unsloth/studio/backend/core/training
Daniel Han a091a862df Add krea-2 to bf16-only preflight; match dataset stems case-insensitively
Two evict/corruption fixes surfaced by review of the diffusion training path:

- krea-2 sets force_bf16 in the DiT trainer spec, but the route-level
  _FORCE_BF16_FAMILIES preflight listed only qwen-image and z-image, so a
  krea-2 start with mixed_precision=fp16 passed the route check, reserved
  training and evicted resident GPU models, and only the child trainer then
  raised. Add krea-2 to the set and a drift-guard test asserting it equals
  the trainer specs whose force_bf16 is set.

- The dataset-upload same-stem duplicate check compared stems
  case-sensitively, so on case-insensitive filesystems (Windows / default
  macOS) sample.png and Sample.jpg both passed even though their caption
  sidecars sample.txt / Sample.txt resolve to the same file, silently
  sharing and corrupting one caption. Compare stems and the same-name guard
  with casefold at both the on-disk and in-batch sites.
2026-07-07 18:07:48 +00:00
..
__init__.py Reduce and tighten code comments and docstrings repo-wide (#6095) 2026-06-08 23:09:51 -07:00
diffusion_dit_trainer.py Merge remote-tracking branch 'origin/diffusion-train-perf2' into fold-integration 2026-07-07 01:06:42 +00:00
diffusion_lora_trainer.py Merge branch 'diffusion-krea2' into diffusion-train-perf2 2026-07-05 07:56:51 +00:00
diffusion_train_common.py Add krea-2 to bf16-only preflight; match dataset stems case-insensitively 2026-07-07 18:07:48 +00:00
diffusion_training_service.py Make the diffusion-training reservation a compare-and-set 2026-07-07 11:57:37 +00:00
resume.py feat(studio): implement S3 dataset loading (completes #5951) (#6222) 2026-06-12 14:52:04 +02:00
s3_dataset.py feat(studio): implement S3 dataset loading (completes #5951) (#6222) 2026-06-12 14:52:04 +02:00
trainer.py Fix Windows Studio UTF-8 startup handling (#6614) 2026-07-01 13:47:33 +01:00
training.py (feat) Add project names to studio training runs (#6512) 2026-06-29 16:06:36 +02:00
worker.py Add MLX-aware public Unsloth trainer API (#6462) 2026-07-02 23:02:26 +01:00