目录
开源地址：
https://github.com/OpenBMB/MiniCPM
openbmb/MiniCPM4-8B-Eagle-vLLM
模型 2.29G
模型repo下载：
modelscope download --model=OpenBMB/MiniCPM-V-2_6 --local_dir ./MiniCPM-V-2_6
单文件GGUF下载：
modelscope download --model=OpenBMB/MiniCPM-V-2_6-gguf --local_dir ./ ggml-model-Q4_K_M.gguf
单图片demo：
自动下载模型：
C:\Users\xxx\.cache\modelscope\hub\models\OpenBMB\MiniCPM-V-2_6
需要显存20G，模型文件15G
# test.py
import torch
from PIL import Image
from modelscope import AutoModel, AutoTokenizer
model = AutoModel.from_pretrained('OpenBMB/MiniCPM-V-2_6', trust_remote_code=True,
attn_implementation='sdpa', torch_dtype=torch.bfloat16) # sdpa or flash_attention_2, no eager
model = model.eval().cuda()
tokenizer = AutoTokenizer.from_pretrained('OpenBMB/MiniCPM-V-2_6', trust_remote_code=True)
image = Image.open(r"B:\360MoveData\Users\Administrator\Pictures\liuying\IMG_20150903_123711.jpg").convert('RGB')
question = 'What is in the image?'
msgs = [{'role': 'user', 'content': [image, question]}]
res = model.chat(
image=None,
msgs=msgs,
tokenizer=tokenizer
)
print(res)
## if you want to use streaming, please make sure sampling=True and stream=True
## the model.chat will return a generator
res = model.chat(
image=None,
msgs=msgs,
tokenizer=tokenizer,
sampling=True,
stream=True
)
generated_text = ""
for new_text in res:
generated_text += new_text
print(new_text, flush=True, end='')
多图推理demo：
import torch
from PIL import Image
from modelscope import AutoModel, AutoTokenizer
model = AutoModel.from_pretrained('OpenBMB/MiniCPM-V-2_6', trust_remote_code=True,
attn_implementation='sdpa', torch_dtype=torch.bfloat16) # sdpa or flash_attention_2, no eager
model = model.eval().cuda()
tokenizer = AutoTokenizer.from_pretrained('OpenBMB/MiniCPM-V-2_6', trust_remote_code=True)
image1 = Image.open('image1.jpg').convert('RGB')
image2 = Image.open('image2.jpg').convert('RGB')
question = 'Compare image 1 and image 2, tell me about the differences between image 1 and image 2.'
msgs = [{'role': 'user', 'content': [image1, image2, question]}]
answer = model.chat(
image=None,
msgs=msgs,
tokenizer=tokenizer
)
print(answer)
视频理解
import torch
from PIL import Image
from modelscope import AutoModel, AutoTokenizer
from decord import VideoReader, cpu # pip install decord
params={}
model = AutoModel.from_pretrained('OpenBMB/MiniCPM-V-2_6', trust_remote_code=True,
attn_implementation='sdpa', torch_dtype=torch.bfloat16) # sdpa or flash_attention_2, no eager
model = model.eval().cuda()
tokenizer = AutoTokenizer.from_pretrained('OpenBMB/MiniCPM-V-2_6', trust_remote_code=True)
MAX_NUM_FRAMES=64
def encode_video(video_path):
def uniform_sample(l, n):
gap = len(l) / n
idxs = [int(i * gap + gap / 2) for i in range(n)]
return [l[i] for i in idxs]
vr = VideoReader(video_path, ctx=cpu(0))
sample_fps = round(vr.get_avg_fps() / 1) # FPS
frame_idx = [i for i in range(0, len(vr), sample_fps)]
if len(frame_idx) > MAX_NUM_FRAMES:
frame_idx = uniform_sample(frame_idx, MAX_NUM_FRAMES)
frames = vr.get_batch(frame_idx).asnumpy()
frames = [Image.fromarray(v.astype('uint8')) for v in frames]
print('num frames:', len(frames))
return frames
video_path="car.mp4"
frames = encode_video(video_path)
question = "Describe the video"
msgs = [
{'role': 'user', 'content': frames + [question]},
]
# Set decode params for video
params["use_image_id"] = False
params["max_slice_nums"] = 2 # 如果cuda OOM且视频分辨率大于448*448 可设为1
answer = model.chat(
image=None,
msgs=msgs,
tokenizer=tokenizer,
**params
)
print(answer)
论文学习笔记：
MiniCPM，能被斯坦福抄袭究竟有何魅力？我们一起看看论文吧！-腾讯云开发者社区-腾讯云
部署完整教程：
MiniCPM-V 2.6：端侧最强多模态大模型探索【推理实战大全】_vllm minicpm-CSDN博客
微调教程：
MiniCPM-o-2.6 多模态大模型微调实战（完整代码）_minicpm-o 2.6-CSDN博客
部署，微调教程，视频实测
多图、视频首上端！面壁「小钢炮」 MiniCPM-V 2.6 模型重磅上新！魔搭推理、微调、部署实战教程modelscope-CSDN博客
BitCPM4 技术报告
而刚刚提到的43页技术报告，我看了一遍，觉得可以拆成以下：
InfLLM v2：Attention 层只看重点
FR-Spec：草稿阶段不全写
BitCPM4：训练时就考虑压缩
CPM.cu + ArkInfer：定制推理 & 部署系统
风洞 2.0：小模型先试，大模型再训