Skip to content

random results with query tokens length > 45 #1

@agoriwmt

Description

@agoriwmt

Hi,
I'm encountering an issue where maxsim_cpu.maxsim_scores_variable returns correct results for query tokens length up to 45, but produces incorrect random scores once the length exceeds 45. Below is a minimal reproducible example demonstrating the behavior:

import numpy as np
import maxsim_cpu

query = np.random.randn(32, 128).astype(np.float32)
query /= np.linalg.norm(query, axis=1, keepdims=True)
scores = maxsim_cpu.maxsim_scores_variable(query, [query])
# returns 32

query = np.random.randn(45, 128).astype(np.float32)
query /= np.linalg.norm(query, axis=1, keepdims=True)
scores = maxsim_cpu.maxsim_scores_variable(query, [query])
# returns 45


query = np.random.randn(46, 128).astype(np.float32)
query /= np.linalg.norm(query, axis=1, keepdims=True)
scores = maxsim_cpu.maxsim_scores_variable(query, [query])
# returns [8.1203795e+08]

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions