| { |
| "architectures": [ |
| "PonderMoELlamaForCausalLM" |
| ], |
| "attention_bias": false, |
| "attention_dropout": 0.0, |
| "bos_token_id": 0, |
| "commit_kv": true, |
| "drift_threshold": 0.98, |
| "dtype": "float16", |
| "eos_token_id": [ |
| 1, |
| 130073 |
| ], |
| "head_dim": 128, |
| "hidden_act": "silu", |
| "hidden_size": 2048, |
| "initializer_range": 0.02, |
| "intermediate_size": 6144, |
| "max_ponder_steps": 8, |
| "max_position_embeddings": 131072, |
| "mlp_bias": false, |
| "model_type": "llama", |
| "moe_aux_weight": 0.01, |
| "moe_num_experts": 8, |
| "moe_route_tau": null, |
| "moe_scope": "ponder", |
| "moe_tau": 0.125, |
| "mtp_loss_weight": 1.0, |
| "mtp_num_tokens": 2, |
| "num_attention_heads": 16, |
| "num_hidden_layers": 42, |
| "num_key_value_heads": 2, |
| "pad_token_id": 1, |
| "ponder_all_positions": false, |
| "ponder_epsilon": 0.01, |
| "ponder_forced_steps": 0, |
| "ponder_head_hidden": 0, |
| "ponder_loss_beta": 0.01, |
| "ponder_prior_p": 0.5, |
| "ponder_signal": "learned", |
| "ponder_start_layer": 34, |
| "ponder_step_ce_weight": 0.0, |
| "ponder_window": 1, |
| "pretraining_tp": 1, |
| "rms_norm_eps": 1e-06, |
| "rope_parameters": { |
| "rope_theta": 5000000, |
| "rope_type": "default" |
| }, |
| "tie_word_embeddings": false, |
| "transformers_version": "5.16.1", |
| "use_cache": true, |
| "vocab_size": 130560, |
| "auto_map": { |
| "AutoConfig": "modeling_samai_pnet.SamaiPnetConfig", |
| "AutoModelForCausalLM": "modeling_samai_pnet.SamaiPnetForCausalLM" |
| } |
| } |
