Simulai builtin optimizers
Built-in Optimizers #
SpaRSA #
Source code in simulai/optimization/_builtin.py
20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 |
|
__init__(lambd=None, alpha_0=None, epsilon=1e-10, sparsity_tol=1e-15, use_mean=False, transform=None)
#
Sparse Regression Algorithm
Parameters:
Name | Type | Description | Default |
---|---|---|---|
lambd |
float
|
Quadratic regularization penalty. |
None
|
alpha_0 |
float
|
Update step lenght. |
None
|
epsilon |
float
|
Error tolerance. |
1e-10
|
sparsity_tol |
float
|
Sparsity tolerance. |
1e-15
|
use_mean |
bool
|
Use mean for evaluating loss or not. |
False
|
transform |
callable
|
A transformation to be applied to the data. |
None
|
Source code in simulai/optimization/_builtin.py
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 |
|
fit(input_data=None, target_data=None)
#
Parameters:
Name | Type | Description | Default |
---|---|---|---|
input_data |
ndarray
|
Input data for training the model. |
None
|
target_data |
ndarray
|
Target data for training the model. |
None
|
Returns:
Source code in simulai/optimization/_builtin.py
111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 |
|
BBI #
Bases: Optimizer
Source code in simulai/optimization/_builtin_pytorch.py
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 |
|
__init__(params=None, lr=0.001, eps1=1e-10, eps2=1e-40, v0=0, threshold0=1000, threshold=3000, deltaEn=0.0, consEn=True, n_fixed_bounces=1)
#
Optimizer based on the BBI model of inflation.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
params |
iterable
|
iterable of parameters to optimize or dicts defining parameter groups |
None
|
lr |
float
|
learning rate |
0.001
|
v0 |
float
|
expected minimum of the potential (Delta V in the paper) |
0
|
threshold0 |
int
|
threshold for fixed bounces (T0 in the paper) |
1000
|
threshold1 |
int
|
threshold for progress-dependent bounces (T1 in the paper) |
required |
deltaEn |
float
|
extra initial energy (delta E in the paper) |
0.0
|
consEn |
bool
|
if True enforces energy conservation at every step |
True
|
n_fixed_bounces |
int
|
number of bounces every T0 iterations (Nb in the paper) |
1
|
Source code in simulai/optimization/_builtin_pytorch.py
23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 |
|
step(closure)
#
Parameters:
Name | Type | Description | Default |
---|---|---|---|
closure |
callable
|
A function which enclosures the loss evaluation. |
required |
Returns: torch.Tensor: The evaluation for the loss function.
Source code in simulai/optimization/_builtin_pytorch.py
72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 |
|