return torch._C._nn.upsample_nearest3d(input, output_size, scale_factors) RuntimeError: "upsample_nearest3d_out_frame" not implemented for 'BFloat16'

[shifuxiao]

when i use this mode,it have this error.my code is:import requests import base64 from io import BytesIO from PIL import Image import matplotlib.pyplot as plt

url = 'http://192.168.1.59:8834/infer'

with open(path, 'rb') as image_file:

img_bytes = image_file.read()

img_base64 = base64.base64encode(img_bytes).decode('utf-8')

input_data_dict = { 'prompt':"生成一个小孩吃饭的视频",

'image_data':img_base64,

'image_data':'None', 'num_inference_steps':50, 'generate_type':'t2v' }

#发送 POST 请求 response = requests.post(url, json=input_data_dict) print(response) data = response.json() print(data) if response.status_code ==200: with open(r"E:\workSpace\tools\output.mp4","wb") as f: f.write(response.content) print("video saved as 'output")

[OleehyO]

The functionality in the cogvideo repository is no longer maintained. We recommend using cogkit to deploy the cogvideo API server, which will be more user-friendly.