-
Notifications
You must be signed in to change notification settings - Fork 31
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
Geshen/async mcts #101
base: yi/search_benchmark
Are you sure you want to change the base?
Geshen/async mcts #101
Conversation
Signed-off-by: Gerald Shen <geshen@nvidia.com> fix Signed-off-by: Gerald Shen <geshen@nvidia.com>
Signed-off-by: Gerald Shen <geshen@nvidia.com>
Signed-off-by: Gerald Shen <geshen@nvidia.com>
for more information, see https://pre-commit.ci
Signed-off-by: Gerald Shen <geshen@nvidia.com>
Signed-off-by: Gerald Shen <geshen@nvidia.com>
for more information, see https://pre-commit.ci
for more information, see https://pre-commit.ci
Signed-off-by: Gerald Shen <geshen@nvidia.com>
for more information, see https://pre-commit.ci
Signed-off-by: Yi Dong <yidong@nvidia.com>
Signed-off-by: Gerald Shen <geshen@nvidia.com>
Signed-off-by: Gerald Shen <geshen@nvidia.com>
Signed-off-by: Gerald Shen <geshen@nvidia.com>
Signed-off-by: Gerald Shen <geshen@nvidia.com>
Signed-off-by: Gerald Shen <geshen@nvidia.com>
Signed-off-by: Gerald Shen <geshen@nvidia.com>
Signed-off-by: Gerald Shen <geshen@nvidia.com>
for more information, see https://pre-commit.ci
Signed-off-by: Gerald Shen <geshen@nvidia.com>
…to geshen/async_mcts
return_value_memory.append((list(spg.value_memory), spg.data_id, backup_root_states[i])) | ||
del parallel_searches[i] | ||
del backup_root_states[i] | ||
if is_terminal: |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
indentation error?
Signed-off-by: Gerald Shen <geshen@nvidia.com>
Signed-off-by: Gerald Shen <geshen@nvidia.com>
cfg.model, | ||
trainer, | ||
strict=True, | ||
load_base_model_only=True, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
will it cause a problem that the second time we load the improved hybrid network, it won't load the value head weights?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
the second time we load the value head, PTL will overwrite this checkpoint that we loaded, so it should be pretty safe
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
okay. so you are loading from ckpt not nemo files
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
yeah, so when we have no checkpoint it loads from the .nemo file, when there is a checkpoint it loads this .nemo file but PTL will overwrite it
Signed-off-by: Gerald Shen <geshen@nvidia.com>
Signed-off-by: Gerald Shen <geshen@nvidia.com>
Signed-off-by: Gerald Shen <geshen@nvidia.com>
Signed-off-by: Gerald Shen <geshen@nvidia.com>
011e5f3
to
4ac5c9b
Compare
Signed-off-by: Gerald Shen <geshen@nvidia.com>
DRAFT tracking only