sd_smartprocess/interrogators/llava_interrogator.py

91 lines
3.1 KiB
Python

import gc
import logging
import os
import torch
from PIL import Image
from transformers import AutoTokenizer
from extensions.sd_smartprocess.interrogators.interrogator import Interrogator
from extensions.sd_smartprocess.model_download import fetch_model
from extensions.sd_smartprocess.mplug_owl import MplugOwlForConditionalGeneration, MplugOwlImageProcessor, \
MplugOwlProcessor
from extensions.sd_smartprocess.process_params import ProcessParams
logger = logging.getLogger(__name__)
class LLAVAInterrogator(Interrogator):
model = None
processor = None
params = {"max_tokens": 75}
def __init__(self, params: ProcessParams):
super().__init__(params)
print("Initializing LLM model...")
model_path = fetch_model('MAGAer13/mplug-owl-llama-7b', "llm")
model_config = os.path.join(model_path, "config.json")
print(f"Loading model from {model_path}...")
self.model = MplugOwlForConditionalGeneration.from_pretrained(
model_path,
torch_dtype=torch.bfloat16,
config=model_config,
)
print("Loading tokenizer...")
self.image_processor = MplugOwlImageProcessor.from_pretrained(model_path)
print("Loading processor...")
self.tokenizer = AutoTokenizer.from_pretrained(model_path)
print("Initializing processor...")
self.processor = MplugOwlProcessor(self.image_processor, self.tokenizer)
logger.debug("Initialized LLM model.")
def interrogate(self, image: Image, params=None, unload: bool = False) -> str:
self.load()
if params is None:
params = {}
raw_image = image.convert('RGB')
max_tokens = params.max_tokens
generate_kwargs = {
'do_sample': True,
'top_k': 5,
'max_length': max_tokens,
}
images = [raw_image]
prompts = ["Human: <image>",
"Human: Give a short one sentence caption for this image with NO punctuation. DO NOT USE ANY PUNCTUATION OR COMMAS. DO NOT USE COMMAS!!!",
f"AI:"]
logger.debug("Processing inputs...")
inputs = self.processor(text=prompts, images=images, return_tensors='pt')
inputs = {k: v.bfloat16() if v.dtype == torch.float else v for k, v in inputs.items()}
inputs = {k: v.to(self.model.device) for k, v in inputs.items()}
logger.debug("Generating response...")
with torch.no_grad():
res = self.model.generate(**inputs, **generate_kwargs)
caption = self.tokenizer.decode(res.tolist()[0], skip_special_tokens=True)
if "," in caption:
parts = caption.split(",")
caption = " ".join([part.strip() for part in parts if part.strip() != ""])
logger.debug(f"Caption: {caption}")
if unload:
self._to_cpu()
return caption
def _to_cpu(self):
self.model.to('cpu')
def _to_gpu(self):
self.model.to(self.device)
def unload(self):
self._to_cpu()
if torch.cuda.is_available():
torch.cuda.empty_cache()
gc.collect()
def load(self):
self._to_gpu()