build tensorflow learn estimator and run -- generate plots along the way. the input function will queue as needed.

tf_learn_example.py

import tensorflow.contrib.learn as learn
from tensorflow.contrib.learn.python.learn.estimators import model_fn as model_fn_lib
import tensorflow as tf
import tensorflow.contrib.slim as sm
import numpy as np
import functools

def build_toy_dataset(N=40, noise_std=0.1):
    D = 1
    x = np.concatenate([np.linspace(0, 2, num=N / 2),
                      np.linspace(6, 8, num=N / 2)])
    y = np.cos(x) + np.random.normal(0, noise_std, size=N)
    x = (x - 4.0) / 4.0
    x = x.astype(np.float32).reshape((N, D))
    y = y.astype(np.float32).reshape((N,1))
    return x, y

x,y=build_toy_dataset()

def train_fn(x,y,queue_all=True):
    def input_fn():
        data=tf.train.batch({'input':x,'target':y},batch_size=1,enqueue_many=queue_all)
        data['input']=tf.reshape(data['input'],[-1,1])
        data['target']=tf.reshape(data['target'],[-1,1])
        return {'input':data['input']},data['target']
    return input_fn




def model_fn(features,targets,mode,params):
    print(features['input'])
    net=sm.repeat(features['input'],3,sm.fully_connected, 100,activation_fn=tf.nn.relu6)
    net=sm.fully_connected(net,1,activation_fn=None)
    prediction=net

    loss=tf.losses.mean_squared_error(targets,prediction)
    train_op = tf.contrib.layers.optimize_loss(
      loss=loss,
      global_step=tf.contrib.framework.get_global_step(),
      learning_rate=params.get('learning_rate',.001),
      optimizer="Adam")
    eval_metric_ops={}
    return learn.ModelFnOps(
        mode=mode,
        predictions=prediction,
        loss=loss,
        train_op=train_op,
        eval_metric_ops=eval_metric_ops)

esto=learn.Estimator(model_fn,params={"learning_rate":.001})
plt.gcf()
plt.gca()
for i in range(3):
    esto.fit(input_fn=train_fn(x,y),steps=10)

    jj=esto.predict(input_fn=train_fn(x,y,queue_all=False),as_iterable=False)
    plt.plot(x,y)
    plt.plot(x,jj.reshape((-1,1)))
    plt.show()

for tensorflow 1.0