Instructions to use FIM4Science/fim-imp-pointwise-base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use FIM4Science/fim-imp-pointwise-base with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("feature-extraction", model="FIM4Science/fim-imp-pointwise-base", trust_remote_code=True)# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("FIM4Science/fim-imp-pointwise-base", trust_remote_code=True, dtype="auto") - Notebooks
- Google Colab
- Kaggle
| { | |
| "auto_map": { | |
| "AutoConfig": "fim.models.imputation_pointwise.FIMImpPointBaseConfig", | |
| "AutoModel": "fim.models.imputation_pointwise.FIMImpPointBase" | |
| }, | |
| "architectures": [ | |
| "FIMImpPointBase" | |
| ], | |
| "branch_net": { | |
| "dim_model": 512, | |
| "dim_time": 512, | |
| "dropout": 0.1, | |
| "name": "fim.models.blocks.Transformer", | |
| "num_encoder_blocks": 4, | |
| "num_heads": 8, | |
| "residual_mlp": { | |
| "dropout": 0.1, | |
| "hidden_act": { | |
| "name": "torch.nn.SELU" | |
| }, | |
| "hidden_layers": [ | |
| 1024 | |
| ], | |
| "in_features": 512, | |
| "name": "fim.models.blocks.MLP", | |
| "out_features": 512, | |
| "output_act": { | |
| "name": "torch.nn.Identity" | |
| } | |
| } | |
| }, | |
| "combiner_net": { | |
| "dropout": 0.1, | |
| "hidden_act": { | |
| "name": "torch.nn.SELU" | |
| }, | |
| "hidden_layers": [ | |
| 1024, | |
| 1024, | |
| 1024, | |
| 1024 | |
| ], | |
| "in_features": 1024, | |
| "name": "fim.models.blocks.MLP", | |
| "out_features": 512, | |
| "output_act": { | |
| "name": "torch.nn.Identity" | |
| } | |
| }, | |
| "init_cond_net": { | |
| "dropout": 0.1, | |
| "hidden_act": { | |
| "name": "torch.nn.SELU" | |
| }, | |
| "hidden_layers": [ | |
| 1024, | |
| 1024, | |
| 1024, | |
| 1024 | |
| ], | |
| "in_features": 513, | |
| "name": "fim.models.blocks.MLP", | |
| "out_features": 2, | |
| "output_act": { | |
| "name": "torch.nn.Identity" | |
| } | |
| }, | |
| "load_in_8bit": false, | |
| "loss_configs": { | |
| "loss_scale_drift": 1.0, | |
| "loss_scale_init_cond": 1.0, | |
| "loss_scale_unsuperv_loss": 10.0, | |
| "ode_solver": "rk4" | |
| }, | |
| "model_type": "fimode", | |
| "normalization_time": { | |
| "name": "fim.models.blocks.MinMaxNormalization" | |
| }, | |
| "normalization_values": { | |
| "name": "fim.models.blocks.MinMaxNormalization" | |
| }, | |
| "time_encoding": { | |
| "name": "fim.models.blocks.SineTimeEncoding", | |
| "out_features": 512 | |
| }, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.51.3", | |
| "trunk_net": { | |
| "dropout": 0.1, | |
| "hidden_act": { | |
| "name": "torch.nn.SELU" | |
| }, | |
| "hidden_layers": [ | |
| 1024, | |
| 1024, | |
| 1024, | |
| 1024 | |
| ], | |
| "in_features": 512, | |
| "name": "fim.models.blocks.MLP", | |
| "out_features": 512, | |
| "output_act": { | |
| "name": "torch.nn.Identity" | |
| } | |
| }, | |
| "use_bf16": false, | |
| "vector_field_net": { | |
| "dropout": 0.1, | |
| "hidden_act": { | |
| "name": "torch.nn.SELU" | |
| }, | |
| "hidden_layers": [], | |
| "in_features": 512, | |
| "name": "fim.models.blocks.MLP", | |
| "out_features": 2, | |
| "output_act": { | |
| "name": "torch.nn.Identity" | |
| } | |
| } | |
| } | |