Model Intelligence Sheet
1bit-MONSTER/Zamba2-2.7B-instruct-GGUF overview
Zamba2 2.7B instruct — GGUF GGUF conversion of Zyphra's Zamba2 2.7B instruct https://huggingface.co/Zyphra/Zamba2 2.7B instruct : Mamba 2 layers with shared tr…
Runs locally from ~3.79 GB disk (4 GB VRAM class GPUs with llama.cpp / guIDE).
Repository Files & Downloads
1 GGUF files detected
Direct downloads for local inference
| File | Type | Quantization | Size | Link |
|---|---|---|---|---|
| Zamba2-2.7B-instruct-Q8_0.gguf | GGUF | Q8_0 | 3.79 GB | Download |
Model Details
| Model ID | 1bit-MONSTER/Zamba2-2.7B-instruct-GGUF |
|---|---|
| Author | 1bit-MONSTER |
| Pipeline | — |
| License | apache-2.0 |
| Base model | Zyphra/Zamba2-2.7B-instruct |
| Last modified | 2026-09-26T22:17:27.000Z |
Model README
---
license: apache-2.0
base_model:
- Zyphra/Zamba2-2.7B-instruct
tags:
- gguf
- zamba2
- mamba
- hybrid
---
Zamba2-2.7B-instruct — GGUF
GGUF conversion of Zyphra's Zamba2-2.7B-instruct: Mamba-2 layers with shared transformer blocks. It runs from our llama.cpp fork, which carries the Zamba2 model (ggml-org/llama.cpp#21412). The fork also adds a Vulkan Mamba-2 scan for this model's state size and fixes a converter bug that gave every token score -1000.
Contents
Zamba2-2.7B-instruct-Q8_0.gguf
Validation
- Converted after the token-score fix: tokenization matches the model's own tokenizer.
- Measured with
llama-benchon Vulkan (Radeon 8060S), Q8_0: prefill 1541 tok/s (pp256), decode 45.2 tok/s (tg64). Chat answers are correct.
Running it
With the 1bit engine:
1bit serve -m Zamba2-2.7B-instruct-Q8_0.gguf --device vulkanAttribution
- Base model: Zyphra/Zamba2-2.7B-instruct, Apache 2.0.
- GGUF conversion and validation: the 1bit engine project, with the converter and model code in our llama.cpp fork.
- License: Apache 2.0, inherited from the base model.
Run 1bit-MONSTER/Zamba2-2.7B-instruct-GGUF with guIDE
Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.
Source: Hugging Face · Compare models