Internvl3 1B Pretrained

🚀 InternVL3-1B預訓練模型

InternVL3-1B預訓練模型是先進的多模態大語言模型，具備強大的多模態感知和推理能力。相較於InternVL 2.5，它進一步拓展了多模態能力，涵蓋工具使用、GUI代理、工業圖像分析、3D視覺感知等領域。

📂 GitHub 📜 InternVL 1.0 📜 InternVL 1.5 📜 InternVL 2.5 📜 InternVL2.5-MPO 📜 InternVL3

🆕 Blog 🗨️ Chat Demo 🤗 HF Demo 🚀 快速開始 📖 文檔

🚀 快速開始

我們提供了使用transformers運行InternVL3-1B的示例代碼。

⚠️ 重要提示

請使用transformers>=4.37.2以確保模型正常工作。

模型加載

16位（bf16 / fp16）

import torch
from transformers import AutoTokenizer, AutoModel
path = "OpenGVLab/InternVL3-1B"
model = AutoModel.from_pretrained(
    path,
    torch_dtype=torch.bfloat16,
    low_cpu_mem_usage=True,
    use_flash_attn=True,
    trust_remote_code=True).eval().cuda()

BNB 8位量化

import torch
from transformers import AutoTokenizer, AutoModel
path = "OpenGVLab/InternVL3-1B"
model = AutoModel.from_pretrained(
    path,
    torch_dtype=torch.bfloat16,
    load_in_8bit=True,
    low_cpu_mem_usage=True,
    use_flash_attn=True,
    trust_remote_code=True).eval()

多GPU

以下代碼的編寫方式是為了避免在多GPU推理期間由於張量不在同一設備上而發生錯誤。通過確保大語言模型（LLM）的第一層和最後一層在同一設備上，我們可以防止此類錯誤。

import math
import torch
from transformers import AutoTokenizer, AutoModel

def split_model(model_name):
    device_map = {}
    world_size = torch.cuda.device_count()
    config = AutoConfig.from_pretrained(model_path, trust_remote_code=True)
    num_layers = config.llm_config.num_hidden_layers
    # Since the first GPU will be used for ViT, treat it as half a GPU.
    num_layers_per_gpu = math.ceil(num_layers / (world_size - 0.5))
    num_layers_per_gpu = [num_layers_per_gpu] * world_size
    num_layers_per_gpu[0] = math.ceil(num_layers_per_gpu[0] * 0.5)
    layer_cnt = 0
    for i, num_layer in enumerate(num_layers_per_gpu):
        for j in range(num_layer):
            device_map[f'language_model.model.layers.{layer_cnt}'] = i
            layer_cnt += 1
    device_map['vision_model'] = 0
    device_map['mlp1'] = 0
    device_map['language_model.model.tok_embeddings'] = 0
    device_map['language_model.model.embed_tokens'] = 0
    device_map['language_model.output'] = 0
    device_map['language_model.model.norm'] = 0
    device_map['language_model.model.rotary_emb'] = 0
    device_map['language_model.lm_head'] = 0
    device_map[f'language_model.model.layers.{num_layers - 1}'] = 0

    return device_map

path = "OpenGVLab/InternVL3-1B"
device_map = split_model('InternVL3-1B')
model = AutoModel.from_pretrained(
    path,
    torch_dtype=torch.bfloat16,
    low_cpu_mem_usage=True,
    use_flash_attn=True,
    trust_remote_code=True,
    device_map=device_map).eval()

使用Transformers進行推理

import math
import numpy as np
import torch
import torchvision.transforms as T
from decord import VideoReader, cpu
from PIL import Image
from torchvision.transforms.functional import InterpolationMode
from transformers import AutoModel, AutoTokenizer

IMAGENET_MEAN = (0.485, 0.456, 0.406)
IMAGENET_STD = (0.229, 0.224, 0.225)

def build_transform(input_size):
    MEAN, STD = IMAGENET_MEAN, IMAGENET_STD
    transform = T.Compose([
        T.Lambda(lambda img: img.convert('RGB') if img.mode != 'RGB' else img),
        T.Resize((input_size, input_size), interpolation=InterpolationMode.BICUBIC),
        T.ToTensor(),
        T.Normalize(mean=MEAN, std=STD)
    ])
    return transform

def find_closest_aspect_ratio(aspect_ratio, target_ratios, width, height, image_size):
    best_ratio_diff = float('inf')
    best_ratio = (1, 1)
    area = width * height
    for ratio in target_ratios:
        target_aspect_ratio = ratio[0] / ratio[1]
        ratio_diff = abs(aspect_ratio - target_aspect_ratio)
        if ratio_diff < best_ratio_diff:
            best_ratio_diff = ratio_diff
            best_ratio = ratio
        elif ratio_diff == best_ratio_diff:
            if area > 0.5 * image_size * image_size * ratio[0] * ratio[1]:
                best_ratio = ratio
    return best_ratio

def dynamic_preprocess(image, min_num=1, max_num=12, image_size=448, use_thumbnail=False):
    orig_width, orig_height = image.size
    aspect_ratio = orig_width / orig_height

    # calculate the existing image aspect ratio
    target_ratios = set(
        (i, j) for n in range(min_num, max_num + 1) for i in range(1, n + 1) for j in range(1, n + 1) if
        i * j <= max_num and i * j >= min_num)
    target_ratios = sorted(target_ratios, key=lambda x: x[0] * x[1])

    # find the closest aspect ratio to the target
    target_aspect_ratio = find_closest_aspect_ratio(
        aspect_ratio, target_ratios, orig_width, orig_height, image_size)

    # calculate the target width and height
    target_width = image_size * target_aspect_ratio[0]
    target_height = image_size * target_aspect_ratio[1]
    blocks = target_aspect_ratio[0] * target_aspect_ratio[1]

    # resize the image
    resized_img = image.resize((target_width, target_height))
    processed_images = []
    for i in range(blocks):
        box = (
            (i % (target_width // image_size)) * image_size,
            (i // (target_width // image_size)) * image_size,
            ((i % (target_width // image_size)) + 1) * image_size,
            ((i // (target_width // image_size)) + 1) * image_size
        )
        # split the image
        split_img = resized_img.crop(box)
        processed_images.append(split_img)
    assert len(processed_images) == blocks
    if use_thumbnail and len(processed_images) != 1:
        thumbnail_img = image.resize((image_size, image_size))
        processed_images.append(thumbnail_img)
    return processed_images

def load_image(image_file, input_size=448, max_num=12):
    image = Image.open(image_file).convert('RGB')
    transform = build_transform(input_size=input_size)
    images = dynamic_preprocess(image, image_size=input_size, use_thumbnail=True, max_num=max_num)
    pixel_values = [transform(image) for image in images]
    pixel_values = torch.stack(pixel_values)
    return pixel_values

def split_model(model_name):
    device_map = {}
    world_size = torch.cuda.device_count()
    config = AutoConfig.from_pretrained(model_path, trust_remote_code=True)
    num_layers = config.llm_config.num_hidden_layers
    # Since the first GPU will be used for ViT, treat it as half a GPU.
    num_layers_per_gpu = math.ceil(num_layers / (world_size - 0.5))
    num_layers_per_gpu = [num_layers_per_gpu] * world_size
    num_layers_per_gpu[0] = math.ceil(num_layers_per_gpu[0] * 0.5)
    layer_cnt = 0
    for i, num_layer in enumerate(num_layers_per_gpu):
        for j in range(num_layer):
            device_map[f'language_model.model.layers.{layer_cnt}'] = i
            layer_cnt += 1
    device_map['vision_model'] = 0
    device_map['mlp1'] = 0
    device_map['language_model.model.tok_embeddings'] = 0
    device_map['language_model.model.embed_tokens'] = 0
    device_map['language_model.output'] = 0
    device_map['language_model.model.norm'] = 0
    device_map['language_model.model.rotary_emb'] = 0
    device_map['language_model.lm_head'] = 0
    device_map[f'language_model.model.layers.{num_layers - 1}'] = 0

    return device_map

# If you set `load_in_8bit=True`, you will need two 80GB GPUs.
# If you set `load_in_8bit=False`, you will need at least three 80GB GPUs.
path = 'OpenGVLab/InternVL3-1B'
device_map = split_model('InternVL3-1B')
model = AutoModel.from_pretrained(
    path,
    torch_dtype=torch.bfloat16,
    load_in_8bit=False,
    low_cpu_mem_usage=True,
    use_flash_attn=True,
    trust_remote_code=True,
    device_map=device_map).eval()
tokenizer = AutoTokenizer.from_pretrained(path, trust_remote_code=True, use_fast=False)

# set the max number of tiles in `max_num`
pixel_values = load_image('./examples/image1.jpg', max_num=12).to(torch.bfloat16).cuda()
generation_config = dict(max_new_tokens=1024, do_sample=True)

# pure-text conversation (純文本對話)
question = 'Hello, who are you?'
response, history = model.chat(tokenizer, None, question, generation_config, history=None, return_history=True)
print(f'User: {question}\nAssistant: {response}')

question = 'Can you tell me a story?'
response, history = model.chat(tokenizer, None, question, generation_config, history=history, return_history=True)
print(f'User: {question}\nAssistant: {response}')

# single-image single-round conversation (單圖單輪對話)
question = '<image>\nPlease describe the image shortly.'
response = model.chat(tokenizer, pixel_values, question, generation_config)
print(f'User: {question}\nAssistant: {response}')

# single-image multi-round conversation (單圖多輪對話)
question = '<image>\nPlease describe the image in detail.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config, history=None, return_history=True)
print(f'User: {question}\nAssistant: {response}')

question = 'Please write a poem according to the image.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config, history=history, return_history=True)
print(f'User: {question}\nAssistant: {response}')

# multi-image multi-round conversation, combined images (多圖多輪對話，拼接圖像)
pixel_values1 = load_image('./examples/image1.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values2 = load_image('./examples/image2.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values = torch.cat((pixel_values1, pixel_values2), dim=0)

question = '<image>\nDescribe the two images in detail.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
                               history=None, return_history=True)
print(f'User: {question}\nAssistant: {response}')

question = 'What are the similarities and differences between these two images.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
                               history=history, return_history=True)
print(f'User: {question}\nAssistant: {response}')

# multi-image multi-round conversation, separate images (多圖多輪對話，獨立圖像)
pixel_values1 = load_image('./examples/image1.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values2 = load_image('./examples/image2.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values = torch.cat((pixel_values1, pixel_values2), dim=0)
num_patches_list = [pixel_values1.size(0), pixel_values2.size(0)]

question = 'Image-1: <image>\nImage-2: <image>\nDescribe the two images in detail.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
                               num_patches_list=num_patches_list,
                               history=None, return_history=True)
print(f'User: {question}\nAssistant: {response}')

question = 'What are the similarities and differences between these two images.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
                               num_patches_list=num_patches_list,
                               history=history, return_history=True)
print(f'User: {question}\nAssistant: {response}')

# batch inference, single image per sample (單圖批處理)
pixel_values1 = load_image('./examples/image1.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values2 = load_image('./examples/image2.jpg', max_num=12).to(torch.bfloat16).cuda()
num_patches_list = [pixel_values1.size(0), pixel_values2.size(0)]
pixel_values = torch.cat((pixel_values1, pixel_values2), dim=0)

questions = ['<image>\nDescribe the image in detail.'] * len(num_patches_list)
responses = model.batch_chat(tokenizer, pixel_values,
                             num_patches_list=num_patches_list,
                             questions=questions,
                             generation_config=generation_config)
for question, response in zip(questions, responses):
    print(f'User: {question}\nAssistant: {response}')

# video multi-round conversation (視頻多輪對話)
def get_index(bound, fps, max_frame, first_idx=0, num_segments=32):
    if bound:
        start, end = bound[0], bound[1]
    else:
        start, end = -100000, 100000
    start_idx = max(first_idx, round(start * fps))
    end_idx = min(round(end * fps), max_frame)
    seg_size = float(end_idx - start_idx) / num_segments
    frame_indices = np.array([
        int(start_idx + (seg_size / 2) + np.round(seg_size * idx))
        for idx in range(num_segments)
    ])
    return frame_indices

def load_video(video_path, bound=None, input_size=448, max_num=1, num_segments=32):
    vr = VideoReader(video_path, ctx=cpu(0), num_threads=1)
    max_frame = len(vr) - 1
    fps = float(vr.get_avg_fps())

    pixel_values_list, num_patches_list = [], []
    transform = build_transform(input_size=input_size)
    frame_indices = get_index(bound, fps, max_frame, first_idx=0, num_segments=num_segments)
    for frame_index in frame_indices:
        img = Image.fromarray(vr[frame_index].asnumpy()).convert('RGB')
        img = dynamic_preprocess(img, image_size=input_size, use_thumbnail=True, max_num=max_num)
        pixel_values = [transform(tile) for tile in img]
        pixel_values = torch.stack(pixel_values)
        num_patches_list.append(pixel_values.shape[0])
        pixel_values_list.append(pixel_values)
    pixel_values = torch.cat(pixel_values_list)
    return pixel_values, num_patches_list

video_path = './examples/red-panda.mp4'
pixel_values, num_patches_list = load_video(video_path, num_segments=8, max_num=1)
pixel_values = pixel_values.to(torch.bfloat16).cuda()
video_prefix = ''.join([f'Frame{i+1}: <image>\n' for i in range(len(num_patches_list))])
question = video_prefix + 'What is the red panda doing?'
# Frame1: <image>\nFrame2: <image>\n...\nFrame8: <image>\n{question}
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
                               num_patches_list=num_patches_list, history=None, return_history=True)
print(f'User: {question}\nAssistant: {response}')

question = 'Describe this video in detail.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
                               num_patches_list=num_patches_list, history=history, return_history=True)
print(f'User: {question}\nAssistant: {response}')

流式輸出

除了上述方法，你還可以使用以下代碼獲取流式輸出。

from transformers import TextIteratorStreamer
from threading import Thread

# Initialize the streamer
streamer = TextIteratorStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True, timeout=10)
# Define the generation configuration
generation_config = dict(max_new_tokens=1024, do_sample=False, streamer=streamer)
# Start the model chat in a separate thread
thread = Thread(target=model.chat, kwargs=dict(
    tokenizer=tokenizer, pixel_values=pixel_values, question=question,
    history=None, return_history=False, generation_config=generation_config,
))
thread.start()

# Initialize an empty string to store the generated text
generated_text = ''
# Loop through the streamer to get the new text as it is generated
for new_text in streamer:
    if new_text == model.conv_template.sep:
        break
    generated_text += new_text
    print(new_text, end='', flush=True)  # Print each new chunk of generated text on the same line

✨ 主要特性

• 先進的多模態能力：與InternVL 2.5相比，InternVL3展現出更卓越的多模態感知和推理能力，其多模態能力進一步拓展到工具使用、GUI代理、工業圖像分析、3D視覺感知等領域。
• 原生多模態預訓練：提出的原生多模態預訓練方法將語言和視覺學習整合到一個預訓練階段，使模型能夠同時學習語言和多模態表示，增強處理視覺 - 語言任務的能力。
• 多種訓練策略：採用監督微調、混合偏好優化和測試時縮放等訓練策略，提升模型的性能和推理能力。
• 支持多種數據類型：從InternVL 2.0開始，除了支持圖像，還額外引入了對多圖像和視頻數據的支持。
• 更好的長上下文理解能力：在InternVL3中集成了可變視覺位置編碼（V2PE），使模型在長上下文理解方面表現更出色。

📦 安裝指南

LMDeploy

LMDeploy是一個用於壓縮、部署和服務大語言模型（LLMs）和視覺語言模型（VLMs）的工具包。

# 如果lmdeploy<0.7.3，你需要顯式設置chat_template_config=ChatTemplateConfig(model_name='internvl2_5')
pip install lmdeploy>=0.7.3

💻 使用示例

基礎用法

from lmdeploy import pipeline, TurbomindEngineConfig, ChatTemplateConfig
from lmdeploy.vl import load_image

model = 'OpenGVLab/InternVL3-1B'
image = load_image('https://raw.githubusercontent.com/open-mmlab/mmdeploy/main/tests/data/tiger.jpeg')
pipe = pipeline(model, backend_config=TurbomindEngineConfig(session_len=16384, tp=1), chat_template_config=ChatTemplateConfig(model_name='internvl2_5'))
response = pipe(('describe this image', image))
print(response.text)

高級用法

多圖像推理

from lmdeploy import pipeline, TurbomindEngineConfig, ChatTemplateConfig
from lmdeploy.vl import load_image
from lmdeploy.vl.constants import IMAGE_TOKEN

model = 'OpenGVLab/InternVL3-1B'
pipe = pipeline(model, backend_config=TurbomindEngineConfig(session_len=16384, tp=1), chat_template_config=ChatTemplateConfig(model_name='internvl2_5'))

image_urls=[
    'https://raw.githubusercontent.com/open-mmlab/mmdeploy/main/demo/resources/human-pose.jpg',
    'https://raw.githubusercontent.com/open-mmlab/mmdeploy/main/demo/resources/det.jpg'
]

images = [load_image(img_url) for img_url in image_urls]
# 對圖像進行編號有助於多圖像對話
response = pipe((f'Image-1: {IMAGE_TOKEN}\nImage-2: {IMAGE_TOKEN}\ndescribe these two images', images))
print(response.text)

批量提示推理

from lmdeploy import pipeline, TurbomindEngineConfig, ChatTemplateConfig
from lmdeploy.vl import load_image

model = 'OpenGVLab/InternVL3-1B'
pipe = pipeline(model, backend_config=TurbomindEngineConfig(session_len=16384, tp=1), chat_template_config=ChatTemplateConfig(model_name='internvl2_5'))

image_urls=[
    "https://raw.githubusercontent.com/open-mmlab/mmdeploy/main/demo/resources/human-pose.jpg",
    "https://raw.githubusercontent.com/open-mmlab/mmdeploy/main/demo/resources/det.jpg"
]
prompts = [('describe this image', load_image(img_url)) for img_url in image_urls]
response = pipe(prompts)
print(response)

多輪對話

from lmdeploy import pipeline, TurbomindEngineConfig, GenerationConfig, ChatTemplateConfig
from lmdeploy.vl import load_image

model = 'OpenGVLab/InternVL3-1B'
pipe = pipeline(model, backend_config=TurbomindEngineConfig(session_len=16384, tp=1), chat_template_config=ChatTemplateConfig(model_name='internvl2_5'))

image = load_image('https://raw.githubusercontent.com/open-mmlab/mmdeploy/main/demo/resources/human-pose.jpg')
gen_config = GenerationConfig(top_k=40, top_p=0.8, temperature=0.8)
sess = pipe.chat(('describe this image', image), gen_config=gen_config)
print(sess.response.text)
sess = pipe.chat('What is the woman doing?', session=sess, gen_config=gen_config)
print(sess.response.text)

服務部署

lmdeploy serve api_server OpenGVLab/InternVL3-1B --chat-template internvl2_5 --server-port 23333 --tp 1

要使用OpenAI風格的接口，你需要安裝openai：

pip install openai

然後，使用以下代碼進行API調用：

from openai import OpenAI

client = OpenAI(api_key='YOUR_API_KEY', base_url='http://0.0.0.0:23333/v1')
model_name = client.models.list().data[0].id
response = client.chat.completions.create(
    model=model_name,
    messages=[{
        'role':
        'user',
        'content': [{
            'type': 'text',
            'text': 'describe this image',
        }, {
            'type': 'image_url',
            'image_url': {
                'url':
                'https://modelscope.oss-cn-beijing.aliyuncs.com/resource/tiger.jpeg',
            },
        }],
    }],
    temperature=0.8,
    top_p=0.8)
print(response)

📚 詳細文檔

InternVL3家族

以下表格概述了InternVL3系列模型。

模型名稱	視覺部分	語言部分	Hugging Face鏈接
InternVL3-1B	InternViT-300M-448px-V2_5	Qwen2.5-0.5B	🤗 鏈接
InternVL3-2B	InternViT-300M-448px-V2_5	Qwen2.5-1.5B	🤗 鏈接
InternVL3-8B	InternViT-300M-448px-V2_5	Qwen2.5-7B	🤗 鏈接
InternVL3-9B	InternViT-300M-448px-V2_5	internlm3-8b-instruct	🤗 鏈接
InternVL3-14B	InternViT-300M-448px-V2_5	Qwen2.5-14B	🤗 鏈接
InternVL3-38B	InternViT-6B-448px-V2_5	Qwen2.5-32B	🤗 鏈接
InternVL3-78B	InternViT-6B-448px-V2_5	Qwen2.5-72B	🤗 鏈接

模型架構

如下圖所示，InternVL3 保留了與 InternVL 2.5 及其前身InternVL 1.5和2.0相同的模型架構，遵循“ViT - MLP - LLM”範式。在這個新版本中，我們使用隨機初始化的MLP投影器，將新的增量預訓練的InternViT與各種預訓練的大語言模型（LLMs）集成在一起，包括InternLM 3和Qwen 2.5。

與之前的版本一樣，我們應用了像素無序操作，將視覺標記的數量減少到原來的四分之一。此外，我們採用了與InternVL 1.5類似的動態分辨率策略，將圖像劃分為448×448像素的圖塊。從InternVL 2.0開始的主要區別在於，我們額外引入了對多圖像和視頻數據的支持。

值得注意的是，在InternVL3中，我們集成了可變視覺位置編碼（V2PE），它為視覺標記使用更小、更靈活的位置增量。得益於V2PE，InternVL3與之前的版本相比，展現出更好的長上下文理解能力。

訓練策略

原生多模態預訓練

我們提出了一種原生多模態預訓練方法，將語言和視覺學習整合到一個預訓練階段。與先訓練純語言模型，然後再將其適應處理其他模態的標準範式不同，我們的方法將多模態數據（如圖像 - 文本、視頻 - 文本或圖像 - 文本交錯序列）與大規模文本語料庫交織在一起。這種統一的訓練方案使模型能夠同時學習語言和多模態表示，最終增強其處理視覺 - 語言任務的能力，而無需單獨的對齊或橋接模塊。更多詳細信息請參閱我們的論文。

監督微調

在這個階段，InternVL2.5中提出的隨機JPEG壓縮、平方損失重新加權和多模態數據打包技術也在InternVL3系列中得到應用。與InternVL2.5相比，InternVL3的監督微調階段的主要進步在於使用了更高質量和更多樣化的訓練數據。具體來說，我們進一步擴展了工具使用、3D場景理解、GUI操作、長上下文任務、視頻理解、科學圖表、創意寫作和多模態推理的訓練樣本。

混合偏好優化

在預訓練和監督微調期間，模型根據之前的真實標記來預測下一個標記。然而，在推理期間，模型根據其自身的先前輸出預測每個標記。這種真實標記和模型預測標記之間的差異引入了分佈偏移，這可能會削弱模型的思維鏈（CoT）推理能力。為了緩解這個問題，我們採用了MPO，它引入了來自正樣本和負樣本的額外監督，使模型響應分佈與真實分佈對齊，從而提高推理性能。具體來說，MPO的訓練目標是偏好損失 \(\mathcal{L}{\text{p}}\)、質量損失 \(\mathcal{L}{\text{q}}\) 和生成損失 \(\mathcal{L}{\text{g}}\) 的組合，可以表述為： $$ \mathcal{L}=w{p}\cdot\mathcal{L}{\text{p}} + w{q}\cdot\mathcal{L}{\text{q}} + w{g}\cdot\mathcal{L}{\text{g}}, $$ 其中 \(w{*}\) 表示分配給每個損失組件的權重。有關MPO的更多詳細信息，請參閱我們的論文。

測試時縮放

測試時縮放已被證明是增強大語言模型（LLMs）和多模態大語言模型（MLLMs）推理能力的有效方法。在這項工作中，我們使用Best - of - N評估策略，並採用VisualPRM - 8B作為評判模型，為推理和數學評估選擇最佳響應。

多模態能力評估

多模態推理和數學

OCR、圖表和文檔理解

多圖像和真實世界理解

綜合多模態和幻覺評估

視覺定位

多模態多語言理解

視頻理解

GUI定位

空間推理

語言能力評估

我們將InternVL3與Qwen2.5聊天模型進行了比較，Qwen2.5的相應預訓練基礎模型被用作InternVL3中語言組件的初始化。得益於原生多模態預訓練，InternVL3系列在整體文本性能上甚至比Qwen2.5系列更好。請注意，Qwen2.5系列的評估分數可能與官方報告的分數不同，因為我們在所有數據集上都採用了表中提供的提示版本進行OpenCompass評估。

消融研究

原生多模態預訓練

我們在InternVL2 - 8B模型上進行了實驗，同時保持其架構、初始化參數和訓練數據完全不變。傳統上，InternVL2 - 8B採用的訓練管道是先進行MLP預熱階段以進行特徵對齊，然後進行指令微調階段。在我們的實驗中，我們用原生多模態預訓練過程取代了傳統的MLP預熱階段。這種修改隔離了原生多模態預訓練對模型整體多模態能力的貢獻。

下圖中的評估結果表明，經過原生多模態預訓練的模型在大多數基準測試中的性能與經過完整多階段訓練的InternVL2 - 8B基線相當。此外，當在更高質量的數據上進行指令微調後，該模型在評估的多模態任務中表現出進一步的性能提升。這些發現強調了原生多模態預訓練在賦予多模態大語言模型強大多模態能力方面的效率。

混合偏好優化

如下表所示，與未使用MPO的模型相比，使用MPO進行微調的模型在七個多模態推理基準測試中表現出更出色的推理性能。具體來說，InternVL3 - 78B和InternVL3 - 38B分別比其對應模型高出4.1和4.5分。值得注意的是，用於MPO的訓練數據是用於監督微調的訓練數據的子集，這表明性能提升主要源於訓練算法而非訓練數據。

可變視覺位置編碼

如下表所示，引入V2PE在大多數評估指標上帶來了顯著的性能提升。此外，我們通過改變位置增量 \( \delta \) 進行的消融研究表明，即使對於主要涉及常規上下文的任務，相對較小的 \( \delta \) 值也能實現最佳性能。這些發現為未來改進多模態大語言模型中視覺標記的位置編碼策略提供了重要見解。

🔧 技術細節

模型架構細節

• 範式遵循：遵循“ViT - MLP - LLM”範式，將視覺特徵提取、特徵映射和語言處理模塊有機結合。
• 視覺處理：應用像素無序操作減少視覺標記數量，採用動態分辨率策略將圖像劃分為448×448像素的圖塊，從InternVL 2.0開始支持多圖像和視頻數據。
• 位置編碼：集成可變視覺位置編碼（V2PE），為視覺標記提供更靈活的位置信息，增強長上下文理解能力。

訓練策略細節

• 原生多模態預訓練：將語言和視覺學習整合到一個預訓練階段，通過交織多模態數據和大規模文本語料庫，使模型同時學習語言和多模態表示。
• 監督微調：採用隨機JPEG壓縮、平方損失重新加權和多模態數據打包技術，使用更高質量和多樣化的訓練數據，擴展多個領域的訓練樣本。
• 混合偏好優化：引入正樣本和負樣本的額外監督，使模型響應分佈與真實分佈對齊，緩解推理期間的分佈偏移問題。
• 測試時縮放：使用Best - of - N評估策略和VisualPRM - 8B作為評判模型，提高推理和數學評估的性能。

📄 許可證

本項目基於MIT許可證發佈。本項目使用了預訓練的Qwen2.5作為組件，Qwen2.5遵循Apache - 2.0許可證。

引用

如果您在研究中發現本項目有用，請考慮引用以下文獻：

@article{chen2024expanding,
  title={Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling},
  author={Chen, Zhe and Wang, Weiyun and Cao, Yue and Liu, Yangzhou and Gao, Zhangwei and Cui, Erfei and Zhu, Jinguo and Ye, Shenglong and Tian, Hao and Liu, Zhaoyang and others},
  journal={arXiv preprint arXiv:2412.05271},
  year={2024}
}
@article{wang2024mpo,
  title={Enhancing the Reasoning Ability of Multimodal Large Language Models via Mixed Preference Optimization},
  author={Wang, Weiyun and Chen, Zhe and Wang, Wenhai and Cao, Yue and Liu, Yangzhou and Gao, Zhangwei and Zhu, Jinguo and Zhu, Xizhou and Lu, Lewei and Qiao, Yu and Dai, Jifeng},
  journal={arXiv preprint arXiv:2411.10442},
  year={2024}
}
@article{chen2024far,
  title={How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites},
  author={Chen, Zhe and Wang, Weiyun and Tian, Hao and Ye, Shenglong and Gao, Zhangwei and Cui, Erfei and Tong, Wenwen and Hu, Kongzhi and Luo, Jiapeng and Ma, Zheng and others},
  journal={arXiv preprint arXiv:2404.16821},
  year={2024}
}
@inproceedings{chen2024internvl,
  title={Internvl: Scaling up vision foundation models and aligning for generic visual-linguistic tasks},
  author={Chen, Zhe and Wu, Jiannan and Wang, Wenhai and Su, Weijie and Chen, Guo and Xing, Sen and Zhong, Muyan and Zhang, Qinglong and Zhu, Xizhou and Lu, Lewei and others},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  pages={24185--24198},
  year={2024}
}