HsuanLLM's picture
Update README.md
d41ba30 verified
|
raw
history blame
1.58 kB
metadata
tags:
  - merge
  - mergekit
  - lazymergekit
  - cognitivecomputations/dolphin-2.6-mistral-7b-dpo
base_model:
  - cognitivecomputations/dolphin-2.6-mistral-7b-dpo

BreezeDolphinDPO-7B-DARE

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

🧩 Configuration

models:
  - model: MediaTek-Research/Breeze-7B-Instruct-v0_1
    # No parameters necessary for base model
  - model: cognitivecomputations/dolphin-2.6-mistral-7b-dpo
    parameters:
      density: 0.5
      weight: 0.5
merge_method: dare_ties
tokenizer_source: base
base_model: MediaTek-Research/Breeze-7B-Instruct-v0_1
parameters:
  #normalize: false
  int8_mask: true
dtype: bfloat16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "HsuanLLM/BreezeDolphinDPO-7B-DARE"
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"])