README.md

---
license: apache-2.0
tags:
- moe
- frankenmoe
- merge
- mergekit
- lazymergekit
- vihangd/DopeyTinyLlama-1.1B-v1
- phanerozoic/Tiny-Pirate-1.1b-v0.1
- Tensoic/TinyLlama-1.1B-3T-openhermes
base_model:
- vihangd/DopeyTinyLlama-1.1B-v1
- phanerozoic/Tiny-Pirate-1.1b-v0.1
- Tensoic/TinyLlama-1.1B-3T-openhermes
---

# Tinyllama-moe1

Tinyllama-moe1 is a Mixture of Experts (MoE) made with the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
* [vihangd/DopeyTinyLlama-1.1B-v1](https://huggingface.co/vihangd/DopeyTinyLlama-1.1B-v1)
* [phanerozoic/Tiny-Pirate-1.1b-v0.1](https://huggingface.co/phanerozoic/Tiny-Pirate-1.1b-v0.1)
* [Tensoic/TinyLlama-1.1B-3T-openhermes](https://huggingface.co/Tensoic/TinyLlama-1.1B-3T-openhermes)

## 🧩 Configuration

```yaml
base_model: vihangd/DopeyTinyLlama-1.1B-v1
experts:
  - source_model: vihangd/DopeyTinyLlama-1.1B-v1
    positive_prompts:
    - "chat"
    - "assistant"
    - "explain"
  - source_model: phanerozoic/Tiny-Pirate-1.1b-v0.1
    positive_prompts:
    - "storywriting"
    - "scene"
    - "story"
  - source_model: Tensoic/TinyLlama-1.1B-3T-openhermes
    positive_prompts:
    - "reason"
    - "instruct"
    - "count"
```

## 💻 Usage

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

from transformers import AutoTokenizer
import transformers
import torch

model = "aipib/Tinyllama-moe1"

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"])
```