| { |
| "model_type": "tolm", |
| "architectures": [ |
| "TOLMForMaskedLM" |
| ], |
| "auto_map": { |
| "AutoConfig": "tolm.TOLMConfig", |
| "AutoModel": "tolm.TOLMModel", |
| "AutoModelForMaskedLM": "tolm.TOLMForMaskedLM", |
| "AutoModelForCausalLM": "tolm.TOLMForCausalLM" |
| }, |
| "vocab_size": 16384, |
| "max_seq_len": 512, |
| "hidden_size": 384, |
| "num_hidden_layers": 12, |
| "num_attention_heads": 6, |
| "intermediate_size": 1280, |
| "position_buckets": 32, |
| "dropout": 0.1, |
| "attention_dropout": 0.1, |
| "initializer_range": 0.03227486121839514, |
| "value_gating": true, |
| "residual_mixing": true, |
| "pad_token_id": 3, |
| "bos_token_id": 1, |
| "eos_token_id": 2, |
| "mask_token_id": 4, |
| "absolute_positions": false, |
| "use_rope": false, |
| "use_alibi": false, |
| "recurrent_steps": 1, |
| "num_experts": 1, |
| "experts_per_token": 1, |
| "expert_intermediate_size": null, |
| "future_offsets": [], |
| "state_mixer_kernel": 0, |
| "rtd_auxiliary": true, |
| "geometry_lexical_dim": 0, |
| "geometry_curvature": 1.0, |
| "cognitive_readout_layer": 0, |
| "cognitive_readout_weight": 0.0, |
| "direct_sum_dims": [], |
| "direct_sum_heads": [], |
| "direct_sum_intermediate_sizes": [], |
| "lexical_residual_buckets": 0, |
| "lexical_residual_dim": 0, |
| "lexical_residual_scale": 1.0, |
| "structured_projection_dim": 0 |
| } |
|
|
