| {"step": 0, "acc_step": 0, "data_loader_state": {"it_state": {"start_token": 0, "it_state": {"it_state": {"root_dir": "/scratch/craffel/lingua/data/", "sources": {"fineweb_edu_10bt_shuffled": 1.0}, "source_to_state": {"fineweb_edu_10bt_shuffled": {"file_path": "/scratch/craffel/lingua/data/fineweb_edu_10bt_shuffled/fineweb_edu_10bt.chunk.38.jsonl", "position": 0, "block_size": 1, "offset": 0, "current_iter": 0}}, "rng_state": {"bit_generator": "PCG64", "state": {"state": 232471632667331820145564819885404856715, "inc": 95963489890761403814531195999220475639}, "has_uint32": 0, "uinteger": 0}}, "add_bos": true, "add_eos": true, "name": "bytes", "path": null}, "output_seq_len": 4096, "n_views": 2, "seq_len": 0}, "seq_idx": 0, "rng_state": {"bit_generator": "PCG64", "state": {"state": 169925910427912386056949567841398549015, "inc": 72545526324180839152750112646078969085}, "has_uint32": 0, "uinteger": 0}, "batch_size": 4, "prefetch_size": 1024}, "scheduler": {"base_lrs": [0.003], "last_epoch": 0, "verbose": false, "_step_count": 1, "_get_lr_called_within_step": false, "_last_lr": [0.0], "lr_lambdas": [{}]}} |