Lily-MoE-2x7b

Lily-MoE-2x7b is a Mixure of Experts (MoE) made with the following models using LazyMergekit:

• cognitivecomputations/dolphin-2.6-mistral-7b-dpo-laser
• LunaticPython161/CyberWitch-7B

🧩 Configuration

base_model: cognitivecomputations/dolphin-2.6-mistral-7b-dpo-laser
gate_mode: hidden # one of "hidden", "cheap_embed", or "random"
dtype: bfloat16 # output dtype (float32, float16, or bfloat16)
experts:
  - source_model: cognitivecomputations/dolphin-2.6-mistral-7b-dpo-laser
    positive_prompts:
    - "chat"
    - "assistant"
    - "tell me"
    - "explain"
    - "code"
    - "programming"
  - source_model: LunaticPython161/CyberWitch-7B
    positive_prompts:
    - "solve"
    - "count"
    - "math"
    - "mathematics"
    - "algorithm"
    - "cypher"
    - "cybersecurity"
    - "penetration testing"
    - "red team"
    - "blue team"
    - "hacking"

💻 Usage

!pip install -qU transformers bitsandbytes accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "LunaticPython161/Lily-MoE-2x7b"

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"])