| { | |
| "architectures": [ | |
| "HybridMoRMoEForCausalLM" | |
| ], | |
| "complexity_hidden_dim": 64, | |
| "d_ff": 1536, | |
| "d_model": 576, | |
| "dropout": 0.05, | |
| "dtype": "float32", | |
| "eos_token_id": 151645, | |
| "max_recursions": 3, | |
| "max_seq_len": 4096, | |
| "model_size": "small", | |
| "model_type": "hybrid_mor_moe", | |
| "moe_aux_loss_coef": 0.0001, | |
| "n_heads": 8, | |
| "num_base_layers": 6, | |
| "num_experts": 4, | |
| "num_experts_per_tok": 1, | |
| "num_hidden_layers": 30, | |
| "num_recursions": 3, | |
| "num_shared_blocks": 6, | |
| "num_unique_last_layers": 2, | |
| "pad_token_id": 151643, | |
| "router_aux_loss_coef": 0.0001, | |
| "router_percentile": 0.7, | |
| "transformers_version": "5.4.0", | |
| "use_cache": false, | |
| "vocab_size": 151665 | |
| } | |