Edits tensorflow model layers
Project description
TFSurgeon
Allows for dynamic tensorflow keras model edition.
You can install this package with pip and call it as from tfsurgeon import edit_model
.
How to use
- remove layers based on layer type
- remove layers based on name or index
- replace layers from one type to another
- replace specific layer (name) by another layer (callable)
see docstring for examples in src/tfsurgeon/tfsurgeon.py
.
import
from tfsurgeon import edit_model
import tensorflow as tf
model = tf.keras.applications.MobileNetV2(weights=None)
new_model = edit_model(
model=model,
replace_layers_by_types=[
[tf.keras.layers.BatchNormalization, tf.keras.layers.LayerNormalization]
],
verbose=True,
)
CLI
no supported yet
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
tfsurgeon-0.0.2.tar.gz
(7.5 kB
view hashes)
Built Distribution
tfsurgeon-0.0.2-py3-none-any.whl
(12.7 kB
view hashes)
Close
Hashes for tfsurgeon-0.0.2-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 86d9ee018d71423534029795ed2026b6ecbe728d5570c1da80260d181cd2965e |
|
MD5 | 583c8a5e4ad5735096a8fc28ab4a8343 |
|
BLAKE2b-256 | 7153be6ec9c45330498ada811cb97821ab431cf9e6a1a9e0e07d033aff52c64f |