README.md

---
license: apache-2.0
tags:
- moe
- frankenmoe
- merge
- mergekit
- lazymergekit
---

# Lumina-3.5

Lumina-3.5 is a Mixture of Experts (MoE) made with [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing). This model uses a context window of up to 32k.

## 🏆 Open LLM Leaderboard Evaluation Results 

|             Metric              |Value|
|---------------------------------|----:|
|Avg.                             |75.41|
|AI2 Reasoning Challenge (25-Shot)|71.59|
|HellaSwag (10-Shot)              |88.82|
|MMLU (5-Shot)                    |64.48|
|TruthfulQA (0-shot)              |75.66|
|Winogrande (5-shot)              |83.98|
|GSM8k (5-shot)                   |67.93|

# Quantized

Special thanks to GGUFs made by [mradermacher](https://huggingface.co/mradermacher)
* [mradermacher/Lumina-3.5-GGUF](https://huggingface.co/mradermacher/Lumina-3.5-GGUF)

## 💻 Usage

```python
!pip install -qU transformers bitsandbytes accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "Ppoyaa/Lumina-3.5"

tokenizer = AutoTokenizer.from_pretrained(model)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    model_kwargs={"torch_dtype": torch.float16, "load_in_4bit": True},
)

messages = [{"role": "user", "content": "Explain what a Mixture of Experts is in less than 100 words."}]
prompt = pipeline.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])
```