You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
- Added --resume/-r flag to remote_train.sh for continuing interrupted training
- Script now passes resume mode through SSH to remote server
- Updated README with resume documentation for remote training
- Supports combining --kill and --resume flags for restart scenarios
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Copy file name to clipboardExpand all lines: README.md
+7-1Lines changed: 7 additions & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -238,15 +238,21 @@ Once Git credentials are configured on your server, run `remote_train.sh` **from
238
238
# From your local machine, start training on the remote GPU server
239
239
./remote_train.sh
240
240
241
+
# Resume training from existing checkpoints
242
+
./remote_train.sh --resume # or -r
243
+
241
244
# Kill existing training sessions and optionally start new one
242
245
./remote_train.sh --kill # or -k
243
246
247
+
# Kill and resume (restart interrupted training)
248
+
./remote_train.sh --kill --resume
249
+
244
250
# You'll be prompted for:
245
251
# - Server address (hostname or IP)
246
252
# - Username
247
253
```
248
254
249
-
**What this script does:** The `remote_train.sh` script connects to your GPU server via SSH and executes `run_llm_stylometry.sh --train -y` in a `screen` session. This allows you to disconnect your local machine while the GPU server continues training.
255
+
**What this script does:** The `remote_train.sh` script connects to your GPU server via SSH and executes `run_llm_stylometry.sh --train -y`(or `--train --resume -y` if resuming) in a `screen` session. This allows you to disconnect your local machine while the GPU server continues training.
0 commit comments