| { | |
| "architectures": [ | |
| "TransformerForCausalLM" | |
| ], | |
| "model_type": "transformer", | |
| "vocab_size": 5000, | |
| "max_position_embeddings": 128, | |
| "hidden_size": 128, | |
| "num_attention_heads": 8, | |
| "num_hidden_layers": 2, | |
| "intermediate_size": 512, | |
| "hidden_act": "gelu", | |
| "initializer_range": 0.02, | |
| "layer_norm_eps": 1e-05, | |
| "use_cache": true, | |
| "pad_token_id": 3, | |
| "bos_token_id": 1, | |
| "eos_token_id": 2, | |
| "num_experts": 4, | |
| "top_k_experts": 2 | |
| } | |