metadata
tags:
- merge
- mergekit
- moe
- frankenmoe
- abacusai/Llama-3-Smaug-8B
- cognitivecomputations/dolphin-2.9-llama3-8b
- Weyaxi/Einstein-v6.1-Llama3-8B
- dreamgen-preview/opus-v1.2-llama-3-8b-base-run3.4-epoch2
base_model:
- abacusai/Llama-3-Smaug-8B
- cognitivecomputations/dolphin-2.9-llama3-8b
- Weyaxi/Einstein-v6.1-Llama3-8B
- dreamgen-preview/opus-v1.2-llama-3-8b-base-run3.4-epoch2
🚀 Skyro-4X8B
Skyro-4X8B is a Mixure of Experts (MoE) made with the following models using Mergekit:
- abacusai/Llama-3-Smaug-8B
- cognitivecomputations/dolphin-2.9-llama3-8b
- Weyaxi/Einstein-v6.1-Llama3-8B
- dreamgen-preview/opus-v1.2-llama-3-8b-base-run3.4-epoch2
🧩 Configuration
base_model: meta-llama/Meta-Llama-3-8B
gate_mode: hidden
experts:
- source_model: abacusai/Llama-3-Smaug-8B
positive_prompts:
- "chat"
- "assistant"
- "tell me"
- "explain"
- "I want"
- source_model: cognitivecomputations/dolphin-2.9-llama3-8b
positive_prompts:
- "math"
- "mathematics"
- "code"
- "engineering"
- "solve"
- "logic"
- "rationality"
- "puzzle"
- "solve"
- source_model: Weyaxi/Einstein-v6.1-Llama3-8B
positive_prompts:
- "science"
- "medical"
- "physics"
- "engineering"
- "math"
- "logic"
- "rationality"
- "mathematics"
- "solve"
- source_model: dreamgen-preview/opus-v1.2-llama-3-8b-base-run3.4-epoch2
positive_prompts:
- "story"
- "roleplay"
- "role-play"
- "storywriting"
- "character"
- "narrative"
- "creative"
💻 Usage
!pip install -qU transformers accelerate
from transformers import AutoTokenizer
import transformers
import torch
model = "saucam/Skyro-4X8B"
messages = [{"role": "user", "content": "What is a large language model?"}]
tokenizer = AutoTokenizer.from_pretrained(model)
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
pipeline = transformers.pipeline(
"text-generation",
model=model,
torch_dtype=torch.float16,
device_map="auto",
)
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"])