Add summary to a tensor, scalar summary if the tensor is 1D, else scalar and histogram summary

tefla.core.summary.summary_metric (tensor, name=None, collections=None)

Args

• tensor: a tensor to add summary
• name: name of the tensor
• collections: training or validation collections

---

Add summary to a tensor, scalar summary if the tensor is 1D, else scalar and histogram summary

tefla.core.summary.summary_activation (tensor, name=None, collections=None)

Args

• tensor: a tensor to add summary
• name: name of the tensor
• collections: training or validation collections

---

creates the summar writter for training and validation

tefla.core.summary.create_summary_writer (summary_dir, sess)

Args

• summary_dir: the directory to write summary
• sess: the session to sun the ops

Returns

training and vaidation summary writter

---

Add summary as per the ops mentioned

tefla.core.summary.summary_param (op, tensor, ndims, name, collections=None)

Args

• op: name of the summary op; e.g. 'stddev' available ops: ['scalar', 'histogram', 'sparsity', 'mean', 'rms', 'stddev', 'norm', 'max', 'min']
• tensor: the tensor to add summary
• ndims: dimension of the tensor
• name: name of the op
• collections: training or validation collections

---

Add summary to all trainable tensors

tefla.core.summary.summary_trainable_params (summary_types, collections=None)

Args

• summary_type: a list of all sumary types to add e.g.: ['scalar', 'histogram', 'sparsity', 'mean', 'rms', 'stddev', 'norm', 'max', 'min']
• collections: training or validation collections

---

Add summary to all gradient tensors

tefla.core.summary.summary_gradients (grad_vars, summary_types, collections=None)

Args

• grads_vars: grads and vars list
• summary_type: a list of all sumary types to add e.g.: ['scalar', 'histogram', 'sparsity', 'mean', 'rms', 'stddev', 'norm', 'max', 'min']
• collections: training or validation collections

---

Add image summary to a image tensor

tefla.core.summary.summary_image (tensor, name=None, max_images=10, collections=None)

Args

• tensor: a tensor to add summary
• name: name of the tensor
• max_images: num of images to add summary
• collections: training or validation collections

---