BG

bge-small-en-v1.5

BGEGGUF by CompendiumLabsUpdated 2024-02-17
Download on Hugging Face
Parameters33M
Context512
Layers12
Architecturebert
Smallest quant0.0 GB
Quick answer

bge-small-en-v1.5 needs about 0.7 GB of memory at Q4_K_M (a 0.0 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 4 quants, from 0.0 GB (Q4_K_M) to 0.1 GB (F32).

Small BERT-based embedding model for efficient text representation

bge-small-en-v1.5 is a 35‑million‑parameter BERT model quantized for GGUF. It generates embeddings for English text with a 512‑token context window. The model is lightweight and fast on CPU and GPU with minimal accuracy loss.

Good at

  • Fast embeddings on low‑resource hardware
  • Small file size and low VRAM usage
  • Compatible with llama.cpp and llama‑cpp‑python

Limits

  • No vision or generation capabilities
  • 512‑token context limit
  • MIT license requires attribution

Best for

  • Semantic search indexing
  • Document similarity tasks
  • Embedding‑based clustering

Who it’s for. Developers needing quick, low‑memory embeddings for English text. Ideal for edge devices or batch processing where speed and size matter.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_K_M0.0 GB 0.5 GB
Fits Get file →
Q8_00.0 GB 0.5 GB
Fits Get file →
F160.1 GB 0.6 GB
Fits Get file →
F320.1 GB 0.6 GB
Fits Get file →

For 8 GB VRAM, Q4_K_M (0.02 GB) or Q8_0 (0.04 GB) are suitable, both using ~0.7 GB memory at 8 k context. For higher precision, F16 (0.07 GB) or F32 (0.13 GB) can be used with similar memory usage.

What GPU do you need for bge-small-en-v1.5?

These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.

  • 4 GB VRAM: the best quant that fits is F32 (0.1 GB file, about 0.8 GB in use).

Which bge-small-en-v1.5 quant should you choose?

Q4_K_M is the usual starting point for bge-small-en-v1.5: it balances file size and quality well. Here it is a 0.0 GB file.

The Q8_0 file is 0.0 GB, about 1.5 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturebert
Parameters33M
Layers12
Trained context512 tokens
KV cache (fp16)18 KB per token

Source

Quantized byCompendiumLabs
FamilyBGE
Licensemit
Downloads (HF)160,108
Last updated2024-02-17

Frequently asked questions about bge-small-en-v1.5

How much VRAM does bge-small-en-v1.5 need?

About 0.7 GB at Q4_K_M with an 8K context window. That is the 0.0 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.

Can I run bge-small-en-v1.5 on an 8 GB GPU?

Yes, with F32: the 0.1 GB file needs about 0.8 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of bge-small-en-v1.5 GGUF?

It depends on the quant: Q4_K_M is 0.0 GB, Q8_0 is 0.0 GB, F16 is 0.1 GB, F32 is 0.1 GB.

What context length does bge-small-en-v1.5 support?

The GGUF header lists a trained context length of 512 tokens (512). The KV cache grows with the context you use: about 0.2 GB at 8K and 0.0 GB at 512.

Where can I download bge-small-en-v1.5 GGUF?

The files are on Hugging Face in the CompendiumLabs/bge-small-en-v1.5-gguf repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does bge-small-en-v1.5 use?

Hugging Face lists the license as mit. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Can I run this model on 4 GB VRAM?

Yes, the Q4_K_M quantization uses only about 0.7 GB memory at 8 k context, fitting comfortably in 4 GB VRAM.

Does it support multi‑language embeddings?

No, the model is trained only on English text.

Is there a GPU acceleration option?

Yes, llama.cpp can use GPU for inference, providing modest speedups over CPU.

Similar models

Browse all →

More from CompendiumLabs

All by CompendiumLabs →

More BGE models

All BGE →

Other tiny models

All tiny models →