#!/usr/bin/python3
# -*- coding: utf-8 -*-
import os
from sklearn import manifold
import numpy as np
import matplotlib.pylab as plt
from mpl_toolkits.mplot3d import Axes3D
from model.utils import arg, Param
if __name__ == '__main__':
args = arg()
config_path = args.cfg
params = Param(config_path)
vecs = np.loadtxt(os.path.join(config_path, f"emb/{args.epoch}vecs.tsv"),
dtype=np.float,
delimiter='\t')
tsne = manifold.TSNE(n_components=3, learning_rate=100, n_iter=350)
out = tsne.fit_transform(vecs)
fig = plt.figure()
ax = Axes3D(fig)
for i in range(18):
ax.scatter(out[i * 100:(i + 1) * 100, 0],
out[i * 100:(i + 1) * 100, 1], out[i * 100:(i + 1) * 100,
2])
plt.show()
print(len(gpus), "Physical GPUs,", len(logical_gpus), "Logical GPU")
except RuntimeError as e:
print(e)
from tensorflow.keras.applications import MobileNetV2
from tensorflow.keras.models import Model
# from model.balance_input_fn import dataset_pipeline
from model.input_fn import dataset_pipeline
from model.triplet_loss import batch_hard_triplet_loss
from model.utils import Param, arg
if __name__ == "__main__":
args = arg(True)
config_path = args.cfg
params = Param(config_path)
# dataset
train_ds, train_count = dataset_pipeline(params)
test_ds, test_count = dataset_pipeline(params, True)
# create model
baseModel = MobileNetV2(include_top=False,
weights='imagenet',
input_shape=(224, 224, 3),
pooling="avg")
fc = tf.keras.layers.Dense(params.NUM_CLASSES,
activation="softmax",
name="dense_final")(baseModel.output)