Searched refs:model (Results 1 – 6 of 6) sorted by relevance
| /honee/tests/createPyTorchModel/ |
| H A D | update_weights.py | 40 model = anisoSGS()
41 model.load_state_dict(torch.load(f'{model_name}.pt', map_location=torch.device('cpu')))
42 model.double()
47 m, n = model.net[layer].weight.shape
50 model.net[layer].weight[...] = torch.from_numpy(weights[i])[...]
51 model.net[layer].bias[...] = torch.from_numpy(biases[i])[...]
59 model = torch.jit.trace(model, dummy_input)
60 torch.jit.save(model, f"{model_name}_fp64_jit.pt")
62 return model
67 model = load_n_trace_model(model_name)
|
| H A D | README.md | 1 This directory exists to create a PyTorch model with certain weights and biases. It is mostly setup…
|
| /honee/problems/torch/ |
| H A D | sgs_model_torch.cpp | 10 torch::jit::script::Module model; variable
59 PetscCallCXX(model = torch::jit::load(model_path)); in LoadModel_Torch()
60 PetscCallCXX(model.to(torch::Device(device_model))); in LoadModel_Torch()
110 PetscCallCXX(output_tensor = model.forward({input_tensor}).toTensor()); in ModelInference_Torch()
|
| /honee/doc/ |
| H A D | theory.md | 412 For explicit LES, it is defined by a subgrid stress model.
423 - Type of subgrid stress model to use. Currently only `data_driven` is available
428 (sgs-dd-model)=
431 The data-driven SGS model implemented here uses a small neural network to compute the SGS term.
433 More details regarding the theoretical background of the model can be found in {cite}`prakashDDSGS2…
445 The data-driven model parameters in the examples directory are not accurate and are for regression …
450 There are two different modes for using the data-driven model: fused and sequential.
452 In fused mode, the input processing, model inference, and output handling were all done in a single…
453 Fused mode is generally faster than the sequential mode, however fused mode requires that the model…
456 Sequential mode has separate function calls/CeedOperators for input creation, model inference, and …
[all …]
|
| H A D | auxiliary.md | 379 Currently the code is only setup to do *in situ* training for the SGS data-driven model.
380 Training data is split into the model inputs and outputs.
381 The model inputs are calculated as the same model inputs in the SGS Data-Driven model described in …
382 The model outputs (or targets in the case of training) are the subgrid stresses.
407 - Whether to enable *in situ* training of data-driven SGS model. Require building with SmartRedis.
|
| H A D | references.bib | 76 title = {Roofline: an insightful visual performance model for multicore architectures},
|