Wernicke-7B-v8

Wernicke-7B-v8 is a merge of the following models using LazyMergekit:

• kaitchup/Mayonnaise-4in1-022
• macadeliccc/WestLake-7B-v2-laser-truthy-dpo
• vanillaOVO/supermario_v2
• FelixChao/WestSeverus-7B-DPO-v2

🧩 Configuration

models:
 - model: CultriX/Wernicke-7B-v1
 # No parameters necessary for base model
 - model: kaitchup/Mayonnaise-4in1-022
 parameters:
 density: 0.53
 weight: 0.40
 - model: macadeliccc/WestLake-7B-v2-laser-truthy-dpo
 parameters:
 density: 0.53
 weight: 0.25
 - model: vanillaOVO/supermario_v2
 parameters:
 density: 0.53
 weight: 0.25
 - model: FelixChao/WestSeverus-7B-DPO-v2
 parameters:
 density: 0.53
 weight: 0.20
merge_method: dare_ties
base_model: CultriX/Wernicke-7B-v1
parameters:
 int8_mask: true
dtype: float16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "CultriX/Wernicke-7B-v8"
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"])