diff options
| author | Jules Laplace <julescarbon@gmail.com> | 2020-02-18 18:39:45 +0100 |
|---|---|---|
| committer | Jules Laplace <julescarbon@gmail.com> | 2020-02-18 18:39:45 +0100 |
| commit | 840b5ea07cd6871ae0dd041da2aab906d5fff334 (patch) | |
| tree | 08ca91bc6dd9bd2dc6051d12163de7fbfcfe4892 /cli/app | |
| parent | 75fa7f62aa9dbbbe3d69d03ad243a63a4b17c192 (diff) | |
simple sum
Diffstat (limited to 'cli/app')
| -rw-r--r-- | cli/app/search/search_dense.py | 8 |
1 files changed, 4 insertions, 4 deletions
diff --git a/cli/app/search/search_dense.py b/cli/app/search/search_dense.py index f52954c..f0e0a68 100644 --- a/cli/app/search/search_dense.py +++ b/cli/app/search/search_dense.py @@ -302,7 +302,7 @@ def find_dense_embedding_for_images(params, opt_tag="inverse_" + timestamp(), op # img_rec_err = params.lambda_mse * img_mse_err + params.lambda_feat * img_feat_err inv_loss = (params.lambda_mse * mse_loss + params.lambda_feat * feat_loss) - inv_loss_quad = (params.lambda_mse * mse_loss_quad + params.lambda_feat * feat_loss_quad) + # inv_loss_quad = (params.lambda_mse * mse_loss_quad + params.lambda_feat * feat_loss_quad) # inv_loss_quint = params.lambda_mse * mse_loss_quint + params.lambda_feat * feat_loss_quint # -------------------------- @@ -328,9 +328,9 @@ def find_dense_embedding_for_images(params, opt_tag="inverse_" + timestamp(), op inv_train_op = optimizer.minimize(inv_loss, var_list=trained_params, global_step=inv_step) reinit_optimizer = tf.variables_initializer(optimizer.variables()) - optimizer_quad = tf.train.AdamOptimizer(learning_rate=lrate, beta1=0.9, beta2=0.999) - inv_train_op_quad = optimizer_quad.minimize(inv_loss_quad, var_list=trained_params, global_step=inv_step) - reinit_optimizer_quad = tf.variables_initializer(optimizer_quad.variables()) + # optimizer_quad = tf.train.AdamOptimizer(learning_rate=lrate, beta1=0.9, beta2=0.999) + # inv_train_op_quad = optimizer_quad.minimize(inv_loss_quad, var_list=trained_params, global_step=inv_step) + # reinit_optimizer_quad = tf.variables_initializer(optimizer_quad.variables()) # optimizer_quint = tf.train.AdamOptimizer(learning_rate=lrate, beta1=0.9, beta2=0.999) # inv_train_op_quint = optimizer_quint.minimize(inv_loss_quint, var_list=trained_params, global_step=inv_step) |