Merge branch 'master' into arpin_billable_seconds

Author: djarpin

.gitignore

@@ -19,3 +19,5 @@ examples/tensorflow/distributed_mnist/data
 *.iml
 doc/_build
 **/.DS_Store
+venv/
+*~

CHANGELOG.rst

@@ -0,0 +1,48 @@
+=========
+CHANGELOG
+=========
+
+1.0.4
+=====
+
+* feature: Estimators: add support for Amazon Neural Topic Model(NTM) algorithm
+* feature: Documentation: Fix description of an argument of sagemaker.session.train
+* feature: Documentation: Add FM and LDA to the documentation
+* feature: Estimators: add support for async fit
+* bug-fix: Estimators: fix estimator role expansion
+
+1.0.3
+=====
+
+* feature: Estimators: add support for Amazon LDA algorithm
+* feature: Hyperparameters: Add data_type to hyperparameters
+* feature: Documentation: Update TensorFlow examples following API change
+* feature: Session: Support multi-part uploads
+
+
+1.0.2
+=====
+
+* feature: Estimators: add support for Amazon FactorizationMachines algorithm
+* feature: Session: Correctly handle TooManyBuckets error_code in default_bucket method
+* feature: Tests: add training failure tests for TF and MXNet
+* feature: Documentation: show how to make predictions against existing endpoint
+* feature: Estimators: implement write_spmatrix_to_sparse_tensor to support any scipy.sparse matrix
+
+
+1.0.1
+=====
+
+* api-change: Model: Remove support for 'supplemental_containers' when creating Model
+* feature: Documentation: multiple updates
+* feature: Tests: ignore tests data in tox.ini, increase timeout for endpoint creation, capture exceptions during endpoint deletion, tests for input-output functions
+* feature: Logging: change to describe job every 30s when showing logs
+* feature: Session: use custom user agent at all times
+* feature: Setup: add travis file
+
+
+1.0.0
+=====
+
+* Initial commit
+

README.rst

@@ -18,7 +18,7 @@ def __getattr__(cls, name):
                 'tensorflow.python.framework', 'tensorflow_serving', 'tensorflow_serving.apis']
 sys.modules.update((mod_name, Mock()) for mod_name in MOCK_MODULES)
 
-version = '1.0'
+version = '1.0.4'
 project = u'sagemaker'
 
 # Add any Sphinx extension module names here, as strings. They can be extensions

doc/conf.py

@@ -0,0 +1,22 @@
+FactorizationMachines
+-------------------------
+
+The Amazon SageMaker Factorization Machines algorithm.
+
+.. autoclass:: sagemaker.FactorizationMachines
+    :members:
+    :undoc-members:
+    :show-inheritance:
+    :inherited-members:
+    :exclude-members: image, num_factors, predictor_type, epochs, clip_gradient, mini_batch_size, feature_dim, eps, rescale_grad, bias_lr, linear_lr, factors_lr, bias_wd, linear_wd, factors_wd, bias_init_method, bias_init_scale, bias_init_sigma, bias_init_value, linear_init_method, linear_init_scale, linear_init_sigma, linear_init_value, factors_init_method, factors_init_scale, factors_init_sigma, factors_init_value
+
+
+.. autoclass:: sagemaker.FactorizationMachinesModel
+    :members:
+    :undoc-members:
+    :show-inheritance:
+
+.. autoclass:: sagemaker.FactorizationMachinesPredictor
+    :members:
+    :undoc-members:
+    :show-inheritance:

doc/factorization_machines.rst

@@ -38,11 +38,15 @@ A managed environment for TensorFlow training and hosting on Amazon SageMaker
 
 SageMaker First-Party Algorithms
 --------------------------------
-Amazon provides implementations of some common machine learning algortithms optimized for GPU archicture and massive datasets.
+Amazon provides implementations of some common machine learning algortithms optimized for GPU architecture and massive datasets.
 
 .. toctree::
     :maxdepth: 2
 
     kmeans
     pca
     linear_learner
+    sagemaker.amazon.amazon_estimator
+    factorization_machines
+    lda
+    ntm

doc/index.rst

@@ -0,0 +1,22 @@
+LDA
+--------------------
+
+The Amazon SageMaker LDA algorithm.
+
+.. autoclass:: sagemaker.LDA
+    :members:
+    :undoc-members:
+    :show-inheritance:
+    :inherited-members:
+    :exclude-members: image, num_topics, alpha0, max_restarts, max_iterations, mini_batch_size, feature_dim, tol
+
+
+.. autoclass:: sagemaker.LDAModel
+    :members:
+    :undoc-members:
+    :show-inheritance:
+
+.. autoclass:: sagemaker.LDAPredictor
+    :members:
+    :undoc-members:
+    :show-inheritance:

doc/lda.rst

@@ -0,0 +1,23 @@
+NTM
+--------------------
+
+The Amazon SageMaker NTM algorithm.
+
+.. autoclass:: sagemaker.NTM
+    :members:
+    :undoc-members:
+    :show-inheritance:
+    :inherited-members:
+    :exclude-members: image, num_topics, encoder_layers, epochs, encoder_layers_activation, optimizer, tolerance,
+                 num_patience_epochs, batch_norm, rescale_gradient, clip_gradient, weight_decay, learning_rate
+
+
+.. autoclass:: sagemaker.NTMModel
+    :members:
+    :undoc-members:
+    :show-inheritance:
+
+.. autoclass:: sagemaker.NTMPredictor
+    :members:
+    :undoc-members:
+    :show-inheritance:

doc/ntm.rst

@@ -0,0 +1,126 @@
+{
+  "nodes": [
+    {
+      "op": "null", 
+      "name": "data", 
+      "inputs": []
+    }, 
+    {
+      "op": "null", 
+      "name": "sequential0_dense0_weight", 
+      "attr": {
+        "__dtype__": "0", 
+        "__lr_mult__": "1.0", 
+        "__shape__": "(128, 0)", 
+        "__wd_mult__": "1.0"
+      }, 
+      "inputs": []
+    }, 
+    {
+      "op": "null", 
+      "name": "sequential0_dense0_bias", 
+      "attr": {
+        "__dtype__": "0", 
+        "__init__": "zeros", 
+        "__lr_mult__": "1.0", 
+        "__shape__": "(128,)", 
+        "__wd_mult__": "1.0"
+      }, 
+      "inputs": []
+    }, 
+    {
+      "op": "FullyConnected", 
+      "name": "sequential0_dense0_fwd", 
+      "attr": {"num_hidden": "128"}, 
+      "inputs": [[0, 0, 0], [1, 0, 0], [2, 0, 0]]
+    }, 
+    {
+      "op": "Activation", 
+      "name": "sequential0_dense0_relu_fwd", 
+      "attr": {"act_type": "relu"}, 
+      "inputs": [[3, 0, 0]]
+    }, 
+    {
+      "op": "null", 
+      "name": "sequential0_dense1_weight", 
+      "attr": {
+        "__dtype__": "0", 
+        "__lr_mult__": "1.0", 
+        "__shape__": "(64, 0)", 
+        "__wd_mult__": "1.0"
+      }, 
+      "inputs": []
+    }, 
+    {
+      "op": "null", 
+      "name": "sequential0_dense1_bias", 
+      "attr": {
+        "__dtype__": "0", 
+        "__init__": "zeros", 
+        "__lr_mult__": "1.0", 
+        "__shape__": "(64,)", 
+        "__wd_mult__": "1.0"
+      }, 
+      "inputs": []
+    }, 
+    {
+      "op": "FullyConnected", 
+      "name": "sequential0_dense1_fwd", 
+      "attr": {"num_hidden": "64"}, 
+      "inputs": [[4, 0, 0], [5, 0, 0], [6, 0, 0]]
+    }, 
+    {
+      "op": "Activation", 
+      "name": "sequential0_dense1_relu_fwd", 
+      "attr": {"act_type": "relu"}, 
+      "inputs": [[7, 0, 0]]
+    }, 
+    {
+      "op": "null", 
+      "name": "sequential0_dense2_weight", 
+      "attr": {
+        "__dtype__": "0", 
+        "__lr_mult__": "1.0", 
+        "__shape__": "(10, 0)", 
+        "__wd_mult__": "1.0"
+      }, 
+      "inputs": []
+    }, 
+    {
+      "op": "null", 
+      "name": "sequential0_dense2_bias", 
+      "attr": {
+        "__dtype__": "0", 
+        "__init__": "zeros", 
+        "__lr_mult__": "1.0", 
+        "__shape__": "(10,)", 
+        "__wd_mult__": "1.0"
+      }, 
+      "inputs": []
+    }, 
+    {
+      "op": "FullyConnected", 
+      "name": "sequential0_dense2_fwd", 
+      "attr": {"num_hidden": "10"}, 
+      "inputs": [[8, 0, 0], [9, 0, 0], [10, 0, 0]]
+    }
+  ], 
+  "arg_nodes": [0, 1, 2, 5, 6, 9, 10], 
+  "node_row_ptr": [
+    0, 
+    1, 
+    2, 
+    3, 
+    4, 
+    5, 
+    6, 
+    7, 
+    8, 
+    9, 
+    10, 
+    11, 
+    12
+  ], 
+  "heads": [[11, 0, 0]], 
+  "attrs": {"mxnet_version": ["int", 1100]}
+}

examples/cli/host/data/model.json

@@ -0,0 +1,3 @@
+#!/bin/bash
+
+sagemaker mxnet host --role-name <your-sagemaker-execution-role>

examples/cli/host/data/model.params

@@ -0,0 +1,41 @@
+from __future__ import print_function
+
+import json
+import mxnet as mx
+from mxnet import gluon
+
+
+def model_fn(model_dir):
+    """
+    Load the gluon model. Called once when hosting service starts.
+
+    :param: model_dir The directory where model files are stored.
+    :return: a model (in this case a Gluon network)
+    """
+    symbol = mx.sym.load('%s/model.json' % model_dir)
+    outputs = mx.symbol.softmax(data=symbol, name='softmax_label')
+    inputs = mx.sym.var('data')
+    param_dict = gluon.ParameterDict('model_')
+    net = gluon.SymbolBlock(outputs, inputs, param_dict)
+    net.load_params('%s/model.params' % model_dir, ctx=mx.cpu())
+    return net
+
+
+def transform_fn(net, data, input_content_type, output_content_type):
+    """
+    Transform a request using the Gluon model. Called once per request.
+
+    :param net: The Gluon model.
+    :param data: The request payload.
+    :param input_content_type: The request content type.
+    :param output_content_type: The (desired) response content type.
+    :return: response payload and content type.
+    """
+    # we can use content types to vary input/output handling, but
+    # here we just assume json for both
+    parsed = json.loads(data)
+    nda = mx.nd.array(parsed)
+    output = net(nda)
+    prediction = mx.nd.argmax(output, axis=1)
+    response_body = json.dumps(prediction.asnumpy().tolist())
+    return response_body, output_content_type

examples/cli/host/run_hosting_example.sh

@@ -0,0 +1,10 @@
+from mxnet import gluon
+
+
+def download_training_data():
+    gluon.data.vision.MNIST('./data/training', train=True)
+    gluon.data.vision.MNIST('./data/training', train=False)
+
+
+if __name__ == "__main__":
+    download_training_data()

examples/cli/host/script.py

@@ -0,0 +1,7 @@
+{
+  "batch_size": 100,
+  "epochs": 10,
+  "learning_rate": 0.1,
+  "momentum": 0.9,
+  "log_interval": 100
+}

examples/cli/train/data/training/t10k-images-idx3-ubyte.gz

@@ -0,0 +1,4 @@
+#!/bin/bash
+
+python ./download_training_data.py
+sagemaker mxnet train --role-name <your-sagemaker-execution-role>