| { | |
| "architectures": [ | |
| "MeridianForCausalLM" | |
| ], | |
| "bos_token_id": 151643, | |
| "dtype": "float32", | |
| "eos_token_id": 151645, | |
| "ewc_lambda": 100.0, | |
| "expert_intermediate_size": 896, | |
| "hidden_size": 768, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 1792, | |
| "max_position_embeddings": 2048, | |
| "model_type": "meridian", | |
| "moe_layer_frequency": 2, | |
| "num_attention_heads": 12, | |
| "num_experts": 8, | |
| "num_experts_per_token": 2, | |
| "num_key_value_heads": 4, | |
| "num_layers": 14, | |
| "numeracy_embed_dim": 64, | |
| "pad_token_id": 151643, | |
| "rms_norm_eps": 1e-06, | |
| "rope_theta": 500000.0, | |
| "router_aux_loss_coef": 0.01, | |
| "tie_word_embeddings": true, | |
| "transformers_version": "5.1.0", | |
| "use_cache": true, | |
| "use_numeracy_encoding": true, | |
| "vocab_size": 151665 | |
| } | |