ncbi / feature-extraction updated 2 years ago

MedCPT-Query-Encoder

Name: MedCPT-Query-Encoder
Author: ncbi

MedCPT generates embeddings of biomedical texts that can be used for semantic search (dense retrieval). The model contains two encoders: - MedCPT Query Encoder: compute the embeddings of short texts (e.g., questions, search queries, sentences). - MedCPT Article Encoder: compute the embeddings of articles (e.g., PubMed titles & abstracts).

Params

110 M

Context

512

Downloads 30d

454 K

Likes

Licence: other Not gated SAFETENSORS View on Hugging Face ↗

Download history

daily snapshots · 11 days

526 K454 K

Jul 28Jul 31Aug 4Aug 7

Can you run it?

Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.

File	Quant	Size	Est. VRAM	Verdict on RTX 4090 · 24 GB
model.safetensors	f32	0.4 GB	1.0 GB	✅ Runs comfortably

Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.

Specifications

Architecture: BertModel
Parameters: 110 M
Tensor type: F32
Context length: 512
Vocabulary: 30,522
Layers / heads: 12 / 12
Licence: other
First seen on the Hub: 2023-10-24
Training datasets: undisclosed
Added to our catalog: 2026-07-28

Compare with

bge-small-en-v1.5 70.9 M bge-large-en-v1.5 12.7 M bge-base-en-v1.5 9.9 M Qwen3-Embedding-0.6B 9.3 M all-MiniLM-L6-v2 2.8 M bge-base-en-v1.5 2.2 M

Appears in

Best open coding models Commercial-use text models Models that fit in 12 GB