README.md

metadata

license: cc-by-nc-4.0
base_model:
  - SvalTek/ColdBrew-Aphid
  - SvalTek/ColdBrew-Aphid
tags:
  - moe
  - frankenmoe
  - merge
  - mergekit
  - lazymergekit
  - SvalTek/ColdBrew-Aphid

ColdBrew-Icarus

ColdBrew-Icarus is a Mixture of Experts (MoE) made with the following models using LazyMergekit:

• SvalTek/ColdBrew-Aphid
• SvalTek/ColdBrew-Aphid

🧩 Configuration

base_model: SvalTek/ColdBrew-Aphid
gate_mode: random
dtype: bfloat16
experts:
  - source_model: SvalTek/ColdBrew-Aphid
  - source_model: SvalTek/ColdBrew-Aphid

💻 Usage

!pip install -qU transformers bitsandbytes accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "SvalTek/ColdBrew-Icarus"

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"])