MI

Lora8B

MinimaxGGUF by pytravelerUpdated 2026-08-18
Download on Hugging Face
Parameters700M
Context—
Layers—
Architectureqwen3vl
Smallest quant0.7 GB
Quick answer

Lora8B needs about 1.3 GB of memory at Q8_0 (a 0.7 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 2 quants, from 0.7 GB (Q8_0) to 1.4 GB (F16).

LoRA adapter for MiniMax‑H3 prompt rewriting with Qwen3‑VL‑8B base

Lora8B is a LoRA adapter that rewrites user requests into production‑oriented prompts for MiniMax‑H3 audio‑video generation. It is built on Qwen3‑VL‑8B‑Instruct and supports text‑only and keyframe‑conditioned rewriting. The adapter is delivered as a GGUF file for llama.cpp.

Good at

  • Text‑to‑audio‑video rewriting
  • Keyframe‑conditioned prompt generation
  • Structured prompt with shot timeline

Limits

  • No training data included
  • No license information provided
  • Requires base Qwen3‑VL‑8B‑Instruct weights

Best for

  • Prompt rewriting for MiniMax‑H3 video generation
  • Text‑only prompt generation
  • Keyframe‑conditioned prompt creation

Who it’s for. Developers and content creators who need to convert short user requests into detailed prompts for MiniMax‑H3 audio‑video synthesis. The adapter works with llama.cpp and requires the Qwen3‑VL‑8B‑Instruct base model.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q8_00.7 GB 1.3 GB
Fits Get file →
F161.4 GB 2.1 GB
Fits Get file →

For 8 GB VRAM, use the Q8_0 quantization (file size 0.74 GB, estimated 1.3 GB memory at 8 k context). If more VRAM is available, F16 (1.4 GB, 2.1 GB memory) offers higher precision. Both formats are supported by llama.cpp.

What GPU do you need for Lora8B?

These figures add the file size, an fp16 KV cache at an 8K context window 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 F16 (1.4 GB file, about 2.1 GB in use).

Which Lora8B quant should you choose?

Q8_0 (0.7 GB) is the middle option in this repository. Close to lossless. It costs roughly twice the memory of a 4-bit file, so it suits cases where you have plenty of VRAM.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureqwen3vl
Parameters700M

Source

Quantized bypytraveler
FamilyMinimax
Downloads (HF)2,082
Last updated2026-08-18

Frequently asked questions about Lora8B

How much VRAM does Lora8B need?

About 1.3 GB at Q8_0 with an 8K context window. That is the 0.7 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 Lora8B on an 8 GB GPU?

Yes, with F16: the 1.4 GB file needs about 2.1 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Lora8B GGUF?

It depends on the quant: Q8_0 is 0.7 GB, F16 is 1.4 GB.

Where can I download Lora8B GGUF?

The files are on Hugging Face in the pytraveler/MiniMax-H3-Prompt-Rewriter-LoRA-8B-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

Can I run this on 8 GB VRAM?

Yes, the Q8_0 quant uses about 1.3 GB memory at 8 k context, well within 8 GB VRAM.

Does it support vision inputs?

Yes, the adapter reads reference frames from the Qwen3‑VL‑8B base model.

Is this a full model?

No, it is only a LoRA adapter; the base Qwen3‑VL‑8B‑Instruct weights are required separately.

Similar models

Browse all →

More from pytraveler

All by pytraveler →

More Minimax models

All Minimax →

Other tiny models

All tiny models →