Downloads · 30 days
15
20% of all-time downloads
bsca-bge-micro-v2-contrastive-v15-gold-v3-multidomain-384
Labradorlabs/bsca-bge-micro-v2-contrastive-v15-gold-v3-multidomain-384
bsca-bge-micro-v2-contrastive-v15-gold-v3-multidomain-384 is a feature extraction model from Labradorlabs. Use it when you need embeddings to search or compare text. The card lists the license as mit.
Embedding model for stripped pseudo-C → source retrieval.
All-time downloads
76
Public
Parameters
17.4M
69.6 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors69.6 MB · 99%
At a glance
- Task
- Feature Extraction
- License
- mit
- Model type
- bert
- Access
- Public
- Created
- Aug 11, 2026
- Updated
- Aug 12, 2026
- SHA
- dd05a368
Base models
Model card
From the Hugging Face model README
BSCA BGE-Micro-v2 Gold v3 Multidomain
Embedding model for stripped pseudo-C → source retrieval.
- Model ID:
Labradorlabs/bsca-bge-micro-v2-contrastive-v15-gold-v3-multidomain-384 - Backbone:
TaylorAI/bge-micro-v2 - Initialization / training mode: weights-restart from
artifacts/reference_v2/models/gold-scale-v2-seed44; optimizer, scheduler, and RNG state were reinitialized (fresh-or-weights-restart) - Dimension / maximum tokens:
384 / 384 - Development-only selected seed:
44(R@1 0.3186, MRR 0.3553) - Frozen test result: R@1 0.3294, R@10 0.6488, MRR 0.4431
- Training data:
Labradorlabs/bsca-binary-source-gold-v3-multidomain
Embed query pseudo-C and source code with attention-mask mean pooling followed by L2 normalization. Use the exact model revision recorded by the associated ReferenceDB vector manifest; do not combine it with vectors from another model revision.
Scope is limited to the accompanying Gold distribution: ELF/Linux and PE x86_64 Windows-GNU direct-TU views. The model is not promoted by an external blind SCA gate yet.
- Task
- Feature Extraction
- Type
- bert
- License
- mit
- Languages
- code
- Created
- Aug 11, 2026
- Updated
- Aug 12, 2026