| { | |
| "architectures": [ | |
| "Pebble10MLM" | |
| ], | |
| "model_type": "pebble_10m", | |
| "auto_map": { | |
| "AutoConfig": "configuration_pebble.PebbleConfig", | |
| "AutoModelForCausalLM": "modeling_pebble.PebbleForCausalLM" | |
| }, | |
| "vocab_size": 2048, | |
| "hidden_size": 384, | |
| "intermediate_size": 1536, | |
| "num_hidden_layers": 8, | |
| "num_attention_heads": 6, | |
| "block_pattern": "mmma|mmma", | |
| "hybrid_ratio": "3:1 mamba2:attention", | |
| "max_position_embeddings": 512, | |
| "rms_norm_eps": 1e-06, | |
| "tie_word_embeddings": true, | |
| "mamba2": { | |
| "d_state": 128, | |
| "d_conv": 4, | |
| "expand": 2, | |
| "headdim": 96, | |
| "use_mem_eff_path": true | |
| }, | |
| "attention": { | |
| "rope_theta": 10000.0, | |
| "is_causal": true | |
| }, | |
| "torch_dtype": "float32" | |
| } |