Merge pull request #241 from microsoft/master

merge master

src/sdk/pynni/nni/compression/torch/pruners.py

@@ -212,7 +212,7 @@ class AGP_Pruner(Pruner):
         if epoch > 0:
             self.now_epoch = epoch
             for wrapper in self.get_modules_wrapper():
-                wrapper.if_calculated.copy_(torch.tensor(0)) # pylint: disable=not-callable
+                wrapper.if_calculated = False
 
 class SlimPruner(Pruner):
     """
@@ -329,10 +329,6 @@ class LotteryTicketPruner(Pruner):
         reset_weights : bool
             Whether reset weights and optimizer at the beginning of each round.
         """
-        super().__init__(model, config_list, optimizer)
-        self.curr_prune_iteration = None
-        self.prune_iterations = config_list[0]['prune_iterations']
-
         # save init weights and optimizer
         self.reset_weights = reset_weights
         if self.reset_weights:
@@ -344,6 +340,10 @@ class LotteryTicketPruner(Pruner):
             if lr_scheduler is not None:
                 self._scheduler_state = copy.deepcopy(lr_scheduler.state_dict())
 
+        super().__init__(model, config_list, optimizer)
+        self.curr_prune_iteration = None
+        self.prune_iterations = config_list[0]['prune_iterations']
+
     def validate_config(self, model, config_list):
         """
         Parameters

src/sdk/pynni/nni/nas/pytorch/darts/trainer.py

@@ -129,6 +129,7 @@ class DartsTrainer(Trainer):
         self.mutator.reset()
         logits = self.model(X)
         loss = self.loss(logits, y)
+        self._write_graph_status()
         return logits, loss
 
     def _backward(self, val_X, val_y):

src/sdk/pynni/nni/nas/pytorch/enas/trainer.py

@@ -126,6 +126,7 @@ class EnasTrainer(Trainer):
 
             with torch.no_grad():
                 self.mutator.reset()
+            self._write_graph_status()
             logits = self.model(x)
 
             if isinstance(logits, tuple):
@@ -159,6 +160,7 @@ class EnasTrainer(Trainer):
                 self.mutator.reset()
                 with torch.no_grad():
                     logits = self.model(x)
+                self._write_graph_status()
                 metrics = self.metrics(logits, y)
                 reward = self.reward_function(logits, y)
                 if self.entropy_weight:

src/sdk/pynni/nni/nas/pytorch/trainer.py

@@ -3,6 +3,8 @@
 
 import json
 import logging
+import os
+import time
 from abc import abstractmethod
 
 import torch
@@ -90,6 +92,9 @@ class Trainer(BaseTrainer):
         self.batch_size = batch_size
         self.workers = workers
         self.log_frequency = log_frequency
+        self.log_dir = os.path.join("logs", str(time.time()))
+        os.makedirs(self.log_dir, exist_ok=True)
+        self.status_writer = open(os.path.join(self.log_dir, "log"), "w")
         self.callbacks = callbacks if callbacks is not None else []
         for callback in self.callbacks:
             callback.build(self.model, self.mutator, self)
@@ -168,3 +173,22 @@ class Trainer(BaseTrainer):
         Return trainer checkpoint.
         """
         raise NotImplementedError("Not implemented yet")
+
+    def enable_visualization(self):
+        """
+        Enable visualization. Write graph and training log to folder ``logs/<timestamp>``.
+        """
+        sample = None
+        for x, _ in self.train_loader:
+            sample = x.to(self.device)[:2]
+            break
+        if sample is None:
+            _logger.warning("Sample is %s.", sample)
+        _logger.info("Creating graph json, writing to %s. Visualization enabled.", self.log_dir)
+        with open(os.path.join(self.log_dir, "graph.json"), "w") as f:
+            json.dump(self.mutator.graph(sample), f)
+        self.visualization_enabled = True
+
+    def _write_graph_status(self):
+        if hasattr(self, "visualization_enabled") and self.visualization_enabled:
+            print(json.dumps(self.mutator.status()), file=self.status_writer, flush=True)

src/webui/package.json

@@ -92,5 +92,8 @@
     "presets": [
       "react-app"
     ]
+  },
+  "resolutions": {
+    "npm": ">=6.14.4"
   }
 }

src/webui/src/App.tsx

@@ -17,8 +17,9 @@ interface AppState {
 }
 
 class App extends React.Component<{}, AppState> {
-    private timerId!: number | null;
+    private timerId!: number | undefined;
     private dataFormatimer!: number;
+    private firstLoad: boolean = false; // when click refresh selector options
 
     constructor(props: {}) {
         super(props);
@@ -68,12 +69,18 @@ class App extends React.Component<{}, AppState> {
     }
 
     changeInterval = (interval: number): void => {
-        this.setState({ interval });
-        if (this.timerId === null && interval !== 0) {
-            window.setTimeout(this.refresh);
-        } else if (this.timerId !== null && interval === 0) {
+        
         window.clearTimeout(this.timerId);
+        if (interval === 0) {
+            return;   
         }
+        // setState will trigger page refresh at once.
+        // setState is asyc, interval not update to (this.state.interval) at once.
+        this.setState({interval}, () => {
+            this.firstLoad = true;
+            this.refresh();
+        });
+
     }
 
     // TODO: use local storage
@@ -123,6 +130,10 @@ class App extends React.Component<{}, AppState> {
     }
 
     private refresh = async (): Promise<void> => {
+
+        // resolve this question: 10s -> 20s, page refresh twice.
+        // only refresh this page after clicking the refresh options
+        if (this.firstLoad !== true) {
             const [experimentUpdated, trialsUpdated] = await Promise.all([EXPERIMENT.update(), TRIALS.update()]);
             if (experimentUpdated) {
                 this.setState(state => ({ experimentUpdateBroadcast: state.experimentUpdateBroadcast + 1 }));
@@ -130,17 +141,19 @@ class App extends React.Component<{}, AppState> {
             if (trialsUpdated) {
                 this.setState(state => ({ trialsUpdateBroadcast: state.trialsUpdateBroadcast + 1 }));
             }
+        } else {
+            this.firstLoad = false;
+        }
 
         if (['DONE', 'ERROR', 'STOPPED'].includes(EXPERIMENT.status)) {
             // experiment finished, refresh once more to ensure consistency
-            if (this.state.interval > 0) {
             this.setState({ interval: 0 });
             this.lastRefresh();
+            return;
         }
 
-        } else if (this.state.interval !== 0) {
         this.timerId =  window.setTimeout(this.refresh, this.state.interval * 1000);
-        }
+
     }
 
     public async lastRefresh(): Promise<void> {

test/config/naive_trial/naive_trial.py

+import time
+import nni
+
+if __name__ == '__main__':
+    print('trial start')
+    params = nni.get_next_parameter()
+    print('params:', params)
+    epochs = 2
+
+    for i in range(epochs):
+        nni.report_intermediate_result(0.1 * (i+1))
+        time.sleep(1)
+    nni.report_final_result(0.8)
+    print('trial done')

test/config/pr_tests.yml

@@ -70,12 +70,18 @@ testCases:
   config:
     maxTrialNum: 2
     trialConcurrency: 2
+    trial:
+      codeDir: ../naive_trial
+      command: python3 naive_trial.py
 
 - name: assessor-medianstop
   configFile: test/config/assessors/medianstop.yml
   config:
     maxTrialNum: 2
     trialConcurrency: 2
+    trial:
+      codeDir: ../naive_trial
+      command: python3 naive_trial.py
 
 #########################################################################
 # nni tuners test
@@ -89,7 +95,7 @@ testCases:
     searchSpacePath: ../naive_trial/search_space.json
     trial:
       codeDir: ../naive_trial
-      command: python3 trial.py
+      command: python3 naive_trial.py
 
 - name: tuner-evolution
   configFile: test/config/tuners/evolution.yml
@@ -100,7 +106,7 @@ testCases:
     searchSpacePath: ../naive_trial/search_space.json
     trial:
       codeDir: ../naive_trial
-      command: python3 trial.py
+      command: python3 naive_trial.py
 
 - name: tuner-random
   configFile: test/config/tuners/random.yml
@@ -111,7 +117,7 @@ testCases:
     searchSpacePath: ../naive_trial/search_space.json
     trial:
       codeDir: ../naive_trial
-      command: python3 trial.py
+      command: python3 naive_trial.py
 
 - name: tuner-tpe
   configFile: test/config/tuners/tpe.yml
@@ -122,7 +128,7 @@ testCases:
     searchSpacePath: ../naive_trial/search_space.json
     trial:
       codeDir: ../naive_trial
-      command: python3 trial.py
+      command: python3 naive_trial.py
 
 - name: tuner-batch
   configFile: test/config/tuners/batch.yml
@@ -144,7 +150,7 @@ testCases:
     searchSpacePath: ../naive_trial/search_space.json
     trial:
       codeDir: ../naive_trial
-      command: python3 trial.py
+      command: python3 naive_trial.py
 
 - name: tuner-grid
   configFile: test/config/tuners/gridsearch.yml

test/pipelines/pipelines-it-local-windows.yml

@@ -10,7 +10,7 @@ jobs:
       python -m pip install scikit-learn==0.20.0 --user
       python -m pip install keras==2.1.6 --user
       python -m pip install torchvision===0.4.1 torch===1.3.1 -f https://download.pytorch.org/whl/torch_stable.html --user
-      python -m pip install tensorflow-gpu==1.11.0 --user
+      python -m pip install tensorflow-gpu==1.15.2 --user
     displayName: 'Install dependencies for integration tests'
   - script: |
       cd test

test/pipelines/pipelines-it-pai-windows.yml

@@ -63,6 +63,7 @@ jobs:
       cd test
       set PATH=$(ENV_PATH)
       python --version
-      python nni_test/nnitest/generate_ts_config.py --ts pai --pai_host $(pai_host) --pai_user $(pai_user) --pai_pwd $(pai_pwd) --vc $(pai_virtual_cluster) --nni_docker_image $(docker_image) --data_dir $(data_dir) --output_dir $(output_dir) --nni_manager_ip $(nni_manager_ip)
+      mount -o anon $(pai_nfs_uri) $(local_nfs_uri)
+      python nni_test/nnitest/generate_ts_config.py --ts pai --pai_token $(pai_token) --pai_host $(pai_host) --pai_user $(pai_user) --nni_docker_image $(docker_image) --pai_storage_plugin $(pai_storage_plugin) --nni_manager_nfs_mount_path $(nni_manager_nfs_mount_path) --container_nfs_mount_path $(container_nfs_mount_path) --nni_manager_ip $(nni_manager_ip)
       python nni_test/nnitest/run_tests.py --config config/integration_tests.yml --ts pai --exclude multi-phase
     displayName: 'Examples and advanced features tests on pai'
\ No newline at end of file

test/pipelines/pipelines-it-remote-windows.yml

@@ -52,9 +52,3 @@ jobs:
       runOptions: commands
       commands: python3 /tmp/nnitest/$(Build.BuildId)/nni-remote/test/nni_test/nnitest/remote_docker.py --mode stop --name $(Build.BuildId) --os windows
     displayName: 'Stop docker'
-  - task: SSH@0
-    inputs:
-      sshEndpoint: $(end_point)
-      runOptions: commands
-      commands: sudo rm -rf /tmp/nnitest/$(Build.BuildId)
-    displayName: 'Clean the remote files'  

test/scripts/model_compression.sh

@@ -19,6 +19,15 @@ do
     python3 model_prune_torch.py --pruner_name $name --pretrain_epochs 1 --prune_epochs 1
 done
 
+echo 'testing level pruner pruning'
+python3 model_prune_torch.py --pruner_name level --pretrain_epochs 1 --prune_epochs 1
+
+echo 'testing agp pruning'
+python3 model_prune_torch.py --pruner_name agp --pretrain_epochs 1 --prune_epochs 2
+
+echo 'testing mean_activation pruning'
+python3 model_prune_torch.py --pruner_name mean_activation --pretrain_epochs 1 --prune_epochs 1
+
 #echo "testing lottery ticket pruning..."
 #python3 lottery_torch_mnist_fc.py
 

test/scripts/nas.sh

@@ -28,9 +28,10 @@ cd $EXAMPLE_DIR/enas
 python3 search.py --search-for macro --epochs 1
 python3 search.py --search-for micro --epochs 1
 
-echo "testing naive..."
-cd $EXAMPLE_DIR/naive
-python3 train.py
+#disabled for now
+#echo "testing naive..."
+#cd $EXAMPLE_DIR/naive
+#python3 train.py
 
 echo "testing pdarts..."
 cd $EXAMPLE_DIR/pdarts