1experiment:2 seed: 13 model: configs/prisme_103m.yaml4 model_overrides:5 readout_zero_init: false6 readout_init_std: 0.27 ctrl_input: h8 ctrl_use_entropy: true9 init:10 backbone_checkpoint: experiments/runs/pretrain103m_seed01_20261029232706/checkpoint11 freeze_backbone: true12 auto_readout_scale: true13 data:14 kind: tokens15 train_bin: data/processed/fineweb_edu_300m/train.bin16 val_bin: data/processed/fineweb_edu_300m/val.bin17 budgets:18 - 019 - 120 - 221 - 422 - 823 budget_probs:24 - 0.225 - 0.226 - 0.227 - 0.228 - 0.229 budget_per: position30 train:31 batch_size: 1632 seq_len: 51233 steps: 300034 lr: 0.000335 warmup: 10036 weight_decay: 0.137 grad_clip: 1.038 log_every: 2039 eval_every: 50040 eval_batches: 1041 ckpt_every: 100042 final_exact_match: false43 name: editor_pilot_103m_lr3e444 mode: prisme45model:46 name: prisme_103m47 vocab_size: 3276848 d_model: 76849 n_layers: 1250 n_heads: 1251 n_kv_heads: 352 ffn_width: 204853 max_seq_len: 409654 rope_theta: 10000.055 norm_eps: 1.0e-0556 tie_embeddings: true57 init_std: 0.0258 use_workspace: true59 w: 25660 ffn_w: 76861 n_slots: 1662 n_editor_blocks: 263 n_editor_heads: 464 r_max: 865 eta_max: 0.2566 clip_c: 1.067 clip_eps: 1.0e-0668 readout_zero_init: false69 readout_init_std: 0.270 readout_scale: 1.071 controller_hidden: 072 ctrl_input: h73 ctrl_use_entropy: true74 lambda_cost: 0.075 kappa_risk: 0.076 dense_writes: false77resolved_train:78 batch_size: 1679 seq_len: 51280 steps: 300081 lr: 0.000382 warmup: 10083 weight_decay: 0.184 grad_clip: 1.085 log_every: 2086 eval_every: 50087 eval_batches: 1088 ckpt_every: 100089