| { | |
| "_name_or_path": "mamba-1.4b-hf-supernet/last", | |
| "architecture": { | |
| "layer_1": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_10": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_11": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_12": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_13": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_14": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_15": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_16": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_17": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_18": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_19": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_2": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_20": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_21": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_22": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_23": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_24": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_25": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_26": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_27": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_28": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_29": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_3": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_30": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_31": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_32": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_33": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_34": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_35": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_36": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_37": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_38": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_39": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_4": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_40": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_41": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_42": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_43": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_44": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_45": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_46": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_47": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_48": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_5": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_6": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_7": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_8": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| }, | |
| "layer_9": { | |
| "atten_out": "None", | |
| "inter_hidden": 1280, | |
| "residual_hidden": "None" | |
| } | |
| }, | |
| "architectures": [ | |
| "MambaForCausalLM" | |
| ], | |
| "bos_token_id": 0, | |
| "conv_kernel": 4, | |
| "d_model": 2048, | |
| "elastic_config": { | |
| "atten_out_space": [ | |
| "None" | |
| ], | |
| "inter_hidden_space": [ | |
| 4096, | |
| 1280 | |
| ], | |
| "residual_hidden_space": [ | |
| "None" | |
| ] | |
| }, | |
| "eos_token_id": 0, | |
| "expand": 2, | |
| "fused_add_norm": true, | |
| "hidden_act": "silu", | |
| "hidden_size": 2048, | |
| "initializer_range": 0.1, | |
| "intermediate_size": 4096, | |
| "layer_norm_epsilon": 1e-05, | |
| "model_type": "mamba", | |
| "n_layer": 48, | |
| "num_hidden_layers": 48, | |
| "pad_token_id": 0, | |
| "pad_vocab_size_multiple": 8, | |
| "rescale_prenorm_residual": false, | |
| "residual_in_fp32": true, | |
| "rms_norm": true, | |
| "ssm_cfg": {}, | |
| "state_size": 16, | |
| "time_step_floor": 0.0001, | |
| "time_step_init_scheme": "random", | |
| "time_step_max": 0.1, | |
| "time_step_min": 0.001, | |
| "time_step_rank": 128, | |
| "time_step_scale": 1.0, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.40.0.dev0", | |
| "use_bias": false, | |
| "use_cache": true, | |
| "use_conv_bias": true, | |
| "vocab_size": 50280 | |
| } | |