Exl2 version of Undi95/OpenDolphinMaid-4x7b

branch

main : 8bpw h8
b8h8 : 8bpw h8

Using ThePile 0007.parquet as dataset

Quantization settings : python convert.py -i models/flemmingmiguel_TurdusDareBeagle-7B -o TurdusDareBeagle-7B-temp -cf TurdusDareBeagle-7B-8bpw-h8-exl2 -c 0007.parquet -l 8192 -b 8 -hb 8 -ml 8192

below this line is original readme

TurdusDareBeagle-7B

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

As an experiment to find the best base merge to further fine-tuning, expect a lot of experiments named using parts of the component models until a clear winner emerges in the benchmarks

In this case .

🧩 Configuration

slices:
    - sources:
      - model: udkai/Turdus
        layer_range: [0, 32]
      - model: flemmingmiguel/DareBeagle-7B
        layer_range: [0, 32]
merge_method: slerp
base_model: flemmingmiguel/DareBeagle-7B
parameters:
  t:
    - filter: self_attn
      value: [0, 0.5, 0.3, 0.7, 1]
    - filter: mlp
      value: [1, 0.5, 0.7, 0.3, 0]
    - value: 0.45 # fallback for rest of tensors
dtype: float16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "flemmingmiguel/TurdusDareBeagle-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"])