Skip to content

wzqvip/openpi-onnx

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

231 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

OpenPI-ONNX

This repository covers the OpenPI ONNX/TensorRT workflow: JAX → PyTorch conversion, INT8/FP4 quantization, and benchmark results for LIBERO tasks.

Documentation Map

JAX → PyTorch

ONNX Export & INT8 Quantization

FP4 Notes

Benchmarking & Results

Results Summary

Precision Accuracy Latency VRAM Notes
FP32 (PyTorch) 93.75% (750/800) 262.41 ms mean (P99 278.79 ms) 8.10 GB Inference latency
INT8 (TensorRT v1) 98.25% (786/800) ~162 ms mean (P99 ~167 ms) ~4.95 GB Inference latency

INT8 latency is reported from inference logs (mean ~162 ms, P99 ~167 ms).

FP32 PyTorch Baseline (20 trials per task)

Config: 20 trials per task × 4 suites = 800 episodes | seed: 42 | date: 2026-02-11/12

Suite Accuracy Success/Total Avg Latency (ms) Median (ms) P99 (ms) GPU Memory (GB)
libero_spatial 99.5% 199/200 263.23 261.68 286.90 8.10
libero_goal 91.0% 182/200 259.49 258.43 271.99 8.10
libero_object 95.0% 190/200 264.36 263.70 283.00 8.10
libero_10 89.5% 179/200 262.56 262.46 273.26 8.10
Overall 93.75% 750/800 262.41 261.57 278.79 8.10

Full table and logs: benchmark_results/FP32_RESULTS_20TRIALS.md

INT8 TensorRT (v1) (20 trials per task)

Accuracy:

Suite Accuracy Success/Total
libero_spatial 99.00% 198/200
libero_goal 98.50% 197/200
libero_object 99.50% 199/200
libero_10 96.00% 192/200
Overall 98.25% 786/800

Latency (inference): mean ~162 ms, median ~161 ms, P99 ~167 ms (from evaluation logs)

GPU Memory: 4954 MiB (~4.95 GB) with the engine loaded (measured via nvidia-smi while serve_trt.py is running)

Full details: INT8_FINAL_RESULTS.md and INT8_SUMMARY.md

About

ONNX quant and deployment

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

  • Python 96.3%
  • Shell 2.2%
  • PowerShell 1.2%
  • Other 0.3%