# BICLab/SpikingBrain-7B

Spiking Brain-inspired Large Models, integrating hybrid efficient attention, MoE modules and spike encoding into its architecture

Repository: https://github.com/BICLab/SpikingBrain-7B
Canonical: https://ross.abutalabs.com/products/spikingbrain-7b
Language: Python
License: MIT
License Family: permissive
Last push: 2026-05-14T09:52:16+00:00

## Health v2 (maintenance only)
Score: 54/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 82, release rhythm 35, longevity 26
- inputs: {"age_days": 364, "days_push": 111, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1369, forks 189 (observed 2026-08-28T04:04:31.721185+00:00)

## What it is
SpikingBrain-7B is a brain-inspired large language model that combines hybrid efficient attention, MoE modules, and spike encoding, with a conversion pipeline for continual pre-training from open-source models. The repo provides HuggingFace weights, vLLM inference plugins, and quantized (W8ASpike) versions, with support for non-NVIDIA MetaX GPU clusters.

## Use cases
- run a spiking brain-inspired LLM locally
- efficient long-context LLM inference with vLLM
- continual pre-training of LLMs on limited data
- research neuromorphic and spike-based model architectures
- deploy LLMs on non-NVIDIA GPU clusters
- quantized LLM inference with sparsity

## When to choose
- researching brain-inspired or spiking LLM architectures
- needing efficient long-sequence (million-token) inference
- training on MetaX or non-NVIDIA hardware
- wanting pretrained 7B weights with vLLM support

## When to avoid
- needing a mainstream production LLM with broad ecosystem support
- requiring NVIDIA-only optimized tooling without plugin setup
- looking for a general-purpose chat model rather than research code

## Facets
- artifact type: library
- maturity: active
- function: llm-inference, llm-training, machine-learning, deep-learning
- domain: large-language-models, deep-learning, artificial-intelligence, gpu-computing
- platform: python
- tags: spiking-neural-networks, neuromorphic, mixture-of-experts, efficient-attention, vllm-plugin, model-weights, quantization, metax-gpu, gpu, linux, docker

## Member repositories
- BICLab/SpikingBrain-7B (main) score 54

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:31.721185+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:40:56.811692+00:00, confidence not recorded.
  - readme: https://github.com/BICLab/SpikingBrain-7B (fetched 2026-08-28T04:04:31.721185+00:00, sha 1d4a832cb706)
- Data as of 2026-08-30T08:39:29.467469+00:00.