Please provide the following information when requesting support. • Hardware (T4/V100/Xavier/Nano/etc): GPU • Network Type (Detectnet_v2/Faster_rcnn/Yolo_v4/LPRnet/Mask_rcnn/Classification/etc): Yolo_v4 • TLT Version (Please run “tlt info --verbose” and share “docker_tag” here): NA • Training sp…

[image] hyperlight: We tried that and got the same error: [I] [TRT] Calibration table does not match calibrator algorithm type. [E] Error[3]: IExecutionContext::executeV2: Error Code 3: API Usage Error (Parameter check failed, condition: nullPtrAllowed. Tensor "Input" is bound to nullptr, which…

[image] hyperlight: Can I re-export the original .tlt file (trained using TAO 3.0) to .onnx using TAO 5.0? Please try either of below options. Please try to change the .etlt file to .onnx file. Reference: tao_toolkit_recipes/tao_forum_faq/FAQ.md at main · NVIDIA-AI-IOT/tao_toolkit_recipes ·…

@Morganh After I convert the .etlt file to .onnx file and use trtexec to build an int8 engine from the .onnx file. The .onnx file was converted from a .etlt file exported using TAO 3.0 following the link you provided, the calibration file is for the .etlt file, I reuse the same the calibration fil…

[image] hyperlight: yolov4.bin Can you share the yolov4.bin ?

@Morganh Here is the file (I had to add .txt extension because .bin file is not allowed to be uploaded): yolov4.bin.txt (3.1 KB)

[image] hyperlight: After I convert the .etlt file to .onnx file and use trtexec to build an int8 engine from the .onnx file. The .onnx file was converted from a .etlt file exported using TAO 3.0 following the link you provided, the calibration file is for the .etlt file, I reuse the same the c…

@Morganh To remove TAO 3.0 to TAO 5.5 migration complexity, we repeated the exact same steps using a .hdf5 model trained with TAO 5.5. We exported the ONNX and calibration files again, following the same procedure, and we still got the same error. QAT is enabled in both the training and retraining …

[image] hyperlight: QAT is enabled in both the training and retraining processes. OK, this is the important information. For QAT workflow, you can refer to “12.7. Deployment of the QAT model” of tao_tutorials/notebooks/tao_launcher_starter_kit/yolo_v4/yolo_v4.ipynb at tao_5.5_release · NVIDIA…

@Morganh Hardware/Software info: Train PC: TAO 5.5, GPU: GeForce 4090 Engine building PC: DeepStream 8.0, GPU: GeForce 5090 We followed this: Does TAO 5.0 support exporting a model trained by TAO 3.0? - #9 by Morganh But there was no calibration file generated. Is that expected? I went ahead a…

Hi @hyperlight , Thanks for the detailed info. Seems that the two cases have similar behavior. Let us focus on case2 to check further. Use the steps of notebook[ tao_tutorials/notebooks/tao_launcher_starter_kit/yolo_v4/yolo_v4.ipynb at tao_5.5_release · NVIDIA/tao_tutorials · GitHub ] as below. [i…

Hi @Morganh , Based on your suggestion, after running tao model yolo_v4 export, my co-worker inspected the exported ONNX file in Netron and confirmed that there are no QDQ nodes. I’ve attached a screenshot of the exported model for reference. (Note: enable_qat: true was set during the train, prune, …

Does TAO 5.0 support exporting a model trained by TAO 3.0?

Accelerated Computing Intelligent Video Analytics TAO Toolkit

Morganh November 6, 2025, 5:56am 3

Please try either of below options.

Please try to change the .etlt file to .onnx file. Reference: tao_toolkit_recipes/tao_forum_faq/FAQ.md at main · NVIDIA-AI-IOT/tao_toolkit_recipes · GitHub.
Please try to change the .tlt file to .hdf5 file. Reference: tao_toolkit_recipes/tao_forum_faq/FAQ.md at main · NVIDIA-AI-IOT/tao_toolkit_recipes · GitHub. And then export to .onnx file using TAO5.0.

Topic		Replies	Views
Tao deploy error - TAO Toolkit jetson , deepstream	43	902	August 25, 2025
Unable to deploy TAO 4.0.1 yolov4 model on deepstream6.0 TAO Toolkit deepstream	42	1702	August 18, 2023
Convert TAO Yolov4 model to DLA engine fails TAO Toolkit	21	2042	February 15, 2022
Unable to generate tensorrt engine using ds-tao-detection app for yolov4_tiny for QAT trained etlt model DeepStream SDK	15	839	May 31, 2023
How to integrate TAO custom trained model 's tao export and tao deploy files with Deep Streem DeepStream SDK	7	554	January 18, 2023
Unable to export QAT yolov3 in int8 TAO Toolkit	6	693	April 25, 2023
Convert model to Jetson Error during model export step in TAO notebook TAO Toolkit	20	2373	January 30, 2022
Error when generating engine file from a TAO trained yolov4_tiny model in Deepstream 6.1.1 DeepStream SDK	10	615	June 12, 2023
TAO Toolkit Yolov4 DeepStream SDK	19	783	November 30, 2022
The effect is very poor when converted to trt TAO Toolkit tensorrt , ubuntu	60	2256	August 28, 2023

Does TAO 5.0 support exporting a model trained by TAO 3.0?

Related topics