.xmodel file generation

[ravi54116]

I extracted ".xmodel and .protxt" from model.yaml using below command but I didn't get any quantized.h5 file. i directly got model file. For this when I am running I am getting following error in smartcam docker.

• wget https://www.xilinx.com/bin/public/openDownload?filename=yolov3_coco_416_tf2-zcu102_zcu104_kv260-r2.5.0.tar.gz

name: yolov3_coco_416_tf2
type: xmodel
board: zcu102 & zcu104 & kv260
download link: https://www.xilinx.com/bin/public/openDownload?filename=yolov3_coco_416_tf2-zcu102_zcu104_kv260-r2.5.0.tar.gz
checksum: ae417567b1462c7d5b6708285643f140

---

[I 2025-02-15 17:51:28.097 ServerApp] Kernel restarted: 127238e8-e26c-4698-bb71-2edcbd07c57e
WARNING: Logging before InitGoogleLogging() is written to STDERR
W0215 17:51:48.107272 146 dpu_runner_base_imp.cpp:676] CHECK fingerprint fail ! model_fingerprint 0x101000016010407 dpu_fingerprint 0x101000016010406
F0215 17:51:48.107425 146 dpu_runner_base_imp.cpp:648] fingerprint check failure.

Hello, I am trying to generate .xmodel file using quantized.h5 and arch.json. I used Vitis AI 2.5 version for this. Because new vitis ai models don't have tf2_yolov3_coco_416_416_65.9G_2.5. I am facing the following error. Please help me resolving it.

(vitis-ai-tensorflow2) Vitis-AI /workspace > vai_c_tensorflow2 -m /workspace/tf2_yolov3_coco_416_416_65.9G_2.5/quantized/quantized.h5 -a /workspace/arch.json -o /

workspace -n yolov3_coco_416_tf2

---

• VITIS_AI Compilation - Xilinx Inc.

---

[INFO] Namespace(batchsize=1, inputs_shape=None, layout='NHWC', model_files=['/workspace/tf2_yolov3_coco_416_416_65.9G_2.5/quantized/quantized.h5'], model_type='tensorflow2', named_inputs_shape=None, out_filename='/tmp/yolov3_coco_416_tf2_0x101000016010406_org.xmodel', proto=None)
[INFO] tensorflow2 model: /workspace/tf2_yolov3_coco_416_416_65.9G_2.5/quantized/quantized.h5
[INFO] keras version: 2.6.0
[INFO] Tensorflow Keras model type: functional
[INFO] parse raw model : 0%| | 0/181 [00:00<?, ?it/s]
Traceback (most recent call last):
File "/opt/vitis_ai/conda/envs/vitis-ai-tensorflow2/bin/xnnc-run", line 33, in
sys.exit(load_entry_point('xnnc==2.5.0', 'console_scripts', 'xnnc-run')())
File "/opt/vitis_ai/conda/envs/vitis-ai-tensorflow2/lib/python3.7/site-packages/xnnc/main.py", line 49, in main
runner.normal_run(args)
File "/opt/vitis_ai/conda/envs/vitis-ai-tensorflow2/lib/python3.7/site-packages/xnnc/runner.py", line 123, in normal_run
target=target,
File "/opt/vitis_ai/conda/envs/vitis-ai-tensorflow2/lib/python3.7/site-packages/xnnc/xconverter.py", line 145, in run
model_files, model_type, _layout, in_shapes, batchsize
File "/opt/vitis_ai/conda/envs/vitis-ai-tensorflow2/lib/python3.7/site-packages/xnnc/core.py", line 123, in make_xmodel
model_type=model_t,
File "/opt/vitis_ai/conda/envs/vitis-ai-tensorflow2/lib/python3.7/site-packages/xnnc/translator/tensorflow_translator.py", line 107, in to_xmodel
model_type,
File "/opt/vitis_ai/conda/envs/vitis-ai-tensorflow2/lib/python3.7/site-packages/xnnc/translator/tensorflow_translator.py", line 177, in create_xmodel
name, layers, layout, in_shapes, batchsize
File "/opt/vitis_ai/conda/envs/vitis-ai-tensorflow2/lib/python3.7/site-packages/xnnc/translator/tensorflow_translator.py", line 458, in __create_xmodel_from_tf2
), f"[ERROR] Invalid shape of input layer: shape: {shape} (N,H,W,C), name: {xnode.op_name}"
AssertionError: [ERROR] Invalid shape of input layer: shape: [1, None, None, 3] (N,H,W,C), name: image_input

[jimheaton]

Please see the following from the VAI-2.5 User Guide, I believe this describes what you are seeing:

Sometimes, the TensorFlow model does not contain input tensor shape information because itmight cause the compilation to fail. You can specify the input tensor shape with an extra optionlike --options '{"input_shape": "1,224,224,3"}'.