| { | |
| "attention_o_bias": false, | |
| "attention_qk_norm": true, | |
| "attention_qkv_bias": false, | |
| "codebook_size": 4096, | |
| "dim": 1024, | |
| "dropout": 0.0, | |
| "fast_attention_o_bias": false, | |
| "fast_attention_qk_norm": false, | |
| "fast_attention_qkv_bias": false, | |
| "fast_dim": 1024, | |
| "fast_head_dim": 64, | |
| "fast_intermediate_size": 3072, | |
| "fast_n_head": 16, | |
| "fast_n_local_heads": 8, | |
| "head_dim": 128, | |
| "initializer_range": 0.03125, | |
| "intermediate_size": 3072, | |
| "max_seq_len": 8192, | |
| "model_type": "dual_ar", | |
| "n_fast_layer": 4, | |
| "n_head": 16, | |
| "n_layer": 28, | |
| "n_local_heads": 8, | |
| "norm_eps": 1e-06, | |
| "num_codebooks": 10, | |
| "rope_base": 1000000, | |
| "scale_codebook_embeddings": true, | |
| "tie_word_embeddings": false, | |
| "use_gradient_checkpointing": true, | |
| "vocab_size": 155776 | |
| } |