src/requirements.txt

# for generate (gradio server) and finetune
datasets==2.13.0
sentencepiece==0.1.99
gradio==3.41.2
huggingface_hub==0.16.4
appdirs==1.4.4
fire==0.5.0
docutils==0.20.1
torch==2.0.1; sys_platform != "darwin" and platform_machine != "arm64"
evaluate==0.4.0
rouge_score==0.1.2
sacrebleu==2.3.1
scikit-learn==1.2.2
# optional (need to uncomment code in gradio_runner.py for import of better_profanity)
# alt-profanity-check==1.2.2
# better-profanity==0.7.0
numpy==1.24.3
pandas==2.0.2
matplotlib==3.7.1
loralib==0.1.1
bitsandbytes==0.41.1
accelerate==0.22.0
peft==0.5.0
transformers==4.33.1
tokenizers==0.13.3
APScheduler==3.10.1

# optional for generate
pynvml==11.5.0
psutil==5.9.5
boto3==1.26.101
botocore==1.29.101

# optional for finetune
tensorboard==2.13.0
neptune==1.2.0

# for gradio client
gradio_client==0.5.0
beautifulsoup4==4.12.2
markdown==3.4.3

# data and testing
pytest==7.2.2
pytest-xdist==3.2.1
nltk==3.8.1
textstat==0.7.3
# pandoc==2.3
pypandoc==1.11; sys_platform == "darwin" and platform_machine == "arm64"
pypandoc_binary==1.11; platform_machine == "x86_64"
pypandoc_binary==1.11; sys_platform == "win32"
python-magic-bin==0.4.14; sys_platform == "win32"
openpyxl==3.1.2
lm_dataformat==0.0.20
bioc==2.0

# falcon
einops==0.6.1
instructorembedding==1.0.1

# for gpt4all .env file, but avoid worrying about imports
python-dotenv==1.0.0

text-generation==0.6.0
# for tokenization when don't have HF tokenizer
tiktoken==0.4.0

requests>=2.31.0
urllib3>=1.26.16
filelock>=3.12.2
joblib>=1.3.1
tqdm>=4.65.0
tabulate>=0.9.0
packaging>=23.1