ViperEkura
84ed2327f5
feat: add --resume flag to decouple weight loading from training resumption
...
- Add --resume bool flag to train.py CLI
- --param_path always loads weights only by default
- --resume restores epoch, consumed_samples, optimizer & scheduler
- Checkpoint.load() now preserves full meta dict
- Update test_early_stopping to use new param_path/resume API
2026-07-16 14:23:23 +08:00
ViperEkura
b092316385
feat : add distributed checkpoint via executor checkpoint_context
...
- Add checkpoint_context context manager to BaseExecutor with entry/exit barrier
- Add _gather_state_dict hook overridden per executor (template method)
- DDPExecutor skips unwrap on non-rank-0 to avoid redundant state_dict gather
- FSDPExecutor uses rank0_only=True to reduce memory on non-writers
- Remove redundant rank-0 guard from Checkpoint.save and manual barrier from Callback
2026-07-13 12:27:09 +08:00
ViperEkura
4af7acd449
fix: support single .h5 file loading in load_h5
2026-07-07 00:11:16 +08:00
ViperEkura
8999ca89b8
feat: add JSONL dataset store with on-the-fly tokenization
...
- Add JsonlStore registered under "jsonl" in astrai/dataset/storage.py
- Reuse PipelineConfig schema for JSONL dataset configuration
- Update detect_format to recognize JSONL directories and files
- Move save_h5/load_h5/save_bin/load_bin to astrai/serialization
- Split astrai/serialization.py into checkpoint/dataset submodules
- Add tests for JSONL detection, seq/SFT stores, and config roundtrip
2026-07-04 15:42:33 +08:00
ViperEkura