Convllava JP 1.3b 1280

🚀 ConvLLaVA-JP Model Card

ConvLLaVA-JP is a vision - language model designed to engage in conversations about input images, offering high - resolution input support.

🚀 Quick Start

To use the ConvLLaVA - JP model, you need to follow the steps below:

1. Download the necessary dependencies.
2. Perform inference using the provided code example.

📦 Installation

git clone https://github.com/tosiyuki/LLaVA-JP.git

💻 Usage Examples

Basic Usage

import requests
import torch
import transformers
from PIL import Image

from transformers.generation.streamers import TextStreamer
from llava.constants import DEFAULT_IMAGE_TOKEN, IMAGE_TOKEN_INDEX
from llava.conversation import conv_templates, SeparatorStyle
from llava.model.llava_gpt2 import LlavaGpt2ForCausalLM
from llava.train.dataset import tokenizer_image_token


if __name__ == "__main__":
    model_path = 'toshi456/ConvLLaVA-JP-1.3b-1280'
    device = "cuda" if torch.cuda.is_available() else "cpu"
    torch_dtype = torch.bfloat16 if device=="cuda" else torch.float32

    model = LlavaGpt2ForCausalLM.from_pretrained(
        model_path, 
        low_cpu_mem_usage=True,
        use_safetensors=True,
        torch_dtype=torch_dtype,
        device_map=device,
    )
    tokenizer = transformers.AutoTokenizer.from_pretrained(
        model_path,
        model_max_length=1532,
        padding_side="right",
        use_fast=False,
    )
    model.eval()

    conv_mode = "v1"
    conv = conv_templates[conv_mode].copy()

    # image pre-process
    image_url = "https://huggingface.co/rinna/bilingual-gpt-neox-4b-minigpt4/resolve/main/sample.jpg"
    image = Image.open(requests.get(image_url, stream=True).raw).convert('RGB')
    
    if device == "cuda":
        image_tensor = model.get_model().vision_tower.image_processor(image).unsqueeze(0).half().cuda().to(torch_dtype)
    else:
        image_tensor = model.get_model().vision_tower.image_processor(image).unsqueeze(0).to(torch_dtype)

    # create prompt
    # ユーザー: <image>\n{prompt}
    prompt = "猫の隣には何がありますか？"
    inp = DEFAULT_IMAGE_TOKEN + '\n' + prompt
    conv.append_message(conv.roles[0], inp)
    conv.append_message(conv.roles[1], None)
    prompt = conv.get_prompt()

    input_ids = tokenizer_image_token(
        prompt, 
        tokenizer, 
        IMAGE_TOKEN_INDEX, 
        return_tensors='pt'
    ).unsqueeze(0)
    if device == "cuda":
        input_ids = input_ids.to(device)

    input_ids = input_ids[:, :-1] # </sep>がinputの最後に入るので削除する
    stop_str = conv.sep if conv.sep_style != SeparatorStyle.TWO else conv.sep2
    keywords = [stop_str]
    streamer = TextStreamer(tokenizer, skip_prompt=True, timeout=20.0)

    # predict
    with torch.inference_mode():
        output_id = model.generate(
            inputs=input_ids,
            images=image_tensor,
            do_sample=False,
            temperature=1.0,
            top_p=1.0,
            max_new_tokens=256,
            streamer=streamer,
            use_cache=True,
        )
    """猫の隣にはノートパソコンがあります。"""

📚 Documentation

Model Details

Model Type: ConvLLaVA - JP is a vision - language model capable of conversing about input images. This LVLM model uses [laion/CLIP - convnext_large_d_320.laion2B - s29B - b131K - ft](https://huggingface.co/laion/CLIP - convnext_large_d_320.laion2B - s29B - b131K - ft) as the image encoder and [llm - jp/llm - jp - 1.3b - v1.0](https://huggingface.co/llm - jp/llm - jp - 1.3b - v1.0) as the text decoder, supporting an input resolution of 1280 x 1280.

Training: This model was initially trained with the Vision Projector and Stage 5 using LLaVA - Pretrain - JA. In the second phase, it was trained on the Image Encoder, Vision Projector, Stage 5, and LLM using LLaVA - Pretrain - JA. In the third phase, it was fine - tuned with the Vision Projector and LLM using LLaVA - v1.5 - Instruct - 620K - JA.

For more information, refer to: https://github.com/tosiyuki/LLaVA - JP/tree/main

Comparing VLMs

Training Dataset

Stage 1 and Stage 2 Pretrain:

• [LLaVA - Pretrain - JA](https://huggingface.co/datasets/turing - motors/LLaVA - Pretrain - JA)

Stage 3 Fine - tuning:

• [LLaVA - v1.5 - Instruct - 620K - JA](https://huggingface.co/datasets/turing - motors/LLaVA - v1.5 - Instruct - 620K - JA)

Acknowledgement

• ConvLLaVA
• [LLM - jp](https://llm - jp.nii.ac.jp/)
• Open CLIP

📄 License

This project is licensed under the CC - BY - NC - 4.0 license.