Part Number: AM69A
Tool/software:
Hello,
I am looking to compile a Conv2D model with the following options :
compile_options = {
'tidl_tools_path' : os.environ['TIDL_TOOLS_PATH'],
'artifacts_folder' : output_dir,
'tensor_bits' : 8,
'accuracy_level' : 1,
'advanced_options:calibration_frames' : len(calib_images),
'advanced_options:calibration_iterations' : 16,
'advanced_options:inference_mode' : 1,
'advanced_options:num_cores' : 4,
'core_start_idx' : 1,
}
However, my kernel crashes (see error).
The Kernel crashed while executing code in the current cell or a previous cell. Please review the code in the cell(s) to identify a possible cause of the failure. Click here for more info. View Jupyter log for further details.
By removing certain options ('advanced_options:num_cores' : 4, 'core_start_idx' : 1,), the kernel doesn't crash anymore but I get this other error:
========================= [Model Compilation Started] ========================= Model compilation will perform the following stages: 1. Parsing 2. Graph Optimization 3. Quantization & Calibration 4. Memory Planning ============================== [Version Summary] ============================== ------------------------------------------------------------------------------- | TIDL Tools Version | 10_00_08_00 | ------------------------------------------------------------------------------- | C7x Firmware Version | 10_00_02_00 | ------------------------------------------------------------------------------- ============================== [Parsing Started] ============================== Number of OD backbone nodes = 86 Size of odBackboneNodeIds = 86 Total Nodes = 104 ------------------------------------------------------------------------------- | Core | No. of Nodes | Number of Subgraphs | ------------------------------------------------------------------------------- ... =================== [Optimization for subgraph_264 started] =================== [TIDL Import] ERROR: High Throughtput Inference Mode is not supported when partial batch is detected in graph -- [tidl_import_core.cpp, 2960] [TIDL Import] ERROR: Network Optimization failed - Failed in function: TIDL_runtimesOptimizeNet -- [tidl_runtimes_import_common.cpp, 1268] Output is truncated. View as a scrollable element or open in a text editor. Adjust cell output settings... 0%| | 0/4 [00:00<?, ?it/s] Number of subgraphs:1 , 104 nodes delegated out of 104 nodes
Could you explain to me what this error means and what I can do to work around it? I have already tried inference_mode 0 and 2, and the results are not relevant.
Thank you very much,
Azer
