{ "gpu_hours": 47.786, "tokens_per_step": 262144, "attention": "flash_attention_2", "sparse_prediction": true, "final_step": 84974, "tokens_seen": 22275424256, "args": { "train": "/work/hdd/bbov/smeher/cb2/packed/train_1024_cb2full", "eval": "/work/hdd/bbov/smeher/cb2/packed/eval_random_1024_cb2full", "out": "/work/hdd/bbov/smeher/cb2/runs/conflibert-v2-cb2-wsd-full", "base": "/work/hdd/bbov/smeher/cb2/models/modernbert-stable-hf", "mlm_prob": 0.3, "lr": 0.0002, "bsz": 16, "accum": 4, "epochs": 1.0, "max_steps": -1, "warmup_ratio": 0.03, "weight_decay": 0.01, "scheduler": "wsd", "decay_ratio": 0.2, "save_steps": 500, "eval_steps": 1000, "eval_max_blocks": 8192, "log_steps": 25, "attn": "auto", "doc_attention": true, "dense_pred": false, "compile": true, "grad_checkpointing": false, "mem_fraction": 0.0, "new_embed_warmup_steps": 0, "new_embed_first_row": 50368, "lora": false, "lora_r": 16, "resume": false, "auto_resume": true, "report": "tensorboard", "run_name": "", "tb_root": "/work/hdd/bbov/smeher/cb2/tb" } }