Spaetzle V60 7b

🚀 Spaetzle-v60-7b

Spaetzle-v60-7b is a progressive merge model (mostly using dare-ties, and also slerp in some cases). It aims to strike a suitable balance for English and German local tasks.

Spaetzle-v60-7b is a merge of the following models using LazyMergekit:

• abideen/AlphaMonarch-dora
• cstr/Spaetzle-v58-7b

✨ Features

• Multi - language Adaptability: Suitable for both English and German local tasks, offering a balanced performance.
• Progressive Merge: Utilizes a combination of dare - ties and slerp techniques for merging models.

📚 Documentation

Benchmarks

The performance looks ok so far: e.g., in EQ - Bench, we get Score (v2_de): 65.08 (Parseable: 171.0).

From the Occiglot Euro LLM Leaderboard:

And for the int4 - inc quantized version, from Low - bit Quantized Open LLM Leaderboard:

Contamination check results (reference model: Mistral instruct 7b v0.1):

• MMLU: result < 0.1, %: 0.19
• TruthfulQA: result < 0.1, %: 0.34
• GSM8k: result < 0.1, %: 0.39

🧩 Configuration

models:
  - model: cstr/Spaetzle-v58-7b
    # no parameters necessary for base model
  - model: abideen/AlphaMonarch-dora
    parameters:
      density: 0.60
      weight: 0.30
merge_method: dare_ties
base_model: cstr/Spaetzle-v58-7b
parameters:
  int8_mask: true
dtype: bfloat16
random_seed: 0
tokenizer_source: base

💻 Usage Examples

Basic Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "cstr/Spaetzle-v60-7b"
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"])

📄 License

This project is licensed under the cc - by - nc - 4.0 license.