| { | |
| "_name_or_path": "mlsquare/130M_Seshu", | |
| "architectures": [ | |
| "MambaForCausalLM" | |
| ], | |
| "bias": false, | |
| "conv_bias": true, | |
| "d_conv": 4, | |
| "d_inner": 5120, | |
| "d_model": 2560, | |
| "d_state": 16, | |
| "dt_rank": 160, | |
| "expand": 2, | |
| "initializer_range": 0.02, | |
| "model_type": "mamba", | |
| "n_layer": 64, | |
| "pad_vocab_size_multiple": 8, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.38.1", | |
| "vocab_size": 20000 | |
| } | |