models:
  - model: Kaoeiri/L3MaidRPKraiKei-V1.6-8B
    parameters:
      density: .5
      weight: 1
  - model: openlynn/Llama-3-Soliloquy-8B-v2
    parameters:
      density: .35
      weight: .4
  - model: Sao10K/L3-8B-Stheno-v3.2
    parameters:
      density: .5
      weight: .7
merge_method: ties

base_model: mlabonne/NeuralDaredevil-8B-abliterated
parameters:
  normalize: true
  int8_mask: true
dtype: bfloat16

💻 Usage

python

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "Kaoeiri/L3MaidRPKraiKei-V1.65-8B"
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"])