| { |
| "attention_qkv_bias": false, |
| "codebook_size": 2048, |
| "dim": 768, |
| "dropout": 0.1, |
| "fast_attention_qkv_bias": false, |
| "fast_dim": 768, |
| "fast_head_dim": 64, |
| "fast_intermediate_size": 3072, |
| "fast_n_head": 12, |
| "fast_n_local_heads": 4, |
| "head_dim": 64, |
| "initializer_range": 0.041666666666666664, |
| "intermediate_size": 3072, |
| "is_reward_model": false, |
| "max_seq_len": 2048, |
| "model_type": "dual_ar", |
| "n_fast_layer": 4, |
| "n_head": 12, |
| "n_layer": 10, |
| "n_local_heads": 4, |
| "depthwise_wte": true, |
| "depthwise_output": true, |
| "norm_eps": 1e-5, |
| "num_codebooks": 8, |
| "rope_base": 100000, |
| "scale_codebook_embeddings": false, |
| "share_codebook_embeddings": true, |
| "tie_word_embeddings": true, |
| "use_gradient_checkpointing": true, |
| "vocab_size": 2368 |
| } |
|
|