CK0607's picture
Round 3 3B domain expansion results
4838960 verified
raw
history blame
684 Bytes
[2026-05-04 20:39:46] Dataset audit kept=3 dropped=[] domain_counts={'math': 1, 'code': 1, 'science': 1}
[2026-05-04 20:39:46] Launching 6 LoRA trainings across 3 workers
[2026-05-04 20:39:51] [SKIP_TRAIN_EXISTS] meta-llama/Llama-3.2-3B-Instruct / gsm8k
[2026-05-04 20:39:51] [SKIP_TRAIN_EXISTS] Qwen/Qwen2.5-3B-Instruct / gsm8k
[2026-05-04 20:39:51] [SKIP_TRAIN_EXISTS] Qwen/Qwen2.5-3B-Instruct / arc_easy
[2026-05-04 20:39:51] [SKIP_TRAIN_EXISTS] meta-llama/Llama-3.2-3B-Instruct / mbpp
[2026-05-04 20:39:51] [SKIP_TRAIN_EXISTS] Qwen/Qwen2.5-3B-Instruct / mbpp
[2026-05-04 20:39:51] [SKIP_TRAIN_EXISTS] meta-llama/Llama-3.2-3B-Instruct / arc_easy
[2026-05-04 20:39:52] [SMOKE_DONE]