NinjaDolphin-7B

NinjaDolphin-7B is a merge of the following models using:

• beowolx/CodeNinja-1.0-OpenChat-7B
• beowolx/MistralHermes-CodePro-7B-v1

Improving coding ability from FelixChao/WizardDolphin-7B.

HumanEval (uninstructed and no post-process)

Metric	Value
humaneval-python	52.4390243902439

🧩 Configuration

models:
  - model: FelixChao/WizardDolphin-7B
  - model: beowolx/CodeNinja-1.0-OpenChat-7B
    parameters:
      density: 0.53
      weight: 0.3
  - model: beowolx/MistralHermes-CodePro-7B-v1
    parameters:
      density: 0.53
      weight: 0.3
merge_method: dare_ties
base_model: FelixChao/WizardDolphin-7B
parameters:
  int8_mask: true
dtype: bfloat16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "FelixChao/NinjaDolphin-7B"
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"])

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	69.74
AI2 Reasoning Challenge (25-Shot)	65.61
HellaSwag (10-Shot)	85.35
MMLU (5-Shot)	64.43
TruthfulQA (0-shot)	54.94
Winogrande (5-shot)	80.27
GSM8k (5-shot)	67.85