refactor : replace iteration with consumed_samples
- Replace context.iteration with consumed_samples (global sample count) - Add optimizer_step property derived from consumed_samples - Checkpoint meta.json stores consumed_samples, drops iteration - CLI --start_batch renamed to --start_samples (per-rank samples) - Checkpoint dir naming: epoch_X_step_Y instead of epoch_X_iter_Y - Metric log entries use step and consumed_samples fields - Backward compat removed (old iteration checkpoints unsupported)
This commit is contained in:
@@ -175,7 +175,10 @@ def parse_args() -> argparse.Namespace:
|
||||
"--start_epoch", type=int, default=0, help="Start epoch for training."
|
||||
)
|
||||
parser.add_argument(
|
||||
"--start_batch", type=int, default=0, help="Start batch for training."
|
||||
"--start_samples",
|
||||
type=int,
|
||||
default=0,
|
||||
help="Start samples (per rank) for training.",
|
||||
)
|
||||
|
||||
parser.add_argument(
|
||||
@@ -317,7 +320,7 @@ def train(
|
||||
n_epoch: int,
|
||||
batch_per_device: int,
|
||||
start_epoch: int,
|
||||
start_batch: int,
|
||||
start_samples: int,
|
||||
grad_accum_steps: int,
|
||||
warmup_ratio: float,
|
||||
ckpt_interval: int,
|
||||
@@ -444,7 +447,7 @@ def train(
|
||||
n_epoch=n_epoch,
|
||||
batch_per_device=batch_per_device,
|
||||
start_epoch=start_epoch,
|
||||
start_batch=start_batch,
|
||||
start_samples=start_samples,
|
||||
ckpt_interval=ckpt_interval,
|
||||
grad_accum_steps=grad_accum_steps,
|
||||
max_grad_norm=max_grad_norm,
|
||||
|
||||
Reference in New Issue
Block a user