Tobi Lütke 8cc7d8c138 Add sampled /only: variants (399) for training balance пре 3 месеци
..
grpo.yaml 354744af53 Finetune 2.0: consolidate and simplify the entire training pipeline пре 3 месеци
sft.yaml 8572c2fd94 Deploy fine-tuned GRPO model as default for query expansion пре 3 месеци
sft_v4.yaml 8cc7d8c138 Add sampled /only: variants (399) for training balance пре 3 месеци