CultriX_-_MonaTrix-v4-8bits

Quantization made by Richard Erkhov.

Github

Discord

Request more models

MonaTrix-v4 - bnb 8bits

• Model creator: https://huggingface.co/CultriX/
• Original model: https://huggingface.co/CultriX/MonaTrix-v4/

Original model description:

tags:

• merge
• mergekit
• lazymergekit
• Kukedlc/NeuralMaxime-7B-slerp
• eren23/ogno-monarch-jaskier-merge-7b
• eren23/dpo-binarized-NeutrixOmnibe-7B base_model:
• Kukedlc/NeuralMaxime-7B-slerp
• eren23/ogno-monarch-jaskier-merge-7b
• eren23/dpo-binarized-NeutrixOmnibe-7B license: apache-2.0

---

MonaTrix-v4

MonaTrix-v4 is a merge of the following models using LazyMergekit:

• Kukedlc/NeuralMaxime-7B-slerp
• eren23/ogno-monarch-jaskier-merge-7b
• eren23/dpo-binarized-NeutrixOmnibe-7B

🧩 Configuration

models:
  - model: mistralai/Mistral-7B-v0.1
    # No parameters necessary for base model
  - model: Kukedlc/NeuralMaxime-7B-slerp
    #Emphasize the beginning of Vicuna format models
    parameters:
      weight: 0.36
      density: 0.65
  - model: eren23/ogno-monarch-jaskier-merge-7b
    parameters:
      weight: 0.34
      density: 0.6
  # Vicuna format
  - model: eren23/dpo-binarized-NeutrixOmnibe-7B
    parameters:
      weight: 0.3
      density: 0.6

merge_method: dare_ties
base_model: mistralai/Mistral-7B-v0.1
parameters:
  int8_mask: true
dtype: bfloat16
random_seed: 0

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "CultriX/MonaTrix-v4"
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"])