6 Commits

Author	SHA1	Message	Date
Dan Saunders	10ba1622f7	checkpoint model on first step callback (#2906) ... * checkpoint model on first step callback * remove debug * add test cases; update existing tests not to save on first step * move test out of solo * delete * default to False * typo	2025-07-15 15:00:48 -04:00
Wing Lian	81893c775c	Accelerate 1.8.1 and BNB 0.46.0 update (#2815) ... * update accelerate to v1.8.0 * update bnb also * fix multigpu ci timeout * fix test set size * use latest accelerate 1.8.1 * disable default dtype	2025-06-28 15:29:19 -04:00
Wing Lian	a85efffbef	bump transformers==4.52.4 (#2800) [skip ci] ... * bump transformers==4.52.4 * don't use hf offline for qwen tokenizer * increase timeout * don't use methodtype * increase timeout * better assertion logging * upgrade deepspeed version too	2025-06-18 15:46:14 -04:00
Dan Saunders	1d91d905c9	remove deprecated wandb env var (#2751) ... * remove deprecated wandb env var * remove os.environ wandb setting; unused loggers * remove os.environ wandb setting; unused loggers	2025-06-03 14:04:15 -07:00
salman	65c5481120	Rank 0-only logging (#2608) ... Co-authored-by: Wing Lian <wing@axolotl.ai>	2025-05-28 14:57:30 +01:00
Wing Lian	328d598114	gemma3 packing fixes (#2449) ... * make gemma3 work with packing * multi-gpu e2e for ci * update gemma3 model namespace to use mirror * add gradient checkpointing to multigpu e2e ci * update gemma3 examples for use_reentrant and fix ddp find unused params * fix tests for gemma3 * fix import for test utils * set correct train loss for gemma3 e2e	2025-03-31 17:15:23 -04:00