---
language:
  - en
license: apache-2.0
library_name: transformers
tags:
  - mixture-of-experts
  - MoE
  - coding
  - 8-bit
  - bitsandbytes
  - quantized
  - LFM2
  - Qwen
base_model:
  - Akahsizrr/fuse-1-Lite
  - LiquidAI/LFM2.5-2.6B
  - Qwen/Qwen3.6-35B-A3B
pipeline_tag: text-generation
inference: false
---

# fuse-1 Lite — 8-bit Quantized

> **8-bit quantized version of [fuse-1 Lite](https://huggingface.co/Akahsizrr/fuse-1-Lite). 6.00 GB VRAM — runs on T4, L4, and consumer GPUs.**

This is the bitsandbytes 8-bit quantized version of fuse-1 Lite, offering a balance between precision and memory usage.

## Usage

```python
from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
import torch

bnb_config = BitsAndBytesConfig(load_in_8bit=True)

model = AutoModelForCausalLM.from_pretrained(
    "Akahsizrr/fuse-1-Lite-8bit",
    quantization_config=bnb_config,
    device_map="auto",
    trust_remote_code=True,
)
tokenizer = AutoTokenizer.from_pretrained("Akahsizrr/fuse-1-Lite-8bit")

messages = [{"role": "user", "content": "Write a Python function to check if a number is prime."}]
text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = tokenizer(text, return_tensors="pt").to(model.device)

with torch.no_grad():
    outputs = model.generate(**inputs, max_new_tokens=512, do_sample=True, temperature=0.1)

print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:], skip_special_tokens=True))
```

## VRAM Requirements

| Precision | VRAM | GPU |
|-----------|------|-----|
| 4-bit | 3.36 GB | T4, RTX 3060, M2 Pro |
| **8-bit (this model)** | **6.00 GB** | T4, L4, RTX 3060 |
| bfloat16 | ~12 GB | L4, A10G, RTX 4090 |

See the [main model card](https://huggingface.co/Akahsizrr/fuse-1-Lite) for full architecture details, training info, and technical report.