Quantization made by Richard Erkhov.

Maxine-7B-0401-stock - GGUF

Model creator: https://huggingface.co/louisbrulenaudet/
Original model: https://huggingface.co/louisbrulenaudet/Maxine-7B-0401-stock/

Name	Quant method	Size
Maxine-7B-0401-stock.Q2_K.gguf	Q2_K	2.53GB
Maxine-7B-0401-stock.IQ3_XS.gguf	IQ3_XS	2.81GB
Maxine-7B-0401-stock.IQ3_S.gguf	IQ3_S	2.96GB
Maxine-7B-0401-stock.Q3_K_S.gguf	Q3_K_S	2.95GB
Maxine-7B-0401-stock.IQ3_M.gguf	IQ3_M	3.06GB
Maxine-7B-0401-stock.Q3_K.gguf	Q3_K	3.28GB
Maxine-7B-0401-stock.Q3_K_M.gguf	Q3_K_M	3.28GB
Maxine-7B-0401-stock.Q3_K_L.gguf	Q3_K_L	3.56GB
Maxine-7B-0401-stock.IQ4_XS.gguf	IQ4_XS	3.67GB
Maxine-7B-0401-stock.Q4_0.gguf	Q4_0	3.83GB
Maxine-7B-0401-stock.IQ4_NL.gguf	IQ4_NL	3.87GB
Maxine-7B-0401-stock.Q4_K_S.gguf	Q4_K_S	3.86GB
Maxine-7B-0401-stock.Q4_K.gguf	Q4_K	4.07GB
Maxine-7B-0401-stock.Q4_K_M.gguf	Q4_K_M	4.07GB
Maxine-7B-0401-stock.Q4_1.gguf	Q4_1	4.24GB
Maxine-7B-0401-stock.Q5_0.gguf	Q5_0	4.65GB
Maxine-7B-0401-stock.Q5_K_S.gguf	Q5_K_S	4.65GB
Maxine-7B-0401-stock.Q5_K.gguf	Q5_K	4.78GB
Maxine-7B-0401-stock.Q5_K_M.gguf	Q5_K_M	4.78GB
Maxine-7B-0401-stock.Q5_1.gguf	Q5_1	5.07GB
Maxine-7B-0401-stock.Q6_K.gguf	Q6_K	5.53GB
Maxine-7B-0401-stock.Q8_0.gguf	Q8_0	7.17GB

Original model description:

tags: - merge - mergekit - MTSAIR/multi_verse_model - rwitz/experiment26-truthy-iter-0 - MaziyarPanahi/Calme-7B-Instruct-v0.2 - chemistry - biology - math base_model: - MTSAIR/multi_verse_model - rwitz/experiment26-truthy-iter-0 - MaziyarPanahi/Calme-7B-Instruct-v0.2 license: apache-2.0 language: - en library_name: transformers pipeline_tag: text-generation model-index: - name: Maxine-7B-0401-stock results: - task: type: text-generation metrics: - name: Average type: Average value: 76.73 - name: ARC type: ARC value: 73.12 - name: GSM8K type: GSM8K value: 70.66 - name: Winogrande type: Winogrande value: 85 - name: TruthfulQA type: TruthfulQA value: 78.07 - name: HellaSwag type: HellaSwag value: 89.13 source: name: Open LLM Leaderboard url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard

Maxine-7B-0401-stock, an xtraordinary 7B model

03-22-2024 - To date, louisbrulenaudet/Pearl-34B-ties is the "Best 🤝 base merges and moerges model of around 30B" on the Open LLM Leaderboard.

Configuration

models:
    - model: OpenPipe/mistral-ft-optimized-1227
    - model: MTSAIR/multi_verse_model
    - model: rwitz/experiment26-truthy-iter-0
    - model: MaziyarPanahi/Calme-7B-Instruct-v0.2
merge_method: model_stock
base_model: OpenPipe/mistral-ft-optimized-1227
dtype: bfloat16

Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "louisbrulenaudet/Maxine-7B-0401-stock"
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"])

Citing & Authors

If you use this code in your research, please use the following BibTeX entry.

@misc{louisbrulenaudet2024,
  author =       {Louis Brulé Naudet},
  title =        {Maxine-7B-0401-stock, an xtraordinary 7B model},
  year =         {2024}
  howpublished = {\url{https://huggingface.co/louisbrulenaudet/Maxine-7B-0401-stock}},
}

Feedback

If you have any feedback, please reach out at louisbrulenaudet@icloud.com.