Create Lambda layer of open source machine learning model

I want a lambda layer created and implemented of a open source machine learning model.
The open source model is available here (https://github.com/NVIDIA/sentiment-discovery)


Deliverables
Phase 1
- Verify that the model works and report back accuracy
- Run 2 provided sample data set (provided as CSV’s) and share results (Can be done on your local machine as the model is pre trained)

Phase 2
Create a lambda layer of model
Move code to a private bitbucket repo and create a script as per code below
so layer deployment can be automated (already have automatic code deploy just need the layer to be packaged up as per bellow script)


Provide Node 8 lambda endpoint script (with test) so that data can be sent to the model and
A) A csv that the models scores first column
B) A string of text
C) A word by word heatmap (as per heatmap pull request in repository)

Expected return is
A) A CSV that can be downloaded
B) A string that can be logged with console.log so I can store it in a database
C) The score for the word (as per the heatmap pull request)



Please provide cost for each of the phases as phase 2 will not start unless the results from phase 1 are meeting requirements




git clone --depth=1 https://github.com/alixaxel/chrome-aws-lambda.git && \
cd chrome-aws-lambda && \
brotli --decompress --rm bin/chromium-*.br && \
npm pack && \
mkdir -p nodejs/node_modules/chrome-aws-lambda/ && \
tar --directory nodejs/node_modules/chrome-aws-lambda/ --extract --file chrome-aws-lambda-*.tgz --strip-components=1 && \
rm chrome-aws-lambda-*.tgz && \
npm install puppeteer-core --no-bin-links --no-optional --no-package-lock --no-save --no-shrinkwrap --prefix nodejs/ && \
zip -9 --filesync --move --recurse-paths ../XXX/layers/chrome-aws-lambda.zip nodejs/

PS
Don't bid on this project unless you have demonstrable experience with machine learning