model/llava/eval/run_llava.py

import argparse
import os
from io import BytesIO

import requests
import torch
from llava.conversation import SeparatorStyle, conv_templates
from llava.model import *
from llava.model.utils import KeywordsStoppingCriteria
from llava.utils import disable_torch_init
from PIL import Image
from transformers import (
    AutoModelForCausalLM,
    AutoTokenizer,
    CLIPImageProcessor,
    CLIPVisionModel,
    StoppingCriteria,
)

DEFAULT_IMAGE_TOKEN = "<image>"
DEFAULT_IMAGE_PATCH_TOKEN = "<im_patch>"
DEFAULT_IM_START_TOKEN = "<im_start>"
DEFAULT_IM_END_TOKEN = "<im_end>"


def load_image(image_file):
    if image_file.startswith("http") or image_file.startswith("https"):
        response = requests.get(image_file)
        image = Image.open(BytesIO(response.content)).convert("RGB")
    else:
        image = Image.open(image_file).convert("RGB")
    return image


def eval_model(args):
    # Model
    disable_torch_init()
    model_name = os.path.expanduser(args.model_name)
    tokenizer = AutoTokenizer.from_pretrained(model_name)

    if "mpt" in model_name.lower():
        model = LlavaMPTForCausalLM.from_pretrained(
            model_name,
            low_cpu_mem_usage=True,
            torch_dtype=torch.float16,
            use_cache=True,
        ).cuda()
    else:
        # model = LlavaLlamaForCausalLM.from_pretrained(model_name, low_cpu_mem_usage=True, torch_dtype=torch.float16, use_cache=True).cuda()
        model = LlavaLlamaForCausalLM.from_pretrained(
            model_name, torch_dtype=torch.float16, device_map="auto"
        )  # .cuda()
    image_processor = CLIPImageProcessor.from_pretrained(
        model.config.mm_vision_tower, torch_dtype=torch.float16
    )

    mm_use_im_start_end = getattr(model.config, "mm_use_im_start_end", False)
    tokenizer.add_tokens([DEFAULT_IMAGE_PATCH_TOKEN], special_tokens=True)
    if mm_use_im_start_end:
        tokenizer.add_tokens(
            [DEFAULT_IM_START_TOKEN, DEFAULT_IM_END_TOKEN], special_tokens=True
        )

    vision_tower = model.get_model().vision_tower[0]
    if vision_tower.device.type == "meta":
        vision_tower = CLIPVisionModel.from_pretrained(
            vision_tower.config._name_or_path,
            torch_dtype=torch.float16,
            low_cpu_mem_usage=True,
        ).cuda()
        model.get_model().vision_tower[0] = vision_tower
    else:
        vision_tower.to(device="cuda", dtype=torch.float16)
    vision_config = vision_tower.config
    vision_config.im_patch_token = tokenizer.convert_tokens_to_ids(
        [DEFAULT_IMAGE_PATCH_TOKEN]
    )[0]
    vision_config.use_im_start_end = mm_use_im_start_end
    if mm_use_im_start_end:
        (
            vision_config.im_start_token,
            vision_config.im_end_token,
        ) = tokenizer.convert_tokens_to_ids(
            [DEFAULT_IM_START_TOKEN, DEFAULT_IM_END_TOKEN]
        )
    image_token_len = (vision_config.image_size // vision_config.patch_size) ** 2

    qs = args.query
    if mm_use_im_start_end:
        qs = (
            qs
            + "\n"
            + DEFAULT_IM_START_TOKEN
            + DEFAULT_IMAGE_PATCH_TOKEN * image_token_len
            + DEFAULT_IM_END_TOKEN
        )
    else:
        qs = qs + "\n" + DEFAULT_IMAGE_PATCH_TOKEN * image_token_len

    if "v1" in model_name.lower():
        conv_mode = "llava_v1"
    elif "mpt" in model_name.lower():
        conv_mode = "mpt_multimodal"
    else:
        conv_mode = "multimodal"

    if args.conv_mode is not None and conv_mode != args.conv_mode:
        print(
            "[WARNING] the auto inferred conversation mode is {}, while `--conv-mode` is {}, using {}".format(
                conv_mode, args.conv_mode, args.conv_mode
            )
        )
    else:
        args.conv_mode = conv_mode

    conv = conv_templates[args.conv_mode].copy()
    conv.append_message(conv.roles[0], qs)
    conv.append_message(conv.roles[1], None)
    prompt = conv.get_prompt()
    inputs = tokenizer([prompt])

    image = load_image(args.image_file)
    image_tensor = image_processor.preprocess(image, return_tensors="pt")[
        "pixel_values"
    ][0]

    input_ids = torch.as_tensor(inputs.input_ids).cuda()

    stop_str = conv.sep if conv.sep_style != SeparatorStyle.TWO else conv.sep2
    keywords = [stop_str]
    stopping_criteria = KeywordsStoppingCriteria(keywords, tokenizer, input_ids)

    with torch.inference_mode():
        output_ids = model.generate(
            input_ids,
            images=image_tensor.unsqueeze(0).half().cuda(),
            do_sample=True,
            temperature=0.2,
            max_new_tokens=1024,
            stopping_criteria=[stopping_criteria],
        )

    input_token_len = input_ids.shape[1]
    n_diff_input_output = (input_ids != output_ids[:, :input_token_len]).sum().item()
    if n_diff_input_output > 0:
        print(
            f"[Warning] {n_diff_input_output} output_ids are not the same as the input_ids"
        )
    outputs = tokenizer.batch_decode(
        output_ids[:, input_token_len:], skip_special_tokens=True
    )[0]
    outputs = outputs.strip()
    if outputs.endswith(stop_str):
        outputs = outputs[: -len(stop_str)]
    outputs = outputs.strip()
    print(outputs)

    import pdb

    pdb.set_trace()


if __name__ == "__main__":
    parser = argparse.ArgumentParser()
    parser.add_argument("--model-name", type=str, default="facebook/opt-350m")
    parser.add_argument("--image-file", type=str, required=True)
    parser.add_argument("--query", type=str, required=True)
    parser.add_argument("--conv-mode", type=str, default=None)
    args = parser.parse_args()

    eval_model(args)