Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

AIO CPU Locked Tensor #6592

Merged
merged 26 commits into from
Oct 9, 2024
Merged
Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
26 commits
Select commit Hold shift + click to select a range
1413d69
fixed functionality of cpu locked tensor
jomayeri Sep 30, 2024
87a7c69
Merge branch 'master' into jomayeri/aio-locked-tensor
jomayeri Sep 30, 2024
c13bb10
enabling cpu locked in unittests, and fixing compilation errors
jomayeri Oct 1, 2024
248cec4
Merge branch 'jomayeri/aio-locked-tensor' of github.com:microsoft/Dee…
jomayeri Oct 1, 2024
b909702
passing gds tests
jomayeri Oct 2, 2024
b1ee711
renaming all instances of num_threads
jomayeri Oct 2, 2024
ada1b83
updating function names to match
jomayeri Oct 2, 2024
1cb88ce
fix formatting
jomayeri Oct 2, 2024
f5528da
variable name change to fix compilation
jomayeri Oct 2, 2024
f576d29
formatting
jomayeri Oct 2, 2024
5a47bf3
update references in tutorial
jomayeri Oct 3, 2024
fe93fdc
Merge branch 'master' into jomayeri/aio-locked-tensor
jomayeri Oct 7, 2024
b2866cb
Merge branch 'master' into jomayeri/aio-locked-tensor
tjruwase Oct 7, 2024
9c93d2c
Merge branch 'master' into jomayeri/aio-locked-tensor
tjruwase Oct 8, 2024
884c0fd
async_io operator for CPU accelerator
tjruwase Oct 9, 2024
a5ba643
Merge branch 'jomayeri/aio-locked-tensor' of github.com:microsoft/Dee…
tjruwase Oct 9, 2024
ea0e45b
Merge branch 'master' into jomayeri/aio-locked-tensor
tjruwase Oct 9, 2024
98988cd
Formatting; Use int64_t
tjruwase Oct 9, 2024
b30cda5
Merge branch 'jomayeri/aio-locked-tensor' of github.com:microsoft/Dee…
tjruwase Oct 9, 2024
90e25da
Skip fp16 tests on CPU
tjruwase Oct 9, 2024
60ae3e0
Merge branch 'master' into jomayeri/aio-locked-tensor
tjruwase Oct 9, 2024
a008d4c
Merge branch 'master' into jomayeri/aio-locked-tensor
tjruwase Oct 9, 2024
59d8dfa
Merge branch 'master' into jomayeri/aio-locked-tensor
tjruwase Oct 9, 2024
8a52388
Add Cuda 12.6
tjruwase Oct 9, 2024
d1afe4c
Merge branch 'jomayeri/aio-locked-tensor' of github.com:microsoft/Dee…
tjruwase Oct 9, 2024
0182208
Merge branch 'master' into jomayeri/aio-locked-tensor
tjruwase Oct 9, 2024
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
formatting
  • Loading branch information
jomayeri committed Oct 2, 2024
commit f576d291a7675df1b85e4201b58ea1013b38938c
2 changes: 1 addition & 1 deletion csrc/aio/py_lib/deepspeed_aio_thread.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -29,7 +29,7 @@ void deepspeed_aio_thread_t::run()
{
std::unique_lock<std::mutex> lock(_work_sync._mutex);
_work_sync._cond_var.wait(lock,
[this] { return (!_work_queue.empty() || _time_to_exit); });
[this] { return (!_work_queue.empty() || _time_to_exit); });
if (!_work_queue.empty()) {
next_io_op = _work_queue.front();
_work_queue.pop();
Expand Down
2 changes: 1 addition & 1 deletion csrc/aio/py_lib/deepspeed_py_io_handle.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -148,7 +148,7 @@ std::shared_ptr<struct io_op_desc_t> deepspeed_io_handle_t::_wait_for_aio_work()
for (auto& ctxt : _thread_contexts) {
std::unique_lock<std::mutex> lock(ctxt->_complete_sync._mutex);
ctxt->_complete_sync._cond_var.wait(lock,
[ctxt] { return !ctxt->_complete_queue.empty(); });
[ctxt] { return !ctxt->_complete_queue.empty(); });
completed_op = ctxt->_complete_queue.front();
ctxt->_complete_queue.pop();
}
Expand Down
Loading