Update README.md
Browse files
README.md
CHANGED
@@ -508,3 +508,36 @@ configs:
|
|
508 |
- split: queries
|
509 |
path: ruby-queries/queries-*
|
510 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
508 |
- split: queries
|
509 |
path: ruby-queries/queries-*
|
510 |
---
|
511 |
+
Employing the MTEB evaluation framework's dataset version, utilize the code below for assessment:
|
512 |
+
|
513 |
+
```python
|
514 |
+
import mteb
|
515 |
+
import logging
|
516 |
+
from sentence_transformers import SentenceTransformer
|
517 |
+
from mteb import MTEB
|
518 |
+
|
519 |
+
logger = logging.getLogger(__name__)
|
520 |
+
|
521 |
+
model_name = 'intfloat/e5-base-v2'
|
522 |
+
model = SentenceTransformer(model_name)
|
523 |
+
tasks = mteb.get_tasks(
|
524 |
+
tasks=[
|
525 |
+
"AppsRetrieval",
|
526 |
+
"CodeFeedbackMT",
|
527 |
+
"CodeFeedbackST",
|
528 |
+
"CodeTransOceanContest",
|
529 |
+
"CodeTransOceanDL",
|
530 |
+
"CosQA",
|
531 |
+
"SyntheticText2SQL",
|
532 |
+
"StackOverflowQA",
|
533 |
+
"COIRCodeSearchNetRetrieval",
|
534 |
+
"CodeSearchNetCCRetrieval",
|
535 |
+
]
|
536 |
+
)
|
537 |
+
evaluation = MTEB(tasks=tasks)
|
538 |
+
results = evaluation.run(
|
539 |
+
model=model,
|
540 |
+
overwrite_results=True
|
541 |
+
)
|
542 |
+
print(result)
|
543 |
+
```
|