模型概述
模型特點
模型能力
使用案例
🚀 InternVL3-38B-Instruct
InternVL3-38B-Instruct 是一個先進的多模態大語言模型,在多模態感知、推理等能力上表現出色,拓展了多模態能力至工具使用、GUI 代理等多個領域,且在文本性能上也有優秀表現。
📂 GitHub 📜 InternVL 1.0 📜 InternVL 1.5 📜 InternVL 2.5 📜 InternVL2.5-MPO 📜 InternVL3
🆕 Blog 🗨️ Chat Demo 🤗 HF Demo 🚀 Quick Start 📖 Documents

🚀 快速開始
我們提供了使用 transformers
運行 InternVL3-38B
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/InternVL3-38B"
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/InternVL3-38B"
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
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/InternVL3-38B"
device_map = split_model('InternVL3-38B')
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/InternVL3-38B'
device_map = split_model('InternVL3-38B')
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
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
這是 InternVL3-38B 的 SFT 版本,經過了原生多模態預訓練和 SFT,但未經過 MPO。與 InternVL 2.5 相比,InternVL3 展現出更出色的多模態感知和推理能力,並將多模態能力進一步擴展到工具使用、GUI 代理、工業圖像分析、3D 視覺感知等領域。此外,得益於原生多模態預訓練,InternVL3 系列在整體文本性能上甚至優於 Qwen2.5 系列。
📦 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
📚 詳細文檔
InternVL3 家族
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 | 🤗 link |
InternVL3-2B | InternViT-300M-448px-V2_5 | Qwen2.5-1.5B | 🤗 link |
InternVL3-8B | InternViT-300M-448px-V2_5 | Qwen2.5-7B | 🤗 link |
InternVL3-9B | InternViT-300M-448px-V2_5 | internlm3-8b-instruct | 🤗 link |
InternVL3-14B | InternViT-300M-448px-V2_5 | Qwen2.5-14B | 🤗 link |
InternVL3-38B | InternViT-6B-448px-V2_5 | Qwen2.5-32B | 🤗 link |
InternVL3-78B | InternViT-6B-448px-V2_5 | Qwen2.5-72B | 🤗 link |
模型架構
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,包括 InternLM 3 和 Qwen 2.5。
與之前的版本一樣,我們應用了像素重排操作,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,將圖像分割成 448×448 像素的圖塊。從 InternVL 2.0 開始,關鍵的區別在於我們還增加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 與其前身相比,表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。與先訓練僅語言模型,然後將其適應於處理其他模態的標準範式不同,我們的方法將多模態數據(如圖文、視頻文本或圖文交錯序列)與大規模文本語料庫交織在一起。這種統一的訓練方案允許模型同時學習語言和多模態表示,最終提高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。更多細節請參閱 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被應用於 InternVL3 系列。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,模型根據自己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,這可能會損害模型的思維鏈(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,以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 \(w{*}\) 表示每個損失組件的權重。更多關於 MPO 的細節請參閱 我們的論文。
測試時縮放
測試時縮放已被證明是提高大語言模型(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,並使用 VisualPRM-8B 作為評估模型,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優於 Qwen2.5 系列。請注意,Qwen2.5 系列的評估分數可能與官方報告的不同,因為我們在所有數據集上採用了表中提供的提示版本進行 OpenCompass 評估。
消融實驗
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 採用的訓練流程是先進行 MLP 預熱階段進行特徵對齊,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 進行微調的模型相比,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 導致大多數評估指標的性能顯著提升。此外,我們的消融實驗(通過改變位置增量 \( \delta \))表明,即使對於主要涉及傳統上下文的任務,相對較小的 \( \delta \) 值也可以實現最佳性能。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。
微調
許多倉庫現在支持對 InternVL 系列模型進行微調,包括 InternVL、SWIFT、XTurner 等。有關微調的更多詳細信息,請參考它們的文檔。
部署
LMDeploy
LMDeploy 抽象了多模態視覺語言模型(VLM)複雜的推理過程,將其轉化為一個易於使用的管道,類似於大語言模型(LLM)的推理管道。
“Hello, world” 示例
from lmdeploy import pipeline, TurbomindEngineConfig, ChatTemplateConfig
from lmdeploy.vl import load_image
model = 'OpenGVLab/InternVL3-38B'
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=2), chat_template_config=ChatTemplateConfig(model_name='internvl2_5'))
response = pipe(('describe this image', image))
print(response.text)
如果在執行此示例時出現 ImportError
,請按照提示安裝所需的依賴包。
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/InternVL3-38B'
pipe = pipeline(model, backend_config=TurbomindEngineConfig(session_len=16384, tp=2), 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]
# Numbering images improves multi-image conversations
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/InternVL3-38B'
pipe = pipeline(model, backend_config=TurbomindEngineConfig(session_len=16384, tp=2), 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/InternVL3-38B'
pipe = pipeline(model, backend_config=TurbomindEngineConfig(session_len=16384, tp=2), 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
lmdeploy serve api_server OpenGVLab/InternVL3-38B --chat-template internvl2_5 --server-port 23333 --tp 2
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)
🔧 技術細節
模型架構
InternVL3 保留了與 InternVL 2.5 及其前身相同的 “ViT-MLP-LLM” 範式架構。在新版本中,使用隨機初始化的 MLP 投影器將新的增量預訓練的 InternViT 與多種預訓練的 LLMs 集成。同時,應用像素重排操作減少視覺標記數量,採用動態分辨率策略分割圖像,並從 InternVL 2.0 開始增加對多圖像和視頻數據的支持。此外,集成了可變視覺位置編碼(V2PE),提高了長上下文理解能力。
訓練策略
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,提高處理視覺語言任務的能力。
監督微調
在 InternVL3 中使用了更高質量和更多樣化的訓練數據,擴展了多個領域的訓練樣本,提升了模型在多模態任務上的性能。
混合偏好優化
引入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,緩解推理期間的分佈偏移問題,提高推理性能。
測試時縮放
使用 Best-of-N 評估策略和 VisualPRM-8B 作為評估模型,為推理和數學評估選擇最佳響應,提高模型的推理能力。
消融實驗
原生多模態預訓練
通過實驗證明,原生多模態預訓練能賦予模型強大的多模態能力,且在更高質量數據上進行指令微調後,模型性能進一步提升。
混合偏好優化
使用 MPO 進行微調的模型在多模態推理基準測試中表現更出色,性能提升主要源於訓練算法。
可變視覺位置編碼
引入 V2PE 顯著提升了大多數評估指標的性能,較小的位置增量 \( \delta \) 對傳統上下文任務也能實現最佳性能。
📄 許可證
本項目遵循 MIT 許可證發佈。本項目使用了預訓練的 Qwen2.5 作為組件,該組件遵循 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:
@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}
}









