Buckets:
| [queue3] waiting for pid 8626 | |
| [queue3] qwen35 smoke test | |
| [queue3] qwen35 smoke ok: {'batch_size': 8, 'gradient_checkpointing': False, 'seconds_per_step': 1.3092960234498605, 'samples_per_second': 6.110153744239459, 'peak_allocated_gib': 56.76483774185181, 'peak_reserved_gib': 57.083984375, 'safe': True} est train s 6611 | |
| [queue3] qwen35 training | |
| [queue3] qwen35 exit 0 | |
| wandb: [wandb.login()] Loaded credentials for https://api.forge.coreweave.com from /root/.netrc. | |
| wandb: Currently logged in as: mrishikesh104 (storm-dpo) to https://api.forge.coreweave.com. Use `wandb login --relogin` to force relogin | |
| wandb: setting up run wmgfmd4i | |
| wandb: Tracking run with wandb version 0.31.0 | |
| wandb: Run data is saved locally in /workspace/wandb/run-20261105_111412-wmgfmd4i | |
| wandb: Run `wandb offline` to turn off syncing. | |
| wandb: Resuming run qwen3.5-0.8b-30k | |
| wandb: ⭐️ View project at https://forge.coreweave.com/wandb/storm-dpo/policy-classifier | |
| wandb: 🚀 View run at https://forge.coreweave.com/wandb/storm-dpo/policy-classifier/runs/wmgfmd4i | |
| wandb: Adding directory to artifact (checkpoints/qwen3.5-0.8b-30k/checkpoint-1of4-step158)... Done. 0.2s | |
| logged checkpoint-1of4-step158 | |
| wandb: Adding directory to artifact (checkpoints/qwen3.5-0.8b-30k/checkpoint-2of4-step316)... Done. 0.2s | |
| logged checkpoint-2of4-step316 | |
| wandb: Adding directory to artifact (checkpoints/qwen3.5-0.8b-30k/checkpoint-3of4-step474)... Done. 0.2s | |
| logged checkpoint-3of4-step474 | |
| wandb: Adding directory to artifact (checkpoints/qwen3.5-0.8b-30k/checkpoint-4of4-step632)... Done. 0.2s | |
| logged checkpoint-4of4-step632 | |
| wandb: Adding directory to artifact (checkpoints/qwen3.5-0.8b-30k/final)... Done. 0.2s | |
| wandb: WARNING Artifact "qwen3.5-0.8b-30k-adapter" already exists with the same content. No new version will be created. | |
| logged final | |
| logged evaluation | |
| wandb: updating run metadata | |
| wandb: uploading output.log; uploading wandb-summary.json; uploading config.yaml | |
| wandb: uploading output.log; uploading wandb-summary.json | |
| wandb: uploading summary, console lines 152-152 | |
| wandb: | |
| wandb: Run summary: | |
| wandb: eval/accuracy 0.87415 | |
| wandb: eval/followed_f1 0.85134 | |
| wandb: eval/followed_precision 0.83673 | |
| wandb: eval/followed_recall 0.86647 | |
| wandb: eval/followed_support 1041 | |
| wandb: eval/loss 0.39013 | |
| wandb: eval/macro_f1 0.87335 | |
| wandb: eval/not_applicable_f1 0.91036 | |
| wandb: eval/not_applicable_precision 0.90323 | |
| wandb: eval/not_applicable_recall 0.9176 | |
| wandb: +140 ... | |
| wandb: | |
| wandb: 🚀 View run qwen3.5-0.8b-30k at: https://forge.coreweave.com/wandb/storm-dpo/policy-classifier/runs/wmgfmd4i | |
| wandb: ⭐️ View project at: https://forge.coreweave.com/wandb/storm-dpo/policy-classifier | |
| wandb: Synced 5 W&B file(s), 0 media file(s), 24 artifact file(s) and 0 other file(s) | |
| wandb: Find logs at: ./wandb/run-20261105_111412-wmgfmd4i/logs | |
| [queue3] qwen35_2ep smoke test | |
| [queue3] qwen35_2ep smoke ok: {'batch_size': 8, 'gradient_checkpointing': False, 'seconds_per_step': 1.3060233689611778, 'samples_per_second': 6.125464666350702, 'peak_allocated_gib': 56.76483774185181, 'peak_reserved_gib': 57.083984375, 'safe': True} est train s 8792 | |
| [queue3] qwen35_2ep training | |
Xet Storage Details
- Size:
- 3.25 kB
- Xet hash:
- 1f19feadf98c4c1c3e5c755738f573b2f9cdd0adabc0a705b0ef8b53ae53199c
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.