| { | |
| "hidden_size": 512, | |
| "intermediate_size": 1024, | |
| "num_hidden_layers": 12, | |
| "num_attention_heads": 8, | |
| "num_key_value_heads": 4, | |
| "num_experts": 4, | |
| "num_experts_per_tok": 2, | |
| "progressive_experts": true, | |
| "min_experts": 1, | |
| "max_experts": 2, | |
| "use_hyper_connection": true, | |
| "use_compressed_attention": true, | |
| "compress_ratio": 4, | |
| "max_position_embeddings": 2048, | |
| "vocab_size": 50257, | |
| "bos_token_id": 50256, | |
| "eos_token_id": 50256, | |
| "rope_theta": 10000.0, | |
| "dropout": 0.1, | |
| "layer_norm_eps": 1e-05, | |
| "attention_dropout": 0.0, | |
| "use_flash_attention": true, | |
| "torch_dtype": "float16" | |
| } |