Update README.md

Author: indierodo

README.md

@@ -1,15 +1,38 @@
 # Docker DeepSpeech Server
-
 This is a dockerfile to serve a [deepspeech server](https://github.com/MainRo/deepspeech-server).
 
 # How to use this image
+Create a deepspeech server configuration file as specified in config.example.yaml. Download and copy
+a model tflite and a scorer file and run the following command:
 
-    docker run -dt --name deepspeech -p 0.0.0.0:8080:8080 -v [host-model-path]:/opt/deepspeech romainsah/deepspeech-server:latest
+    docker run -dt --name deepspeech \
+    -p 0.0.0.0:8080:8080 \
+    -v [host-model-path]:/opt/deepspeech \
+    romainsah/deepspeech-server:latest
 
 where [host-model-path] is a host directory that contains the deepspeech server
-configuration file, the model file, and the scorer file.
+configuration file (e.g. config.yaml), the model file (e.g. model.tflite), and the scorer file (e.g. huge-vocabulary.scorer).
 
 ## Running on GPU
+Change the tag of the docker image in above command to latest-gpu, like this:
+
+    docker run -dt --name deepspeech \
+    -p 0.0.0.0:8080:8080 \
+    -v [host-model-path]:/opt/deepspeech \
+    romainsah/deepspeech-server:latest-gpu
+
+## How to get the models
+
+The latest docker image was tested with the following model and scorer files:
+
+https://github.com/coqui-ai/STT-models/releases/download/english/coqui/v1.0.0-huge-vocab/model.tflite
+
+https://github.com/coqui-ai/STT-models/releases/download/english/coqui/v1.0.0-huge-vocab/huge-vocabulary.scorer
+
+## How to build this image
 
-    docker run -dt --gpus all --name deepspeech -p 0.0.0.0:8080:8080 -v [host-model-path]:/opt/deepspeech romainsah/deepspeech-server:latest-gpu
+    cd cpu && \
+    docker build -t deepspeech-server:latest .
 
+    cd gpu && \
+    docker build -t deepspeech-server:latest-gpu .