BioMistral-Carpybara-Slerp

BioMistral-Carpybara-Slerp is a merge of the following models:

• BioMistral/BioMistral-7B-DARE
• argilla/CapybaraHermes-2.5-Mistral-7B

Evaluations

Benchmark	BioMistral-Carpybara-Slerp	Orca-2-7b	llama-2-7b	meditron-7b	meditron-70b
MedMCQA
ClosedPubMedQA
PubMedQA
MedQA
MedQA4
MedicationQA
MMLU Medical
MMLU
TruthfulQA
GSM8K
ARC
HellaSwag
Winogrande

More details on the Open LLM Leaderboard evaluation results can be found here.

🧩 Configuration

slices:
  - sources:
      - model: BioMistral/BioMistral-7B-DARE
        layer_range: [0, 32]
      - model: argilla/CapybaraHermes-2.5-Mistral-7B
        layer_range: [0, 32]
merge_method: slerp
base_model: argilla/CapybaraHermes-2.5-Mistral-7B
parameters:
  t:
    - filter: self_attn
      value: [0, 0.5, 0.3, 0.7, 1]
    - filter: mlp
      value: [1, 0.5, 0.7, 0.3, 0]
    - value: 0.5 # fallback for rest of tensors
dtype: float16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "Technoculture/BioMistral-Carpybara-Slerp"
messages = [{"role": "user", "content": "I am feeling sleepy these days"}]

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"])