Skip to content

Instantly share code, notes, and snippets.

@thistleknot

thistleknot/cross_encoder.py

Created November 30, 2025 02:00
Show Gist options

  • Select an option

    Learn more about clone URLs
  • Save thistleknot/f9697dd6943e77bb1af8ba4ab37cdfc2 to your computer and use it in GitHub Desktop.

Select an option

Learn more about clone URLs
Save thistleknot/f9697dd6943e77bb1af8ba4ab37cdfc2 to your computer and use it in GitHub Desktop.
Download ZIP
cross encoder
Raw
cross_encoder.py
from sentence_transformers import CrossEncoder
MODEL_NAME = 'cross-encoder/ms-marco-MiniLM-L12-v2'
nli_model = CrossEncoder(MODEL_NAME)
m2v_model = distill(model_name=MODEL_NAME)
query = "Explain quantum computing"
# Generate SAPO response group
responses = [
"Quantum computing uses qubits that can be 0 and 1 simultaneously",
"It's when computers get really fast",
"I don't know what that is"
]
# NLI model EXPECTS: [premise, hypothesis] PAIRS
scores = nli_model.predict([
(query, r) for r in responses # ← Treating query as premise!
])
print(scores)
# [ 4.2502446 -10.693266 -10.976209 ]
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment