Kukedlc_-_SuperMente-7B-8bits

Quantization made by Richard Erkhov.

Github

Discord

Request more models

SuperMente-7B - bnb 8bits

• Model creator: https://huggingface.co/Kukedlc/
• Original model: https://huggingface.co/Kukedlc/SuperMente-7B/

Original model description:

tags:

• merge
• mergekit
• lazymergekit
• AetherResearch/Cerebrum-1.0-7b
• Kukedlc/NeuralKybalion-7B-slerp-v3
• mlabonne/AlphaMonarch-7B base_model:
• AetherResearch/Cerebrum-1.0-7b
• Kukedlc/NeuralKybalion-7B-slerp-v3
• mlabonne/AlphaMonarch-7B license: apache-2.0

---

SuperMente-7B

CHAT

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

• AetherResearch/Cerebrum-1.0-7b
• Kukedlc/NeuralKybalion-7B-slerp-v3
• mlabonne/AlphaMonarch-7B

🧩 Configuration

models:
  - model: Kukedlc/NeuralKybalion-7B-slerp-v3
    # no parameters necessary for base model
  - model: AetherResearch/Cerebrum-1.0-7b
    parameters:
      density: 0.55
      weight: 0.3
  - model: Kukedlc/NeuralKybalion-7B-slerp-v3
    parameters:
      density: 0.55
      weight: 0.3
  - model: mlabonne/AlphaMonarch-7B
    parameters:
      density: 0.66
      weight: 0.4
merge_method: dare_ties
base_model: Kukedlc/NeuralKybalion-7B-slerp-v3
parameters:
  int8_mask: true
dtype: bfloat16
random_seed: 0

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "Kukedlc/SuperMente-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"])