Skip to content

New hyperparameter ckpt_save_weights_file (ref #26) #27

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.

Sign up for GitHub

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

Open
wants to merge 1 commit into
base: develop
Choose a base branch
from
Open

New hyperparameter ckpt_save_weights_file (ref #26) #27

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
44 changes: 25 additions & 19 deletions candle/ckpt_utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -209,6 +209,9 @@ def scan_params(self, gParams):
self.save_weights_only = self.param(
"ckpt_save_weights_only", True, ParamType.BOOLEAN
)
self.save_weights_file = self.param(
"ckpt_save_weights_file", True, ParamType.BOOLEAN
)
self.checksum_enabled = self.param("ckpt_checksum", False, ParamType.BOOLEAN)
self.keep_mode = self.param(
"ckpt_keep_mode",
Expand Down Expand Up @@ -349,23 +352,25 @@ def write_model(self, dir_work, epoch):
Do the I/O, report stats
dir_work: A pathlib.Path
"""
self.model_file = dir_work / "model.h5"
self.debug("writing model to: '%s'" % self.relpath(self.model_file))
start = time.time()

# Call down to backend-specific model writer:
self.write_model_backend(self.model, epoch)

stop = time.time()
duration = stop - start
stats = os.stat(self.model_file)
MB = stats.st_size / (1024 * 1024)
rate = MB / duration
self.debug(
"model wrote: %0.3f MB in %0.3f seconds (%0.2f MB/s)."
% (MB, duration, rate)
)
self.checksum(dir_work)
if self.save_weights_file:
self.model_file = dir_work / "model.h5"
self.debug("writing model to: '%s'" %
self.relpath(self.model_file))
start = time.time()

# Call down to backend-specific model writer:
self.write_model_backend(self.model, epoch)

stop = time.time()
duration = stop - start
stats = os.stat(self.model_file)
MB = stats.st_size / (1024 * 1024)
rate = MB / duration
self.debug(
"model wrote: %0.3f MB in %0.3f seconds (%0.2f MB/s)."
% (MB, duration, rate)
)
self.checksum(dir_work)
self.write_json(dir_work / "ckpt-info.json", epoch)

def checksum(self, dir_work):
Expand All @@ -387,8 +392,9 @@ def write_json(self, jsonfile, epoch):
D["epoch"] = epoch
D["save_best_metric"] = self.save_best_metric
D["best_metric_last"] = self.best_metric_last
D["model_file"] = "model.h5"
D["checksum"] = self.cksum_model
if self.save_weights_file:
D["model_file"] = "model.h5"
D["checksum"] = self.cksum_model
D["timestamp"] = now.strftime("%Y-%m-%d %H:%M:%S")
if self.timestamp_last is None:
time_elapsed = "__FIRST__"
Expand Down
7 changes: 7 additions & 0 deletions candle/parsing_utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -187,6 +187,7 @@ class ConfigDict(TypedDict):
ckpt_save_best: bool
ckpt_save_best_metric: str
ckpt_save_weights_only: bool
ckpt_save_weights_file: bool
ckpt_save_interval: int
ckpt_keep_mode: str
ckpt_keep_limit: int
Expand Down Expand Up @@ -612,6 +613,12 @@ class ConfigDict(TypedDict):
"default": False,
"help": "Toggle saving only weights (not optimizer) (NYI).",
},
{
"name": "ckpt_save_weights_file",
"type": str2bool,
"default": True,
"help": "Toggle whether to save weights. JSON will still be saved.",
},
{
"name": "ckpt_save_interval",
"type": int,
Expand Down