README.md

metadata

license: apache-2.0
tags:
  - moe
  - frankenmoe
  - merge
  - mergekit
  - lazymergekit
  - cognitivecomputations/dolphin-2.8-mistral-7b-v02
  - OpenPipe/mistral-ft-optimized-1227
base_model:
  - cognitivecomputations/dolphin-2.8-mistral-7b-v02
  - OpenPipe/mistral-ft-optimized-1227

dolphin-2.8-mistral-v02-2x7b

dolphin-2.8-mistral-v02-2x7b is a Mixture of Experts (MoE) made with the following models using LazyMergekit:

• cognitivecomputations/dolphin-2.8-mistral-7b-v02
• OpenPipe/mistral-ft-optimized-1227

🧩 Configuration

base_model: cognitivecomputations/dolphin-2.8-mistral-7b-v02
experts:
  - source_model: cognitivecomputations/dolphin-2.8-mistral-7b-v02
    positive_prompts:
      - "chat"
      - "assistant"
      - "tell me"
      - "explain"
      - "I want"
  - source_model: OpenPipe/mistral-ft-optimized-1227
    positive_prompts:
      - "code"
      - "python"
      - "javascript"
      - "programming"
      - "algorithm"

💻 Usage

!pip install -qU transformers bitsandbytes accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "frost19k/dolphin-2.8-mistral-v02-2x7b"

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"])