Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/GenSI-THUAIR/MolFM
/ functions
Functions
56 in github.com/GenSI-THUAIR/MolFM
⨍
Functions
56
◇
Types & classes
10
↓ 5 callers
Method
__init__
( self, input_nf, output_nf, hidden_nf, normalization_factor,
sampling/egnn.py:7
↓ 4 callers
Function
T
(t)
sampling/cnf_models.py:8
↓ 4 callers
Function
T_hat
(t)
sampling/cnf_models.py:15
↓ 4 callers
Function
assert_mean_zero_with_mask
(x, node_mask, eps=1e-10)
sampling/cnf_models.py:38
↓ 4 callers
Method
phi
(self, t, x, node_mask, edge_mask, context)
sampling/cnf_models.py:156
↓ 4 callers
Method
sample_combined_position_feature_noise
Samples mean-centered normal noise for z_x, and standard normal noise for z_h.
sampling/cnf_models.py:318
↓ 3 callers
Function
remove_mean_with_mask
(x, node_mask)
sampling/cnf_models.py:23
↓ 2 callers
Method
__init__
( self, dynamics, in_node_nf: int, n_dims: int, timesteps: int = 10000
sampling/cnf_models.py:71
↓ 2 callers
Method
_forward
(self, t, xh, node_mask, edge_mask, context)
sampling/cnf_models.py:513
↓ 2 callers
Function
coord2diff
(x, edge_index, norm_constant=1)
sampling/egnn.py:391
↓ 2 callers
Method
sample_p_xh_given_z0
Samples x ~ p(x|z0).
sampling/cnf_models.py:228
↓ 2 callers
Method
unnormalize
(self, x, h_cat, h_int, node_mask)
sampling/cnf_models.py:200
↓ 2 callers
Function
unsorted_segment_sum
Custom PyTorch op to replicate TensorFlow's `unsorted_segment_sum`. Normalization: 'sum' or 'mean'.
sampling/egnn.py:400
↓ 1 callers
Function
assert_correctly_masked
(variable, node_mask)
sampling/cnf_models.py:46
↓ 1 callers
Method
coord_model
(self, h, coord, edge_index, coord_diff, edge_attr, edge_mask)
sampling/egnn.py:118
↓ 1 callers
Method
decode
(self, z, node_mask, edge_mask, context)
sampling/cnf_models.py:251
↓ 1 callers
Method
decode_chain
(self, z, t, node_mask, edge_mask, context)
sampling/cnf_models.py:284
↓ 1 callers
Method
edge_model
(self, source, target, edge_attr, edge_mask)
sampling/egnn.py:41
↓ 1 callers
Method
get_adj_matrix
(self, n_nodes, batch_size, device)
sampling/cnf_models.py:588
↓ 1 callers
Method
node_model
(self, x, edge_index, edge_attr, node_attr)
sampling/egnn.py:58
↓ 1 callers
Method
reverse
(self, tensor)
sampling/cnf_models.py:679
↓ 1 callers
Method
sample
Draw samples from the generative model.
sampling/cnf_models.py:354
↓ 1 callers
Function
sample_center_gravity_zero_gaussian_with_mask
(size, device, node_mask)
sampling/cnf_models.py:50
↓ 1 callers
Function
sample_gaussian_with_mask
(size, device, node_mask)
sampling/cnf_models.py:61
↓ 1 callers
Method
set_odeint
(self, method="dopri5", rtol=1e-4, atol=1e-4)
sampling/cnf_models.py:133
↓ 1 callers
Method
subspace_dimensionality
Compute the dimensionality on translation-invariant linear subspace where distributions on x are defined.
sampling/cnf_models.py:175
↓ 1 callers
Method
unnormalize_z
(self, z, node_mask)
sampling/cnf_models.py:211
Method
__init__
(self, in_node_nf, context_node_nf, n_dims,
sampling/cnf_models.py:459
Method
__init__
(self, histogram)
sampling/cnf_models.py:614
Method
__init__
(self)
sampling/cnf_models.py:662
Method
__init__
( self, hidden_nf, normalization_factor, aggregation_method, edges_in_
sampling/egnn.py:92
Method
__init__
( self, hidden_nf, edge_feat_nf=2, device="cpu", act_fn=nn.SiLU(),
sampling/egnn.py:158
Method
__init__
( self, in_node_nf, in_edge_nf, hidden_nf, device="cpu", act_f
sampling/egnn.py:238
Method
__init__
( self, in_node_nf, in_edge_nf, hidden_nf, aggregation_method="sum",
sampling/egnn.py:322
Method
__init__
(self, max_res=15.0, min_res=15.0 / 2000.0, div_factor=4)
sampling/egnn.py:376
Function
assert_mean_zero
(x)
sampling/cnf_models.py:33
Method
check_issues_norm_values
(self, num_stdevs=8)
sampling/cnf_models.py:140
Method
forward
(self, t, xh, node_mask, edge_mask, context=None)
sampling/cnf_models.py:500
Method
forward
(self, tensor, node_mask, edge_mask, context)
sampling/cnf_models.py:665
Method
forward
( self, h, edge_index, edge_attr=None, node_attr=None, node_ma
sampling/egnn.py:74
Method
forward
( self, h, coord, edge_index, coord_diff, edge_attr=None,
sampling/egnn.py:141
Method
forward
(self, h, x, edge_index, node_mask=None, edge_mask=None, edge_attr=None)
sampling/egnn.py:213
Method
forward
(self, h, x, edge_index, node_mask=None, edge_mask=None)
sampling/egnn.py:298
Method
forward
(self, h, edges, edge_attr=None, node_mask=None, edge_mask=None)
sampling/egnn.py:360
Method
forward
(self, x)
sampling/egnn.py:384
Method
fwd
(time, state)
sampling/cnf_models.py:505
Method
inflate_batch_array
Inflates the batch array (array) with only a single axis (i.e. shape = (batch_size,), or possibly more empty axes (i.e. shape (batch_
sampling/cnf_models.py:167
Method
log_prob
(self, batch_n_nodes)
sampling/cnf_models.py:637
Method
normalize
(self, x, h, node_mask)
sampling/cnf_models.py:180
Method
sample
(self, n_samples=1)
sampling/cnf_models.py:633
Method
sample_cat_z0
get the catgorical distribution according to coordinate and features.
sampling/cnf_models.py:335
Method
sample_chain
Draw samples from the generative model, keep the intermediate states for visualization purposes.
sampling/cnf_models.py:414
Method
sample_normal
Samples from a Normal distribution.
sampling/cnf_models.py:243
Method
unwrap_forward
(self)
sampling/cnf_models.py:510
Method
wrap_forward
(self, node_mask, edge_mask, context)
sampling/cnf_models.py:503
Method
wrapper
(t, x)
sampling/cnf_models.py:252