RichardErkhov/Gille_-_StrangeMerges_41-7B-dare_ties-8bits

Quantization made by Richard Erkhov.

Github

Discord

Request more models

StrangeMerges_41-7B-dare_ties - bnb 8bits

• Model creator: https://huggingface.co/Gille/
• Original model: https://huggingface.co/Gille/StrangeMerges_41-7B-dare_ties/

Original model description:

license: apache-2.0 tags: - merge - mergekit - lazymergekit - Weyaxi/Einstein-v4-7B - rwitz/experiment26-truthy-iter-0 - kaist-ai/mistral-orpo-beta base_model: - Weyaxi/Einstein-v4-7B - rwitz/experiment26-truthy-iter-0 - kaist-ai/mistral-orpo-beta

StrangeMerges_41-7B-dare_ties

StrangeMerges_41-7B-dare_ties is a merge of the following models using LazyMergekit:

• Weyaxi/Einstein-v4-7B
• rwitz/experiment26-truthy-iter-0
• kaist-ai/mistral-orpo-beta

🧩 Configuration

models:
  - model: Weyaxi/Einstein-v4-7B
    parameters:
      weight: 0.3
      density: 0.8
  - model: rwitz/experiment26-truthy-iter-0
    parameters:
      weight: 0.2
      density: 0.8
  - model: kaist-ai/mistral-orpo-beta
    parameters:
      weight: 0.5
      density: 0.8
base_model: Gille/StrangeMerges_40-7B-dare_ties
merge_method: dare_ties
dtype: bfloat16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "Gille/StrangeMerges_41-7B-dare_ties"
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"])