NeuralQwen-2.5-1.5B-Spanish
NeuralQwen-2.5-1.5B-Spanish is a merge of the following models using LazyMergekit:
🧩 Configuration
models:
- model: Qwen/Qwen2.5-1.5B
# No parameters necessary for base model
- model: Qwen/Qwen2.5-1.5B-Instruct
parameters:
density: 0.66
weight: 0.6
- model: Kukedlc/Qwen2.5-1.5B-Spanish-1.0-DPO
parameters:
density: 0.44
weight: 0.4
merge_method: dare_ties
base_model: Qwen/Qwen2.5-1.5B
parameters:
int8_mask: true
dtype: float16
💻 Usage
!pip install -qU transformers accelerate
from transformers import AutoTokenizer
import transformers
import torch
model = "Kukedlc/NeuralQwen-2.5-1.5B-Spanish"
messages = [{"role": "system", "content": "Eres un asistente de pensamiento logico que piensa paso a paso, por cada pregunta que te hagan deberes comprobar la respuesta por 3 metodos diferentes."},
{"role": "user", "content": "Cuantas letras 'r' tiene la palabra strawberry?"}]
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"])
