onnxruntime
Enable QNN HTP spill fill buffer setting to save RAM usage.
#22853
Merged
Overview
Commits
12
Changes
View On GitHub
Commits
  • Enable QNN HTP spill fill buffer setting to save RAM usage.
    HectorSVC committed 1 year ago
  • Update onnxruntime/core/providers/qnn/builder/qnn_backend_manager.cc
    HectorSVC committed 1 year ago
  • fix format
    HectorSVC committed 1 year ago
  • dump the max spill fill buffer size into the EPContext node while generating context wrapper model
    HectorSVC committed 1 year ago
  • add option enable_htp_spill_fill_buffer since this feature need to load QnnSystem lib which is not available for Windows x86_64 platform, so that not breaking existing workflow on x86_64 system
    HectorSVC committed 1 year ago
  • merge main
    HectorSVC committed 1 year ago
  • fix UT
    HectorSVC committed 1 year ago
  • update docs/ContribOperators.md
    HectorSVC committed 1 year ago
  • address review comments
    HectorSVC committed 1 year ago
  • Merge branch 'qnn_spill_fill' of https://github.com/microsoft/onnxruntime into qnn_spill_fill
    HectorSVC committed 1 year ago
  • resolve issue cause by merge conflict
    HectorSVC committed 1 year ago
  • format fix
    HectorSVC committed 1 year ago
FAQ Terms Privacy Refunds Impressum
Loading