publish grounding leaderboard
Browse files- index.html +0 -1
- leaderboard.json +40 -38
- results/openai-decisions-gpt-6-luna-en.json +26 -0
- results/openai-decisions-gpt-6-luna-multi.json +26 -0
index.html
CHANGED
|
@@ -112,7 +112,6 @@ python benchmark/score.py --model <your-model> --config en --name "Your Mo
|
|
| 112 |
<strong>Models benchmarked:</strong>
|
| 113 |
<a href="https://huggingface.co/nutrientdocs/grounding-en">nutrientdocs/grounding-en</a> ·
|
| 114 |
<a href="https://huggingface.co/nutrientdocs/grounding-multilingual">nutrientdocs/grounding-multilingual</a> ·
|
| 115 |
-
<a href="https://huggingface.co/nutrientdocs/nutrient-document-decision">nutrientdocs/nutrient-document-decision</a> ·
|
| 116 |
<a href="https://huggingface.co/cross-encoder/nli-deberta-v3-base">cross-encoder/nli-deberta-v3-base</a> ·
|
| 117 |
<a href="https://huggingface.co/cross-encoder/nli-MiniLM2-L6-H768">cross-encoder/nli-MiniLM2-L6-H768</a> ·
|
| 118 |
<a href="https://huggingface.co/facebook/bart-large-mnli">facebook/bart-large-mnli</a> ·
|
|
|
|
| 112 |
<strong>Models benchmarked:</strong>
|
| 113 |
<a href="https://huggingface.co/nutrientdocs/grounding-en">nutrientdocs/grounding-en</a> ·
|
| 114 |
<a href="https://huggingface.co/nutrientdocs/grounding-multilingual">nutrientdocs/grounding-multilingual</a> ·
|
|
|
|
| 115 |
<a href="https://huggingface.co/cross-encoder/nli-deberta-v3-base">cross-encoder/nli-deberta-v3-base</a> ·
|
| 116 |
<a href="https://huggingface.co/cross-encoder/nli-MiniLM2-L6-H768">cross-encoder/nli-MiniLM2-L6-H768</a> ·
|
| 117 |
<a href="https://huggingface.co/facebook/bart-large-mnli">facebook/bart-large-mnli</a> ·
|
leaderboard.json
CHANGED
|
@@ -267,6 +267,46 @@
|
|
| 267 |
"ctx": 512,
|
| 268 |
"ours": false
|
| 269 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 270 |
{
|
| 271 |
"name": "RoBERTa-large MNLI",
|
| 272 |
"model": "FacebookAI/roberta-large-mnli",
|
|
@@ -323,43 +363,5 @@
|
|
| 323 |
"hardware": "NVIDIA RTX A6000",
|
| 324 |
"ctx": 512,
|
| 325 |
"ours": false
|
| 326 |
-
},
|
| 327 |
-
{
|
| 328 |
-
"name": "nutrient-document-decision",
|
| 329 |
-
"model": "nutrientdocs/nutrient-document-decision",
|
| 330 |
-
"config": "en",
|
| 331 |
-
"open": false,
|
| 332 |
-
"n": 5000,
|
| 333 |
-
"overall_auc": 0.9231032623776494,
|
| 334 |
-
"auc": {
|
| 335 |
-
"number": 0.9413457777777777,
|
| 336 |
-
"date": 0.999257438256781,
|
| 337 |
-
"string": 0.9968885281385281,
|
| 338 |
-
"table": 0.9200564106251071,
|
| 339 |
-
"prose": 0.9593182032206422
|
| 340 |
-
},
|
| 341 |
-
"pairs_per_s": 8.2,
|
| 342 |
-
"hardware": "NVIDIA RTX A6000",
|
| 343 |
-
"ctx": 8192,
|
| 344 |
-
"ours": true
|
| 345 |
-
},
|
| 346 |
-
{
|
| 347 |
-
"name": "nutrient-document-decision",
|
| 348 |
-
"model": "nutrientdocs/nutrient-document-decision",
|
| 349 |
-
"config": "multi",
|
| 350 |
-
"open": false,
|
| 351 |
-
"n": 8421,
|
| 352 |
-
"overall_auc": 0.948770067424967,
|
| 353 |
-
"auc": {
|
| 354 |
-
"number": 0.9969035555555555,
|
| 355 |
-
"date": null,
|
| 356 |
-
"string": null,
|
| 357 |
-
"table": 0.9954755455811793,
|
| 358 |
-
"prose": 0.8926265024038461
|
| 359 |
-
},
|
| 360 |
-
"pairs_per_s": 7.8,
|
| 361 |
-
"hardware": "NVIDIA RTX A6000",
|
| 362 |
-
"ctx": 8192,
|
| 363 |
-
"ours": true
|
| 364 |
}
|
| 365 |
]
|
|
|
|
| 267 |
"ctx": 512,
|
| 268 |
"ours": false
|
| 269 |
},
|
| 270 |
+
{
|
| 271 |
+
"name": "OpenAI Decisions (gpt-6-luna)",
|
| 272 |
+
"model": "gpt-6-luna",
|
| 273 |
+
"config": "en",
|
| 274 |
+
"open": false,
|
| 275 |
+
"n": 5000,
|
| 276 |
+
"overall_auc": 0.8228269611400488,
|
| 277 |
+
"auc": {
|
| 278 |
+
"number": 0.7689031111111111,
|
| 279 |
+
"date": 0.9890859222932834,
|
| 280 |
+
"string": 0.9343434343434344,
|
| 281 |
+
"table": 0.8094126105399382,
|
| 282 |
+
"prose": 0.9177731738707349
|
| 283 |
+
},
|
| 284 |
+
"pairs_per_s": 49.8,
|
| 285 |
+
"hardware": "OpenAI Decisions API (gpt-6-luna)",
|
| 286 |
+
"ctx": "api",
|
| 287 |
+
"cloud": true,
|
| 288 |
+
"ours": false
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"name": "OpenAI Decisions (gpt-6-luna)",
|
| 292 |
+
"model": "gpt-6-luna",
|
| 293 |
+
"config": "multi",
|
| 294 |
+
"open": false,
|
| 295 |
+
"n": 8421,
|
| 296 |
+
"overall_auc": 0.861610311695087,
|
| 297 |
+
"auc": {
|
| 298 |
+
"number": 0.9336608888888889,
|
| 299 |
+
"date": null,
|
| 300 |
+
"string": null,
|
| 301 |
+
"table": 0.9352640845070422,
|
| 302 |
+
"prose": 0.8087951472355769
|
| 303 |
+
},
|
| 304 |
+
"pairs_per_s": 51.7,
|
| 305 |
+
"hardware": "OpenAI Decisions API (gpt-6-luna)",
|
| 306 |
+
"ctx": "api",
|
| 307 |
+
"cloud": true,
|
| 308 |
+
"ours": false
|
| 309 |
+
},
|
| 310 |
{
|
| 311 |
"name": "RoBERTa-large MNLI",
|
| 312 |
"model": "FacebookAI/roberta-large-mnli",
|
|
|
|
| 363 |
"hardware": "NVIDIA RTX A6000",
|
| 364 |
"ctx": 512,
|
| 365 |
"ours": false
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 366 |
}
|
| 367 |
]
|
results/openai-decisions-gpt-6-luna-en.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"name": "OpenAI Decisions (gpt-6-luna)",
|
| 3 |
+
"model": "gpt-6-luna",
|
| 4 |
+
"config": "en",
|
| 5 |
+
"open": false,
|
| 6 |
+
"n": 5000,
|
| 7 |
+
"overall_auc": 0.8228269611400488,
|
| 8 |
+
"auc": {
|
| 9 |
+
"number": 0.7689031111111111,
|
| 10 |
+
"date": 0.9890859222932834,
|
| 11 |
+
"string": 0.9343434343434344,
|
| 12 |
+
"table": 0.8094126105399382,
|
| 13 |
+
"prose": 0.9177731738707349
|
| 14 |
+
},
|
| 15 |
+
"pairs_per_s": 49.8,
|
| 16 |
+
"hardware": "OpenAI Decisions API (gpt-6-luna)",
|
| 17 |
+
"ctx": "api",
|
| 18 |
+
"perf": {
|
| 19 |
+
"accelerator": "OpenAI Decisions API (gpt-6-luna)",
|
| 20 |
+
"n_pairs": 5000,
|
| 21 |
+
"elapsed_s": 100.36,
|
| 22 |
+
"ctx": "api",
|
| 23 |
+
"pairs_per_s": 49.8
|
| 24 |
+
},
|
| 25 |
+
"cloud": true
|
| 26 |
+
}
|
results/openai-decisions-gpt-6-luna-multi.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"name": "OpenAI Decisions (gpt-6-luna)",
|
| 3 |
+
"model": "gpt-6-luna",
|
| 4 |
+
"config": "multi",
|
| 5 |
+
"open": false,
|
| 6 |
+
"n": 8421,
|
| 7 |
+
"overall_auc": 0.861610311695087,
|
| 8 |
+
"auc": {
|
| 9 |
+
"number": 0.9336608888888889,
|
| 10 |
+
"date": null,
|
| 11 |
+
"string": null,
|
| 12 |
+
"table": 0.9352640845070422,
|
| 13 |
+
"prose": 0.8087951472355769
|
| 14 |
+
},
|
| 15 |
+
"pairs_per_s": 51.7,
|
| 16 |
+
"hardware": "OpenAI Decisions API (gpt-6-luna)",
|
| 17 |
+
"ctx": "api",
|
| 18 |
+
"perf": {
|
| 19 |
+
"accelerator": "OpenAI Decisions API (gpt-6-luna)",
|
| 20 |
+
"n_pairs": 8421,
|
| 21 |
+
"elapsed_s": 162.83,
|
| 22 |
+
"ctx": "api",
|
| 23 |
+
"pairs_per_s": 51.7
|
| 24 |
+
},
|
| 25 |
+
"cloud": true
|
| 26 |
+
}
|