Kukedlc / NeuralTopBench-7B-ties

NeuralTopBench-7B-ties

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

• Kukedlc/Neural4gsm8k
• nlpguy/AlloyIngotNeoX
• automerger/OgnoExperiment27-7B
• vanillaOVO/supermario_v4

🧩 Configuration

models:
  - model: CultriX/NeuralTrix-bf16
    # no parameters necessary for base model
  - model: Kukedlc/Neural4gsm8k
    parameters:
      weight: 0.3
      density: 0.5
  - model: nlpguy/AlloyIngotNeoX
    parameters:
      weight: 0.2
      density: 0.5
  - model: automerger/OgnoExperiment27-7B
    parameters:
      weight: 0.2
      density: 0.5
  - model: vanillaOVO/supermario_v4
    parameters:
      weight: 0.3
      density: 0.5
merge_method: dare_ties
base_model: CultriX/NeuralTrix-bf16

parameters:
  int8_mask: true
  normalize: true
dtype: bfloat16

💻 Usage - Stream

# Requirements
!pip install -qU transformers accelerate bitsandbytes

# Imports & settings
from transformers import AutoModelForCausalLM, AutoTokenizer, TextStreamer
import warnings
import os
os.environ["TOKENIZERS_PARALLELISM"] = "false"
warnings.filterwarnings('ignore')

# Model & Tokenizer
MODEL_NAME = 'Kukedlc/NeuralTopBench-7B-ties'
model = AutoModelForCausalLM.from_pretrained(MODEL_NAME, device_map='cuda:0', load_in_4bit=True)
tok = AutoTokenizer.from_pretrained(MODEL_NAME)

# Inference
prompt = "I want you to generate a theory that unites quantum mechanics with the theory of relativity and cosmic consciousness\n"
inputs = tok([prompt], return_tensors="pt").to('cuda')
streamer = TextStreamer(tok)

# Despite returning the usual output, the streamer will also print the generated text to stdout.
_ = model.generate(**inputs, streamer=streamer, max_new_tokens=512, do_sample=True, num_beams=1, top_p=0.9, temperature=0.7)

💻 Usage - Clasic

!pip install -qU transformers bitsandbytes accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = 'Kukedlc/NeuralTopBench-7B-ties'

tokenizer = AutoTokenizer.from_pretrained(model)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    model_kwargs={"torch_dtype": torch.float16, "load_in_4bit": True},
)

messages = [{"role": "user", "content": "Explain what a Mixture of Experts is in less than 100 words."}]
prompt = pipeline.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
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"])