光合开发者社区-论坛

DeepSeek

faster_whisper 需要ctranslate2 但是只有dtk2404的，没有dtk2504的，怎么解决

2025-06-24 16:36:11

151

dtk环境为：dtk2504
DCU环境为：K100_AI
使用的docker 镜像为：image.sourcefind.cn:5000/dcu/admin/base/pytorch:2.4.1-ubuntu22.04-dtk25.04-py3.10-fixpy

运行报如下错误：

Traceback (most recent call last):
File "/home/agent-code/audio_receiver_agent.py", line 7, in <module>
from faster_whisper.transcribe import BatchedInferencePipeline, WhisperModel
File "/usr/local/lib/python3.10/dist-packages/faster_whisper/init .py", line 2, in <module>
from faster_whisper.transcribe import WhisperModel
File "/usr/local/lib/python3.10/dist-packages/faster_whisper/transcribe.py", line 10, in <module>
import ctranslate2
File "/usr/local/lib/python3.10/dist-packages/ctranslate2/init .py", line 21, in <module>
from ctranslate2._ext import (
ImportError: /usr/local/lib/python3.10/dist-packages/ctranslate2/libcublas.so.11: undefined symbol: rocblas_ztrmm_outofplace

网

网友93kuK7

2楼

同样问题，能否给一个dtk 25.04.1 能用的 ctranslate2 包哇

2025/7/22

潜水1号

1楼

您直接安装 2404的包能用么？https://download.sourcefind.cn:65024/4/main/ctranslate2/DAS1.3 如果还不能用麻烦您提供一下测试用例

2025/6/24

网

网友tscL0p

回复 @潜水1号

是安装的2404，https://download.sourcefind.cn:65024/directlink/4/ctranslate2/DAS1.3/ctranslate2-4.1.0+das.opt1.dtk24043-cp310-cp310-manylinux_2_31_x86_64.whl

2025/6/24

网

网友tscL0p

回复 @潜水1号

# 加载模型

def load_model():

model_path = "/opt/models/angelala00/faster-whisper-small"

device = "cuda" if torch.cuda.is_available() else "cpu"

compute_type = "float16" if device == "cuda" else "int8"

try:

model = WhisperModel(

model_path,

device=device,

compute_type=compute_type

)

print("Model loaded successfully")

return model

except Exception as e:

print(f"Error loading model: {str(e)}")

raise

# 加载模型

def load_model():

model_path = "/opt/models/angelala00/faster-whisper-small"

device = "cuda" if torch.cuda.is_available() else "cpu"

compute_type = "float16" if device == "cuda" else "int8"

try:

model = WhisperModel(

model_path,

device=device,

compute_type=compute_type

)

print("Model loaded successfully")

return model

except Exception as e:

print(f"Error loading model: {str(e)}")

raise

2025/6/24

网

网友93kuK7

回复 @网友tscL0p

解决了吗

2025/12/5

全部4条回复