| { |
| "architectures": [ |
| "ZambaForCausalLM" |
| ], |
| "attention_dropout": 0.0, |
| "attention_head_dim": 32, |
| "attention_hidden_size": 128, |
| "attention_probs_dropout_prob": 0.1, |
| "attn_layer_offset": 1, |
| "attn_layer_period": 8, |
| "bos_token_id": 1, |
| "dtype": "float32", |
| "eos_token_id": 2, |
| "hidden_act": "gelu", |
| "hidden_dropout_prob": 0.1, |
| "hidden_mamba_act": "silu", |
| "hidden_size": 64, |
| "initializer_range": 0.02, |
| "intermediate_size": 37, |
| "is_decoder": true, |
| "layers_block_type": [ |
| "linear_attention", |
| "linear_attention", |
| "hybrid", |
| "linear_attention", |
| "hybrid" |
| ], |
| "mamba_conv_bias": true, |
| "mamba_d_conv": 4, |
| "mamba_d_state": 16, |
| "mamba_dt_rank": 32, |
| "mamba_expand": 2, |
| "mamba_proj_bias": false, |
| "max_position_embeddings": 512, |
| "model_type": "zamba", |
| "n_mamba_heads": 2, |
| "num_attention_heads": 4, |
| "num_hidden_layers": 5, |
| "num_key_value_heads": 4, |
| "num_logits_to_keep": 1, |
| "pad_token_id": 0, |
| "rms_norm_eps": 1e-05, |
| "tie_word_embeddings": true, |
| "time_step_floor": 0.0001, |
| "time_step_max": 0.1, |
| "time_step_min": 0.001, |
| "transformers_version": "5.16.0.dev0", |
| "type_vocab_size": 16, |
| "use_cache": true, |
| "use_mamba_kernels": false, |
| "vocab_size": 32001 |
| } |
|
|
