
|
| 444 |
+
"avg_true_pass": 5.27,
|
| 445 |
+
"num_groups": 45
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 32,
|
| 449 |
+
"loss": 79.077,
|
| 450 |
+
"lr": 5.65e-05,
|
| 451 |
+
"threshold": 0.883,
|
| 452 |
+
"micro_precision": 0.875,
|
| 453 |
+
"micro_recall": 0.6203,
|
| 454 |
+
"micro_f1": 0.7259,
|
| 455 |
+
"macro_f1": 0.4637,
|
| 456 |
+
"exact_set_match": 0.1556,
|
| 457 |
+
"avg_selected": 3.73,
|
| 458 |
+
"avg_true_pass": 5.27,
|
| 459 |
+
"num_groups": 45
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 33,
|
| 463 |
+
"loss": 78.368,
|
| 464 |
+
"lr": 5.65e-05,
|
| 465 |
+
"threshold": 0.864,
|
| 466 |
"micro_precision": 0.8889,
|
| 467 |
+
"micro_recall": 0.6076,
|
| 468 |
+
"micro_f1": 0.7218,
|
| 469 |
+
"macro_f1": 0.477,
|
| 470 |
+
"exact_set_match": 0.1556,
|
| 471 |
+
"avg_selected": 3.6,
|
| 472 |
+
"avg_true_pass": 5.27,
|
| 473 |
+
"num_groups": 45
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 34,
|
| 477 |
+
"loss": 77.548,
|
| 478 |
+
"lr": 5.65e-05,
|
| 479 |
+
"threshold": 0.883,
|
| 480 |
+
"micro_precision": 0.8876,
|
| 481 |
+
"micro_recall": 0.6329,
|
| 482 |
+
"micro_f1": 0.7389,
|
| 483 |
+
"macro_f1": 0.509,
|
| 484 |
+
"exact_set_match": 0.1778,
|
| 485 |
+
"avg_selected": 3.76,
|
| 486 |
+
"avg_true_pass": 5.27,
|
| 487 |
+
"num_groups": 45
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 35,
|
| 491 |
+
"loss": 76.814,
|
| 492 |
+
"lr": 5.65e-05,
|
| 493 |
+
"threshold": 0.907,
|
| 494 |
+
"micro_precision": 0.8613,
|
| 495 |
+
"micro_recall": 0.6287,
|
| 496 |
+
"micro_f1": 0.7268,
|
| 497 |
+
"macro_f1": 0.4984,
|
| 498 |
+
"exact_set_match": 0.1333,
|
| 499 |
+
"avg_selected": 3.84,
|
| 500 |
+
"avg_true_pass": 5.27,
|
| 501 |
+
"num_groups": 45
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 36,
|
| 505 |
+
"loss": 76.473,
|
| 506 |
+
"lr": 5.65e-05,
|
| 507 |
+
"threshold": 0.891,
|
| 508 |
+
"micro_precision": 0.9068,
|
| 509 |
+
"micro_recall": 0.616,
|
| 510 |
+
"micro_f1": 0.7337,
|
| 511 |
+
"macro_f1": 0.4966,
|
| 512 |
+
"exact_set_match": 0.1333,
|
| 513 |
+
"avg_selected": 3.58,
|
| 514 |
+
"avg_true_pass": 5.27,
|
| 515 |
+
"num_groups": 45
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 37,
|
| 519 |
+
"loss": 75.839,
|
| 520 |
+
"lr": 5.65e-05,
|
| 521 |
+
"threshold": 0.87,
|
| 522 |
+
"micro_precision": 0.9387,
|
| 523 |
+
"micro_recall": 0.6456,
|
| 524 |
+
"micro_f1": 0.765,
|
| 525 |
+
"macro_f1": 0.5103,
|
| 526 |
+
"exact_set_match": 0.2,
|
| 527 |
+
"avg_selected": 3.62,
|
| 528 |
+
"avg_true_pass": 5.27,
|
| 529 |
+
"num_groups": 45
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 38,
|
| 533 |
+
"loss": 74.741,
|
| 534 |
+
"lr": 5.65e-05,
|
| 535 |
+
"threshold": 0.883,
|
| 536 |
+
"micro_precision": 0.9221,
|
| 537 |
+
"micro_recall": 0.5992,
|
| 538 |
+
"micro_f1": 0.7263,
|
| 539 |
+
"macro_f1": 0.5004,
|
| 540 |
+
"exact_set_match": 0.1778,
|
| 541 |
+
"avg_selected": 3.42,
|
| 542 |
+
"avg_true_pass": 5.27,
|
| 543 |
+
"num_groups": 45
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 39,
|
| 547 |
+
"loss": 73.661,
|
| 548 |
+
"lr": 5.65e-05,
|
| 549 |
+
"threshold": 0.902,
|
| 550 |
+
"micro_precision": 0.8977,
|
| 551 |
+
"micro_recall": 0.6667,
|
| 552 |
+
"micro_f1": 0.7651,
|
| 553 |
+
"macro_f1": 0.5197,
|
| 554 |
+
"exact_set_match": 0.2,
|
| 555 |
+
"avg_selected": 3.91,
|
| 556 |
+
"avg_true_pass": 5.27,
|
| 557 |
+
"num_groups": 45
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 40,
|
| 561 |
+
"loss": 73.047,
|
| 562 |
+
"lr": 5.65e-05,
|
| 563 |
+
"threshold": 0.873,
|
| 564 |
+
"micro_precision": 0.9595,
|
| 565 |
+
"micro_recall": 0.5992,
|
| 566 |
+
"micro_f1": 0.7377,
|
| 567 |
+
"macro_f1": 0.4757,
|
| 568 |
+
"exact_set_match": 0.1778,
|
| 569 |
+
"avg_selected": 3.29,
|
| 570 |
+
"avg_true_pass": 5.27,
|
| 571 |
+
"num_groups": 45
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 41,
|
| 575 |
+
"loss": 72.159,
|
| 576 |
+
"lr": 5.65e-05,
|
| 577 |
+
"threshold": 0.886,
|
| 578 |
+
"micro_precision": 0.9096,
|
| 579 |
+
"micro_recall": 0.6371,
|
| 580 |
+
"micro_f1": 0.7494,
|
| 581 |
+
"macro_f1": 0.4815,
|
| 582 |
+
"exact_set_match": 0.1333,
|
| 583 |
+
"avg_selected": 3.69,
|
| 584 |
+
"avg_true_pass": 5.27,
|
| 585 |
+
"num_groups": 45
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 42,
|
| 589 |
+
"loss": 71.134,
|
| 590 |
+
"lr": 5.65e-05,
|
| 591 |
+
"threshold": 0.908,
|
| 592 |
+
"micro_precision": 0.9367,
|
| 593 |
+
"micro_recall": 0.6245,
|
| 594 |
+
"micro_f1": 0.7494,
|
| 595 |
+
"macro_f1": 0.5127,
|
| 596 |
+
"exact_set_match": 0.1556,
|
| 597 |
+
"avg_selected": 3.51,
|
| 598 |
+
"avg_true_pass": 5.27,
|
| 599 |
+
"num_groups": 45
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 43,
|
| 603 |
+
"loss": 70.889,
|
| 604 |
+
"lr": 5.65e-05,
|
| 605 |
+
"threshold": 0.902,
|
| 606 |
+
"micro_precision": 0.9268,
|
| 607 |
+
"micro_recall": 0.6414,
|
| 608 |
+
"micro_f1": 0.7581,
|
| 609 |
+
"macro_f1": 0.5085,
|
| 610 |
+
"exact_set_match": 0.1778,
|
| 611 |
+
"avg_selected": 3.64,
|
| 612 |
+
"avg_true_pass": 5.27,
|
| 613 |
+
"num_groups": 45
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 44,
|
| 617 |
+
"loss": 67.678,
|
| 618 |
+
"lr": 2.825e-05,
|
| 619 |
+
"threshold": 0.887,
|
| 620 |
+
"micro_precision": 0.9441,
|
| 621 |
+
"micro_recall": 0.6414,
|
| 622 |
+
"micro_f1": 0.7638,
|
| 623 |
+
"macro_f1": 0.5242,
|
| 624 |
+
"exact_set_match": 0.1556,
|
| 625 |
+
"avg_selected": 3.58,
|
| 626 |
+
"avg_true_pass": 5.27,
|
| 627 |
+
"num_groups": 45
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 45,
|
| 631 |
+
"loss": 66.422,
|
| 632 |
+
"lr": 2.825e-05,
|
| 633 |
+
"threshold": 0.902,
|
| 634 |
+
"micro_precision": 0.9325,
|
| 635 |
+
"micro_recall": 0.6414,
|
| 636 |
+
"micro_f1": 0.76,
|
| 637 |
+
"macro_f1": 0.5053,
|
| 638 |
+
"exact_set_match": 0.1333,
|
| 639 |
+
"avg_selected": 3.62,
|
| 640 |
+
"avg_true_pass": 5.27,
|
| 641 |
+
"num_groups": 45
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 46,
|
| 645 |
+
"loss": 65.293,
|
| 646 |
+
"lr": 2.825e-05,
|
| 647 |
+
"threshold": 0.898,
|
| 648 |
+
"micro_precision": 0.9286,
|
| 649 |
+
"micro_recall": 0.6582,
|
| 650 |
+
"micro_f1": 0.7704,
|
| 651 |
+
"macro_f1": 0.5371,
|
| 652 |
+
"exact_set_match": 0.2,
|
| 653 |
+
"avg_selected": 3.73,
|
| 654 |
+
"avg_true_pass": 5.27,
|
| 655 |
+
"num_groups": 45
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 47,
|
| 659 |
+
"loss": 63.112,
|
| 660 |
+
"lr": 2.825e-05,
|
| 661 |
+
"threshold": 0.929,
|
| 662 |
+
"micro_precision": 0.9186,
|
| 663 |
+
"micro_recall": 0.6667,
|
| 664 |
+
"micro_f1": 0.7726,
|
| 665 |
+
"macro_f1": 0.5465,
|
| 666 |
+
"exact_set_match": 0.1556,
|
| 667 |
+
"avg_selected": 3.82,
|
| 668 |
+
"avg_true_pass": 5.27,
|
| 669 |
+
"num_groups": 45
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 48,
|
| 673 |
+
"loss": 62.482,
|
| 674 |
+
"lr": 2.825e-05,
|
| 675 |
+
"threshold": 0.916,
|
| 676 |
+
"micro_precision": 0.9259,
|
| 677 |
+
"micro_recall": 0.6329,
|
| 678 |
+
"micro_f1": 0.7519,
|
| 679 |
+
"macro_f1": 0.4931,
|
| 680 |
+
"exact_set_match": 0.1333,
|
| 681 |
+
"avg_selected": 3.6,
|
| 682 |
+
"avg_true_pass": 5.27,
|
| 683 |
+
"num_groups": 45
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 49,
|
| 687 |
+
"loss": 61.881,
|
| 688 |
+
"lr": 2.825e-05,
|
| 689 |
+
"threshold": 0.911,
|
| 690 |
+
"micro_precision": 0.9048,
|
| 691 |
+
"micro_recall": 0.6414,
|
| 692 |
+
"micro_f1": 0.7506,
|
| 693 |
+
"macro_f1": 0.5122,
|
| 694 |
+
"exact_set_match": 0.1333,
|
| 695 |
+
"avg_selected": 3.73,
|
| 696 |
+
"avg_true_pass": 5.27,
|
| 697 |
+
"num_groups": 45
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 50,
|
| 701 |
+
"loss": 60.049,
|
| 702 |
+
"lr": 2.825e-05,
|
| 703 |
+
"threshold": 0.927,
|
| 704 |
+
"micro_precision": 0.925,
|
| 705 |
+
"micro_recall": 0.6245,
|
| 706 |
+
"micro_f1": 0.7456,
|
| 707 |
+
"macro_f1": 0.4902,
|
| 708 |
+
"exact_set_match": 0.1111,
|
| 709 |
+
"avg_selected": 3.56,
|
| 710 |
+
"avg_true_pass": 5.27,
|
| 711 |
+
"num_groups": 45
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 51,
|
| 715 |
+
"loss": 57.691,
|
| 716 |
+
"lr": 2.825e-05,
|
| 717 |
+
"threshold": 0.916,
|
| 718 |
+
"micro_precision": 0.9157,
|
| 719 |
+
"micro_recall": 0.6414,
|
| 720 |
+
"micro_f1": 0.7543,
|
| 721 |
+
"macro_f1": 0.4966,
|
| 722 |
+
"exact_set_match": 0.1111,
|
| 723 |
+
"avg_selected": 3.69,
|
| 724 |
+
"avg_true_pass": 5.27,
|
| 725 |
+
"num_groups": 45
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 52,
|
| 729 |
+
"loss": 54.829,
|
| 730 |
+
"lr": 1.4125e-05,
|
| 731 |
+
"threshold": 0.928,
|
| 732 |
+
"micro_precision": 0.9441,
|
| 733 |
+
"micro_recall": 0.6414,
|
| 734 |
+
"micro_f1": 0.7638,
|
| 735 |
+
"macro_f1": 0.4979,
|
| 736 |
+
"exact_set_match": 0.1778,
|
| 737 |
+
"avg_selected": 3.58,
|
| 738 |
+
"avg_true_pass": 5.27,
|
| 739 |
+
"num_groups": 45
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 53,
|
| 743 |
+
"loss": 53.206,
|
| 744 |
+
"lr": 1.4125e-05,
|
| 745 |
+
"threshold": 0.942,
|
| 746 |
+
"micro_precision": 0.9345,
|
| 747 |
+
"micro_recall": 0.6624,
|
| 748 |
+
"micro_f1": 0.7753,
|
| 749 |
+
"macro_f1": 0.502,
|
| 750 |
+
"exact_set_match": 0.2,
|
| 751 |
+
"avg_selected": 3.73,
|
| 752 |
+
"avg_true_pass": 5.27,
|
| 753 |
+
"num_groups": 45
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 54,
|
| 757 |
+
"loss": 52.181,
|
| 758 |
+
"lr": 1.4125e-05,
|
| 759 |
+
"threshold": 0.935,
|
| 760 |
+
"micro_precision": 0.939,
|
| 761 |
+
"micro_recall": 0.6498,
|
| 762 |
+
"micro_f1": 0.7681,
|
| 763 |
+
"macro_f1": 0.5099,
|
| 764 |
+
"exact_set_match": 0.1778,
|
| 765 |
+
"avg_selected": 3.64,
|
| 766 |
+
"avg_true_pass": 5.27,
|
| 767 |
+
"num_groups": 45
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 55,
|
| 771 |
+
"loss": 51.222,
|
| 772 |
+
"lr": 1.4125e-05,
|
| 773 |
+
"threshold": 0.938,
|
| 774 |
+
"micro_precision": 0.939,
|
| 775 |
+
"micro_recall": 0.6498,
|
| 776 |
+
"micro_f1": 0.7681,
|
| 777 |
+
"macro_f1": 0.5015,
|
| 778 |
+
"exact_set_match": 0.1333,
|
| 779 |
+
"avg_selected": 3.64,
|
| 780 |
+
"avg_true_pass": 5.27,
|
| 781 |
+
"num_groups": 45
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 56,
|
| 785 |
+
"loss": 50.869,
|
| 786 |
+
"lr": 1.4125e-05,
|
| 787 |
+
"threshold": 0.934,
|
| 788 |
+
"micro_precision": 0.9341,
|
| 789 |
+
"micro_recall": 0.6582,
|
| 790 |
+
"micro_f1": 0.7723,
|
| 791 |
+
"macro_f1": 0.5126,
|
| 792 |
+
"exact_set_match": 0.1778,
|
| 793 |
+
"avg_selected": 3.71,
|
| 794 |
+
"avg_true_pass": 5.27,
|
| 795 |
+
"num_groups": 45
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 57,
|
| 799 |
+
"loss": 50.799,
|
| 800 |
+
"lr": 1.4125e-05,
|
| 801 |
+
"threshold": 0.935,
|
| 802 |
+
"micro_precision": 0.9277,
|
| 803 |
+
"micro_recall": 0.6498,
|
| 804 |
+
"micro_f1": 0.7643,
|
| 805 |
+
"macro_f1": 0.4713,
|
| 806 |
+
"exact_set_match": 0.1556,
|
| 807 |
+
"avg_selected": 3.69,
|
| 808 |
+
"avg_true_pass": 5.27,
|
| 809 |
+
"num_groups": 45
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 58,
|
| 813 |
+
"loss": 50.473,
|
| 814 |
+
"lr": 1e-05,
|
| 815 |
+
"threshold": 0.935,
|
| 816 |
+
"micro_precision": 0.9268,
|
| 817 |
+
"micro_recall": 0.6414,
|
| 818 |
+
"micro_f1": 0.7581,
|
| 819 |
+
"macro_f1": 0.4834,
|
| 820 |
+
"exact_set_match": 0.1333,
|
| 821 |
+
"avg_selected": 3.64,
|
| 822 |
+
"avg_true_pass": 5.27,
|
| 823 |
+
"num_groups": 45
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 59,
|
| 827 |
+
"loss": 50.023,
|
| 828 |
+
"lr": 1e-05,
|
| 829 |
+
"threshold": 0.936,
|
| 830 |
+
"micro_precision": 0.9506,
|
| 831 |
+
"micro_recall": 0.6498,
|
| 832 |
+
"micro_f1": 0.7719,
|
| 833 |
+
"macro_f1": 0.4921,
|
| 834 |
+
"exact_set_match": 0.1556,
|
| 835 |
+
"avg_selected": 3.6,
|
| 836 |
+
"avg_true_pass": 5.27,
|
| 837 |
+
"num_groups": 45
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 60,
|
| 841 |
+
"loss": 50.07,
|
| 842 |
+
"lr": 1e-05,
|
| 843 |
+
"threshold": 0.934,
|
| 844 |
+
"micro_precision": 0.9441,
|
| 845 |
+
"micro_recall": 0.6414,
|
| 846 |
+
"micro_f1": 0.7638,
|
| 847 |
+
"macro_f1": 0.4922,
|
| 848 |
+
"exact_set_match": 0.1333,
|
| 849 |
+
"avg_selected": 3.58,
|
| 850 |
+
"avg_true_pass": 5.27,
|
| 851 |
+
"num_groups": 45
|
| 852 |
}
|
| 853 |
]
|
| 854 |
}
|
model_meta.json
CHANGED
|
@@ -9,58 +9,58 @@
|
|
| 9 |
},
|
| 10 |
"encoder": "openai/text-embedding-ada-002",
|
| 11 |
"emb_dim": 1539,
|
| 12 |
-
"threshold": 0.
|
| 13 |
-
"lr": 0.
|
| 14 |
"metrics": {
|
| 15 |
-
"threshold": 0.
|
| 16 |
-
"micro_precision": 0.
|
| 17 |
-
"micro_recall": 0.
|
| 18 |
-
"micro_f1": 0.
|
| 19 |
-
"macro_f1": 0.
|
| 20 |
-
"exact_set_match": 0.
|
| 21 |
-
"avg_selected":
|
| 22 |
-
"avg_true_pass": 5.
|
| 23 |
-
"num_groups":
|
| 24 |
-
"best_epoch":
|
| 25 |
},
|
| 26 |
"bench": {
|
| 27 |
"gold_easy": {
|
| 28 |
-
"groups":
|
| 29 |
-
"precision":
|
| 30 |
-
"recall": 0.
|
| 31 |
-
"f1": 0.
|
| 32 |
},
|
| 33 |
"gold_medium": {
|
| 34 |
"groups": 2,
|
| 35 |
-
"precision":
|
| 36 |
-
"recall": 0.
|
| 37 |
-
"f1": 0.
|
| 38 |
},
|
| 39 |
"gold_hard": {
|
| 40 |
"groups": 5,
|
| 41 |
-
"precision":
|
| 42 |
"recall": 0.9667,
|
| 43 |
-
"f1": 0.
|
| 44 |
},
|
| 45 |
"llm_held_out": {
|
| 46 |
"groups": 29,
|
| 47 |
-
"precision": 0.
|
| 48 |
-
"recall": 0.
|
| 49 |
-
"f1": 0.
|
| 50 |
}
|
| 51 |
},
|
| 52 |
"data": {
|
| 53 |
-
"train_groups":
|
| 54 |
-
"gold_train_groups":
|
| 55 |
-
"test_groups":
|
| 56 |
-
"gold_test_groups":
|
| 57 |
-
"train_candidates":
|
| 58 |
-
"train_pass":
|
| 59 |
-
"train_fail":
|
| 60 |
-
"pass_ratio": 0.
|
| 61 |
-
"avg_candidates_per_train_group": 23.
|
| 62 |
"avg_pass_per_train_group": 4.7
|
| 63 |
},
|
| 64 |
-
"trained_at": "2026-07-
|
| 65 |
"source": "https://github.com/s3777091/recursive_models"
|
| 66 |
}
|
|
|
|
| 9 |
},
|
| 10 |
"encoder": "openai/text-embedding-ada-002",
|
| 11 |
"emb_dim": 1539,
|
| 12 |
+
"threshold": 0.942,
|
| 13 |
+
"lr": 0.000113,
|
| 14 |
"metrics": {
|
| 15 |
+
"threshold": 0.942,
|
| 16 |
+
"micro_precision": 0.9345,
|
| 17 |
+
"micro_recall": 0.6624,
|
| 18 |
+
"micro_f1": 0.7753,
|
| 19 |
+
"macro_f1": 0.502,
|
| 20 |
+
"exact_set_match": 0.2,
|
| 21 |
+
"avg_selected": 3.73,
|
| 22 |
+
"avg_true_pass": 5.27,
|
| 23 |
+
"num_groups": 45,
|
| 24 |
+
"best_epoch": 53
|
| 25 |
},
|
| 26 |
"bench": {
|
| 27 |
"gold_easy": {
|
| 28 |
+
"groups": 9,
|
| 29 |
+
"precision": 0.8571,
|
| 30 |
+
"recall": 0.2069,
|
| 31 |
+
"f1": 0.3333
|
| 32 |
},
|
| 33 |
"gold_medium": {
|
| 34 |
"groups": 2,
|
| 35 |
+
"precision": 1.0,
|
| 36 |
+
"recall": 0.9211,
|
| 37 |
+
"f1": 0.9589
|
| 38 |
},
|
| 39 |
"gold_hard": {
|
| 40 |
"groups": 5,
|
| 41 |
+
"precision": 1.0,
|
| 42 |
"recall": 0.9667,
|
| 43 |
+
"f1": 0.9831
|
| 44 |
},
|
| 45 |
"llm_held_out": {
|
| 46 |
"groups": 29,
|
| 47 |
+
"precision": 0.7436,
|
| 48 |
+
"recall": 0.3625,
|
| 49 |
+
"f1": 0.4874
|
| 50 |
}
|
| 51 |
},
|
| 52 |
"data": {
|
| 53 |
+
"train_groups": 1184,
|
| 54 |
+
"gold_train_groups": 48,
|
| 55 |
+
"test_groups": 45,
|
| 56 |
+
"gold_test_groups": 16,
|
| 57 |
+
"train_candidates": 27544,
|
| 58 |
+
"train_pass": 5567,
|
| 59 |
+
"train_fail": 21977,
|
| 60 |
+
"pass_ratio": 0.202,
|
| 61 |
+
"avg_candidates_per_train_group": 23.3,
|
| 62 |
"avg_pass_per_train_group": 4.7
|
| 63 |
},
|
| 64 |
+
"trained_at": "2026-07-07T02:33:14",
|
| 65 |
"source": "https://github.com/s3777091/recursive_models"
|
| 66 |
}
|
training_report.png
CHANGED
|
Git LFS Details
|
|
Git LFS Details
|