OK this is what I got working for my Conda Enviroment: just cut and paste it into the file: requirements.txt

g2p-en==2.1.0
inflect==4.1.0
librosa==0.7.2
matplotlib==3.4.2
numba==0.48.0
numpy==1.22.4
pypinyin==0.39.0
pyworld==0.3.2
pycwt==0.3.0a22
praat-parselmouth==0.3.3
PyYAML==5.4.1
scikit-learn==0.24.1
scipy==1.6.3
soundfile==0.10.3.post1
tensorboard==2.13.0
tgt==1.4.4
torch==1.8.1
tqdm==4.46.1
unidecode==1.1.1
pandas==2.0.3
pandas==2.0.3
python-speech-features==0.6
tensorflow==2.13.0
resampy==0.3.1

Python version that works for me is 3.8
you will also need to unzip the tar files and copy the model file to correct path, which will be prompted when fail as to where to put them right.
This is the inference command line that works for me:

python synthesize.py --restore_step 2 --model naive --mode single --text "I am so excited that I got this TTS working" --dataset LJSpeech

I would have like for it to read a text file and create a wav file instead of reading a sentence like this. Does anyone know how to make this change?