Fixing requirements and udpating README

lit_nlp/examples/gcp/Dockerfile

@@ -95,6 +95,8 @@ COPY ./lit_nlp/examples/gcp/model_server_gunicorn_config.py ./
 # TODO(b/353980272): Replace this with a requirements file specific to the GCP
 # exmaple, this should include the core lit-nlp package.
 COPY requirements_core.txt ./
+COPY lit_nlp/examples/prompt_debugging/requirements.txt \
+     lit_nlp/examples/prompt_debugging/requirements.txt
 COPY lit_nlp/examples/gcp/requirements.txt lit_nlp/examples/gcp/requirements.txt
 RUN python -m pip install -r lit_nlp/examples/gcp/requirements.txt
 

lit_nlp/examples/gcp/README.md

@@ -1,5 +1,35 @@
 # Using LLMs in LIT on Google Cloud Platform
 
+## Developing
+
+### Use a virtual environment
+
+```shell
+# Create and activate the virtual environment
+python3 -m venv ~/.venvs/lit-on-gcp
+source ~/.venvs/lit-on-gcp/bin/activate
+
+# Install the requirements and LIT in editable mode
+pip install -f ./lit_nlp/examples/gcp/requirements.txt
+pip install -e .
+
+# Optionally, install tetsing requirements
+pip install -f ./requirements_test.txt
+```
+
+### Build the Docker image
+
+```shell
+docker build -f ./lit_nlp/examples/gcp/Dockerfile -t lit-app:gcp-dev .
+```
+
+### Run GPT-2 in a Docker container
+
+```shell
+# Runs GPT-2 in Keras on Tensorflow
+docker run --rm -p 5432:5432 -e MODEL_CONFIG=gpt2:gpt2_base_en lit-app:gcp-dev
+```
+
 Architectural Notes
 
 *   The `LitApp` HTTP API assumes that inputs will be passed around as

lit_nlp/examples/gcp/requirements.txt

@@ -14,6 +14,7 @@
 # ==============================================================================
 
 -r ../../../requirements_core.txt
+-r ../prompt_debugging/requirements.txt
 
 google-cloud-aiplatform>=1.60.0
 gunicorn>=20.1.0