unsloth/studio/backend/core/training
Daniel Han a515fbfed3 Version the conditioning cache key and reject non-finite flow_shift
Two correctness fixes:

- The cache keyed the checkpoint and its companion base by name only, so
  a Hub repo advancing to a new commit, or a local directory updated in
  place, kept returning embeddings from the previous text encoder. Pair
  both with a revision marker: the locally resolved commit sha for a Hub
  repo, config plus text-encoder file stats for a directory. Neither
  loads the encoders, so a warm run still keeps them off the GPU.
- flow_shift only checked positivity, but JSON accepts 1e309, which
  floats to inf, and inf <= 0 is False while NaN fails every comparison.
  The sigma table then evaluates s * u / (1 + (s - 1) * u) as NaN, which
  poisons every sampled sigma and saves a corrupted adapter while
  progress looks normal. Require a finite value.
2026-07-26 10:51:01 +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 [pre-commit.ci] auto fixes from pre-commit.com hooks 2026-07-25 09:00:00 +00:00
diffusion_lora_trainer.py Tighten comments across the remaining image stack files 2026-07-12 11:46:23 +00:00
diffusion_train_common.py Version the conditioning cache key and reject non-finite flow_shift 2026-07-26 10:51:01 +00:00
diffusion_train_extras.py [pre-commit.ci] auto fixes from pre-commit.com hooks 2026-07-25 09:00:00 +00:00
diffusion_training_service.py Tighten comments across the remaining image stack files 2026-07-12 11:46:23 +00:00
fsdp2_design_notes.md Tighten torchao configs and note the FSDP2 design for the DiT trainer 2026-07-20 07:22:49 +00:00
resume.py Fix resume training crash recovery and MLX checkpoints (#6796) 2026-07-21 02:34:58 -07:00
s3_dataset.py feat(studio): implement S3 dataset loading (completes #5951) (#6222) 2026-06-12 14:52:04 +02:00
trainer.py feat(studio): add DoRA support to studio (#7315) 2026-07-24 03:24:16 -07:00
training.py Fix training start NameError, the load-order guard test and CPU-only diffusion tests 2026-07-25 19:18:35 -07:00
worker.py feat(studio): add DoRA support to studio (#7315) 2026-07-24 03:24:16 -07:00