Update Neural Insights Getting Started documentation (#1127)

Signed-off-by: bmyrcha <bartosz.myrcha@intel.com>

Author: bmyrcha

neural_insights/README.md

@@ -88,11 +88,13 @@ pip install tensorflow
 # Prepare fp32 model
 wget https://storage.googleapis.com/intel-optimized-tensorflow/models/v1_6/mobilenet_v1_1.0_224_frozen.pb
 ```
+
 ```python
 from neural_compressor.config import PostTrainingQuantConfig
 from neural_compressor.data import DataLoader
 from neural_compressor.data import Datasets
 
+top1 = Metric(name="topk", k=1)
 dataset = Datasets('tensorflow')['dummy'](shape=(1, 224, 224, 3))
 dataloader = DataLoader(framework='tensorflow', dataset=dataset)
 
@@ -101,9 +103,15 @@ q_model = fit(
     model="./mobilenet_v1_1.0_224_frozen.pb",
     conf=PostTrainingQuantConfig(diagnosis=True),
     calib_dataloader=dataloader,
-    eval_dataloader=dataloader)
+    eval_dataloader=dataloader,
+    eval_metric=top1
+)
 ```
 
+When the quantization is started, the workload should appear on the Neural Insights page and successively, new information should be available while quantization is in progress (such as weights distribution and accuracy data).
+
+> Note that above example uses dummy data which is used to describe usage of Neural Insights. For diagnosis purposes you should use real dataset specific for your use case.
+
 ## Research Collaborations
 
 Welcome to raise any interesting research ideas on model compression techniques and feel free to reach us (inc.maintainers@intel.com). Look forward to our collaborations on Neural Insights!

neural_insights/components/workload_manager/quantization_workload.py

@@ -70,8 +70,9 @@ def ratio(self) -> Optional[float]:
         Returns: accuracy ratio if baseline and optimized accuracy are present
                  Otherwise returns None
         """
-        if self.optimized_accuracy is None or self.baseline_accuracy is None:
+        if self.optimized_accuracy in [None, 0] or self.baseline_accuracy in [None, 0]:
             return None
+        
         return (self.optimized_accuracy - self.baseline_accuracy) / self.baseline_accuracy
 
     def serialize(self, serialization_type: str = "default") -> Dict[str, Any]: