本地部署 通义千问视觉大模型 2.5-VL-3B-Instruct-AWQ量化 (windows)

自 Qwen2-VL 发布以来的五个月里,众多开发者在 Qwen2-VL 视觉语言模型的基础上构建了新的模型,并为我们提供了宝贵的反馈。在此期间,我们专注于构建更有用的视觉语言模型。今天,我们非常激动地向大家介绍 Qwen 家族的最新成员:Qwen2.5-VL

视觉理解:Qwen2.5-VL 不仅擅长识别常见的物体,如花、鸟、鱼和昆虫,而且还能高效分析图像中的文本、图表、图标、图形和布局。

自主代理:Qwen2.5-VL 直接作为视觉代理,能够进行推理并动态指导工具使用,具备计算机和手机操作能力。

理解长视频并捕捉事件:Qwen2.5-VL 可以理解超过1小时的视频,并且这次新增了通过定位相关视频片段来捕捉事件的能力。

支持多种格式的视觉定位:Qwen2.5-VL 可以通过生成边界框或点来准确地定位图像中的对象,并能提供稳定的 JSON 输出,包括坐标和属性。

生成结构化输出:对于发票、表格等扫描数据,Qwen2.5-VL 支持其内容的结构化输出,适用于金融、商业等领域。

参考连接:https://modelscope.cn/models/Qwen/Qwen2.5-VL-3B-Instruct-AWQ

 

 

本次使用的操作系统,以及环境配置

操作系统:windows11

CPU:  i7-11800H

内存:16GB

GPU:RTX3050ti  4G显存

CUDA Version: 12.6

 

使用anaconda 创建用于qwen的虚拟环境

conda create -n qwen python=3.9
activate qwen

 

构建Qwen2.5-VL(需要安装git,通过命令从源码构建)

windows安装git教程,参考链接:git的安装与配置教程--超详细版 - 知乎

pip install git+https://github.com/huggingface/transformers accelerate

 

安装工具包,以帮助您更方便地处理各种类型的视觉输入,包括 base64、URL 以及交错的图像

pip install qwen-vl-utils[decord]==0.0.8

 

安装modelscope,下载模型

pip install modelscope

在你的工作区创建一个download_model.py文件,用于下载文件,这里下载的是Qwen/Qwen2.5-VL-3B-Instruct-AWQ模型

# 通过ModelScope 的 snapshot_download函数下载模型

from modelscope import snapshot_download

# 下载Qwen2.5-VL-3B-Instruct-AWQ模型
model_dir = snapshot_download(model_id="Qwen/Qwen2.5-VL-3B-Instruct-AWQ", local_dir="models/Qwen2___5-VL-3B-Instruct-AWQ")

# 下载Qwen2.5-VL-3B-Instruct模型
# model_dir = snapshot_download("Qwen/Qwen2.5-VL-3B-Instruct")
# 下载Qwen2.5-VL-7B-Instruct7模型
# model_dir = snapshot_download("Qwen/Qwen2.5-VL-7B-Instruct7")
# 下载Qwen2.5-VL-72B-Instruct模型
# model_dir = snapshot_download("Qwen/Qwen2.5-VL-72B-Instruct")

print(f"模型已下载到: {model_dir}")

 

安装pytorch 2.3.1依赖( cuda 12.6)

# 安装依赖(cuda 12.1 或 cuda 12.6)
pip install -i https://pypi.tuna.tsinghua.edu.cn/simple torch==2.3.1 torchvision==0.18.1 torchaudio==2.3.1 --index-url https://download.pytorch.org/whl/cu121

# 检测torch、cuda、cudnn版本
import torch
print(torch.__version__)
print(torch.version.cuda)
print(torch.backends.cudnn.version())

#是否可用gpu
flag = torch.cuda.is_available()
print(flag)

 

通过代码实现,调用Qwen/Qwen2.5-VL-3B-Instruct-AWQ模型

# 图像
import torch
from modelscope import Qwen2_5_VLForConditionalGeneration, AutoTokenizer, AutoProcessor
from qwen_vl_utils import process_vision_info

# 刚刚下载后弹出的下载地址,就填入这里
model_dir="models/Qwen2___5-VL-3B-Instruct-AWQ"
# model_dir = "C:/Users/malin/.cache\/modelscope/hub/models/Qwen/Qwen2___5-VL-3B-Instruct-AWQ"

# default: Load the model on the available device(s)
# model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
#     "Qwen/Qwen2.5-VL-3B-Instruct-AWQ", torch_dtype="auto", device_map="auto"
# )
print(111)
model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
    model_dir, torch_dtype=torch.float16, device_map="cuda:0"
)

# We recommend enabling flash_attention_2 for better acceleration and memory saving, especially in multi-image and video scenarios.
# model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
#     "Qwen/Qwen2.5-VL-3B-Instruct-AWQ",
#     torch_dtype=torch.bfloat16,
#     attn_implementation="flash_attention_2",
#     device_map="auto",
# )

# default processer
# processor = AutoProcessor.from_pretrained("Qwen/Qwen2.5-VL-3B-Instruct-AWQ")
print(222)
# default processer
# 这里做了推理时候的图片压缩,平衡精度和速度,默认是全尺寸图片进行推理,很慢也要很大的显存
min_pixels = 256 * 28 * 28
max_pixels = 1280 * 28 * 28
processor = AutoProcessor.from_pretrained(
    model_dir, min_pixels=min_pixels, max_pixels=max_pixels, use_fast=True,
)
# processor = AutoProcessor.from_pretrained(model_dir, use_fast=True,)

# The default range for the number of visual tokens per image in the model is 4-16384.
# You can set min_pixels and max_pixels according to your needs, such as a token range of 256-1280, to balance performance and cost.
# min_pixels = 256*28*28
# max_pixels = 1280*28*28
# processor = AutoProcessor.from_pretrained("Qwen/Qwen2.5-VL-3B-Instruct-AWQ", min_pixels=min_pixels, max_pixels=max_pixels)

# 本地图像路径
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "image": "4cf739bc-effe-4a64-a29b-156dc12a7f73.jpg"},
            {"type": "text", "text": "识别铁罐上的文字,只输出最高准确度的数字"},
        ],
    }
]

# 图像URL
# messages = [
#     {
#         "role": "user",
#         "content": [
#             {
#                 "type": "image",
#                 "image": "http://192.168.31.164/resource/pic_iron2/20250522101928-af4f7146-8dc8-43a0-8554-2f052c2eabe2.jpg",
#             },
#             {"type": "text", "text": "识别铁罐上的文字,只输出最高准确度的数字"},
#         ],
#     }
# ]

## Base64 encoded image
# messages = [
#     {
#         "role": "user",
#         "content": [
#             {"type": "image", "image": "data:image/jpeg;base64,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"},
#             {"type": "text", "text": "识别铁罐上的文字,只输出最高准确度的数字"},
#         ],
#     }
# ]


# Preparation for inference
text = processor.apply_chat_template(
    messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs = process_vision_info(messages)
inputs = processor(
    text=[text],
    images=image_inputs,
    videos=video_inputs,
    padding=True,
    return_tensors="pt",
)
inputs = inputs.to("cuda")

# Inference: Generation of the output
generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [
    out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_text = processor.batch_decode(
    generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print(output_text)

 

 

安装环境时出现的报错

1.ImportError: cannot import name 'shard_checkpoint' from 'transformers.modeling_utils'

这是因为通过 pip install git+https://github.com/huggingface/transformers accelerate源码构建transformers的方式,transformers是最新的4.53.0的版本,但是shard_checkpoint方法在4.47.0以上的版本就已经删掉了,而且也无法手动降低transformers的版本,会导致 KeyError: 'qwen2_5_vl'报错

# 补充shard_checkpoint方法,transformers4.47.0版本以后删除了
def shard_checkpoint(
    state_dict: Dict[str, torch.Tensor], max_shard_size: Union[int, str] = "10GB", weights_name: str = WEIGHTS_NAME
):
    """
    Splits a model state dictionary in sub-checkpoints so that the final size of each sub-checkpoint does not exceed a
    given size.

    The sub-checkpoints are determined by iterating through the `state_dict` in the order of its keys, so there is no
    optimization made to make each sub-checkpoint as close as possible to the maximum size passed. For example, if the
    limit is 10GB and we have weights of sizes [6GB, 6GB, 2GB, 6GB, 2GB, 2GB] they will get sharded as [6GB], [6+2GB],
    [6+2+2GB] and not [6+2+2GB], [6+2GB], [6GB].

    <Tip warning={true}>

    If one of the model's weight is bigger than `max_shard_size`, it will end up in its own sub-checkpoint which will
    have a size greater than `max_shard_size`.

    </Tip>

    Args:
        state_dict (`Dict[str, torch.Tensor]`): The state dictionary of a model to save.
        max_shard_size (`int` or `str`, *optional*, defaults to `"10GB"`):
            The maximum size of each sub-checkpoint. If expressed as a string, needs to be digits followed by a unit
            (like `"5MB"`).
        weights_name (`str`, *optional*, defaults to `"pytorch_model.bin"`):
            The name of the model save file.
    """
    logger.warning(
        "Note that `shard_checkpoint` is deprecated and will be removed in v4.44. We recommend you using "
        "split_torch_state_dict_into_shards from huggingface_hub library"
    )
    max_shard_size = convert_file_size_to_int(max_shard_size)

    sharded_state_dicts = [{}]
    last_block_size = 0
    total_size = 0
    storage_id_to_block = {}

    for key, weight in state_dict.items():
        # when bnb serialization is used the weights in the state dict can be strings
        # check: https://github.com/huggingface/transformers/pull/24416 for more details
        if isinstance(weight, str):
            continue
        else:
            storage_id = id_tensor_storage(weight)

        # If a `weight` shares the same underlying storage as another tensor, we put `weight` in the same `block`
        if storage_id in storage_id_to_block and weight.device != torch.device("meta"):
            block_id = storage_id_to_block[storage_id]
            sharded_state_dicts[block_id][key] = weight
            continue

        weight_size = weight.numel() * dtype_byte_size(weight.dtype)
        # If this weight is going to tip up over the maximal size, we split, but only if we have put at least one
        # weight in the current shard.
        if last_block_size + weight_size > max_shard_size and len(sharded_state_dicts[-1]) > 0:
            sharded_state_dicts.append({})
            last_block_size = 0

        sharded_state_dicts[-1][key] = weight
        last_block_size += weight_size
        total_size += weight_size
        storage_id_to_block[storage_id] = len(sharded_state_dicts) - 1

    # If we only have one shard, we return it
    if len(sharded_state_dicts) == 1:
        return {weights_name: sharded_state_dicts[0]}, None

    # Otherwise, let's build the index
    weight_map = {}
    shards = {}
    for idx, shard in enumerate(sharded_state_dicts):
        shard_file = weights_name.replace(".bin", f"-{idx+1:05d}-of-{len(sharded_state_dicts):05d}.bin")
        shard_file = shard_file.replace(
            ".safetensors", f"-{idx + 1:05d}-of-{len(sharded_state_dicts):05d}.safetensors"
        )
        shards[shard_file] = shard
        for key in shard.keys():
            weight_map[key] = shard_file

    # Add the metadata
    metadata = {"total_size": total_size}
    index = {"metadata": metadata, "weight_map": weight_map}
    return shards, index

 2.ValueError: You are attempting to load an AWQ model with a device_map that contains a CPU or disk device

指定GPU设备,修改以下代码

# 修改前
model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
    model_dir, torch_dtype=torch.float16, device_map="auto"
)

# 修改后
model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
    model_dir, torch_dtype=torch.float16, device_map="cuda:0"
)

 

 
posted @ 2025-05-22 16:17  马铃薯1  阅读(1806)  评论(1)    收藏  举报