metadata

tags:
  - merge
  - mergekit
  - lazymergekit
  - Nexusflow/Starling-LM-7B-beta
  - timpal0l/Mistral-7B-v0.1-flashback-v2-instruct
  - mlabonne/NeuralBeagle14-7B
base_model:
  - Nexusflow/Starling-LM-7B-beta
  - timpal0l/Mistral-7B-v0.1-flashback-v2-instruct
  - mlabonne/NeuralBeagle14-7B

SwedishBeagleDare

SwedishBeagleDare is a merge of the following models using LazyMergekit:

🧩 Configuration

models:
  - model: Nexusflow/Starling-LM-7B-beta
    parameters:
      weight: 0.5
  - model: timpal0l/Mistral-7B-v0.1-flashback-v2-instruct
    parameters:
      weight: 0.5
  - model: mlabonne/NeuralBeagle14-7B
    parameters:
      weight: 0.5
merge_method: task_arithmetic
base_model: mlabonne/NeuralBeagle14-7B
parameters:
  int8_mask: 1.0
  normalize: true
dtype: bfloat16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "Knobi3/SwedishBeagleDare"
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"])