Hibrid-Llama-Linear

Hibrid-Llama-Linear is a merge of the following models using LazyMergekit:

• ytu-ce-cosmos/Turkish-Llama-8b-Instruct-v0.1
• meta-llama/Llama-3.1-8B
• meta-llama/Llama-3.1-8B-Instruct

🧩 Configuration

model_type: llama
dtype: bfloat16
merge_method: linear
models:
  - model: ytu-ce-cosmos/Turkish-Llama-8b-Instruct-v0.1
    parameters:
      weight: 0.5
  - model: meta-llama/Llama-3.1-8B
    parameters:
      weight: 0.25
  - model: meta-llama/Llama-3.1-8B-Instruct
    parameters:
      weight: 0.25

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "matrixportal/Hibrid-Llama-Linear"
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"])