GPTQ
Collection
quantized LLMs by AutoGPTQ
•
31 items
•
Updated
•
2
MaziyarPanahi/LongAlign-13B-64k-GPTQ is a quantized (GPTQ) version of THUDM/LongAlign-13B-64k
pip install --upgrade accelerate auto-gptq transformers
from transformers import AutoTokenizer, pipeline
from auto_gptq import AutoGPTQForCausalLM, BaseQuantizeConfig
import torch
model_id = "MaziyarPanahi/LongAlign-13B-64k-GPTQ"
quantize_config = BaseQuantizeConfig(
bits=4,
group_size=128,
desc_act=False
)
model = AutoGPTQForCausalLM.from_quantized(
model_id,
use_safetensors=True,
device="cuda:0",
quantize_config=quantize_config)
tokenizer = AutoTokenizer.from_pretrained(model_id)
pipe = pipeline(
"text-generation",
model=model,
tokenizer=tokenizer,
max_new_tokens=512,
temperature=0.7,
top_p=0.95,
repetition_penalty=1.1
)
outputs = pipe("What is a large language model?")
print(outputs[0]["generated_text"])
Base model
THUDM/LongAlign-13B-64k