| { | |
| "MIloss_lambda": 0.01, | |
| "architectures": [ | |
| "MultiSpeciesEnformer" | |
| ], | |
| "attn_dim_key": 64, | |
| "attn_dropout": 0.05, | |
| "cvloss_lambda": 0.005, | |
| "depth": 11, | |
| "dim": 1536, | |
| "dim_divisible_by": 128, | |
| "dropout_rate": 0.4, | |
| "heads": 8, | |
| "model_type": "enformer", | |
| "moe": [ | |
| "species", | |
| "tracks" | |
| ], | |
| "num": 1024, | |
| "num_downsamples": 7, | |
| "output_heads": { | |
| "human": 5313, | |
| "mouse": 1643 | |
| }, | |
| "pos_dropout": 0.01, | |
| "seq_length": 131072, | |
| "species_num_experts": 4, | |
| "target_length": 896, | |
| "topk": 3, | |
| "torch_dtype": "float32", | |
| "tracks_num_experts": 8, | |
| "transformers_version": "4.45.2", | |
| "use_checkpointing": false, | |
| "use_tf_gamma": false, | |
| "zloss_lambda": 0.001 | |
| } | |