Commits
  • check
    ydshieh committed 24 days ago
  • check
    ydshieh committed 24 days ago
  • check
    ydshieh committed 24 days ago
  • check
    ydshieh committed 24 days ago
  • tests: skip SIGABRT after destroy_process_group:done as gloo shutdown race
    ydshieh committed 24 days ago
  • tests: richer SIGABRT diagnostics — thread snapshot + ctypes C backtrace
    ydshieh committed 24 days ago
  • tests: restore faulthandler for SIGABRT, add glibc/OS/torch env to breadcrumb
    ydshieh committed 24 days ago
  • try again
    ydshieh committed 24 days ago
  • tests: add torch.utils.collect_env to SIGABRT diagnostics
    ydshieh committed 24 days ago
  • try again
    ydshieh committed 24 days ago
  • check
    ydshieh committed 23 days ago
  • check
    ydshieh committed 23 days ago
  • check both
    ydshieh committed 23 days ago
  • try again
    ydshieh committed 23 days ago
  • try with shared cache
    ydshieh committed 23 days ago
  • check pytest -n 1
    ydshieh committed 23 days ago
  • try with n1
    ydshieh committed 23 days ago
  • check pytest -n 1
    ydshieh committed 23 days ago
  • try with n1
    ydshieh committed 23 days ago
  • check wait
    ydshieh committed 23 days ago
  • check without wait
    ydshieh committed 22 days ago
  • check without torch job
    ydshieh committed 22 days ago
  • check run-slow
    ydshieh committed 22 days ago
  • check torch.set_num_threads(1)
    ydshieh committed 21 days ago
  • check torch.set_num_threads(1)
    ydshieh committed 21 days ago
  • check torch.set_num_threads(1)
    ydshieh committed 21 days ago
  • add richer diagnostics to TP/EP distributed test crash handler
    ydshieh committed 21 days ago
  • Revert "add richer diagnostics to TP/EP distributed test crash handler"
    ydshieh committed 20 days ago
  • fix SIGABRT in TP/EP tests by forcing gc.collect() after destroy_process_group
    ydshieh committed 20 days ago
  • Revert "fix SIGABRT in TP/EP tests by forcing gc.collect() after destroy_process_group"
    ydshieh committed 20 days ago
Loading