相册img2text用不了GPU啊

菠菜x

华勤J5005主板，编译Model直接失败了....

# lspci | grep VGA

00:02.0 VGA compatible controller: Intel Corporation GeminiLake [UHD Graphics 605] (rev 03)

SPDLOG: create spdlog success!
[01:09:32.841] [info] Program starting......
[01:09:32.841] [info] device:GPU
[01:09:32.841] [info] model_format:OPENVINO
[01:09:32.841] [info] log_file:/usr/trim/logs/ai_manager/com.trim.img2vec/app.log
[01:09:32.841] [info] log_level:info
[01:09:32.841] [info] num_threads_onnx:1
[01:09:32.841] [info] onnx_model:
[01:09:32.841] [info] openvino_model:/vol1/@sysappmeta/ai-manager/models/img2vec-large-model-1.0.0/img2vec.xml
[01:09:32.841] [info] openvino_weights:/vol1/@sysappmeta/ai-manager/models/img2vec-large-model-1.0.0/img2vec.bin
[01:09:32.841] [info] openvino_cache_dir:/vol1/@sysappmeta/ai-manager/temp/com.trim.img2vec/openvino-cache
Traceback (most recent call last):
File "/workspace/img2vec/img2vec.py", line 119, in <module>
File "/workspace/img2vec/img2vec.py", line 59, in __init__
File "/usr/trim/lib/ai_manager/pyenv/lib/python3.11/site-packages/openvino/runtime/ie_api.py", line 543, in compile_model
super().compile_model(model, device_name, {} if config is None else config),
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: Exception from src/inference/src/cpp/core.cpp:104:
Exception from src/inference/src/dev/plugin.cpp:53:
Check 'false' failed at src/plugins/intel_gpu/src/plugin/program_builder.cpp:185:
[GPU] ProgramBuilder build failed!
Exception from src/plugins/intel_gpu/src/runtime/ocl/ocl_stream.cpp:429:
[GPU] clWaitForEvents, error code: -14

SPDLOG: create spdlog success!
[01:10:05.834] [info] Program starting......
[01:10:05.846] [info] device:CPU
[01:10:05.846] [info] model_format:ONNX
[01:10:05.846] [info] log_file:/usr/trim/logs/ai_manager/com.trim.img2vec/app.log
[01:10:05.846] [info] log_level:info
[01:10:05.846] [info] num_threads_onnx:2
[01:10:05.846] [info] onnx_model:/vol1/@sysappmeta/ai-manager/models/img2vec-large-model-1.0.0/img2vec.onnx
[01:10:05.846] [info] openvino_model:
[01:10:05.846] [info] openvino_weights:
[01:10:05.846] [info] openvino_cache_dir:

飞牛技术同学

智能识别任务里，有些任务不支持GPU运行会降级到CPU处理 后续我们会持续兼容