From 26d8955cc5ffe865bf83249c125e65f7f72781d8 Mon Sep 17 00:00:00 2001
From: Sharath Turuvekere Sreenivas <sharatht@nvidia.com>
Date: Fri, 8 Oct 2021 01:30:31 -0700
Subject: [PATCH] [PyT/TF/TF2] remove install html2text requirement for NLP
 models

---
 PyTorch/LanguageModeling/BERT/Dockerfile                      | 2 +-
 PyTorch/LanguageModeling/BERT/requirements.txt                | 1 -
 TensorFlow/LanguageModeling/BERT/Dockerfile                   | 2 +-
 TensorFlow/LanguageModeling/BERT/requirements.txt             | 1 -
 TensorFlow2/LanguageModeling/BERT/Dockerfile                  | 1 -
 TensorFlow2/LanguageModeling/BERT/run.sub                     | 1 -
 TensorFlow2/LanguageModeling/ELECTRA/Dockerfile               | 2 +-
 TensorFlow2/LanguageModeling/ELECTRA/scripts/docker/launch.sh | 2 +-
 8 files changed, 4 insertions(+), 8 deletions(-)

diff --git a/PyTorch/LanguageModeling/BERT/Dockerfile b/PyTorch/LanguageModeling/BERT/Dockerfile
index 0e1e0a53..777ef2dd 100755
--- a/PyTorch/LanguageModeling/BERT/Dockerfile
+++ b/PyTorch/LanguageModeling/BERT/Dockerfile
@@ -33,7 +33,7 @@ RUN pip install /workspace/install/python/tensorrtserver*.whl
 WORKDIR /workspace/bert
 RUN pip install --upgrade --no-cache-dir pip \
  && pip install --no-cache-dir \
- tqdm boto3 requests six ipdb h5py html2text nltk progressbar onnxruntime \
+ tqdm boto3 requests six ipdb h5py nltk progressbar onnxruntime \
  git+https://github.com/NVIDIA/dllogger wget
 
 RUN apt-get install -y iputils-ping
diff --git a/PyTorch/LanguageModeling/BERT/requirements.txt b/PyTorch/LanguageModeling/BERT/requirements.txt
index a82a2aca..199a50f5 100755
--- a/PyTorch/LanguageModeling/BERT/requirements.txt
+++ b/PyTorch/LanguageModeling/BERT/requirements.txt
@@ -8,7 +8,6 @@ six
 ipdb
 #Data processing
 h5py
-html2text
 nltk
 progressbar
 #Others
diff --git a/TensorFlow/LanguageModeling/BERT/Dockerfile b/TensorFlow/LanguageModeling/BERT/Dockerfile
index 1ed83c35..c063e543 100644
--- a/TensorFlow/LanguageModeling/BERT/Dockerfile
+++ b/TensorFlow/LanguageModeling/BERT/Dockerfile
@@ -4,7 +4,7 @@ FROM ${FROM_IMAGE_NAME}
 
 RUN apt-get update && apt-get install -y pbzip2 pv bzip2 libcurl4 curl libb64-dev
 RUN pip install --upgrade pip
-RUN pip install toposort networkx pytest nltk tqdm html2text progressbar pynvml
+RUN pip install toposort networkx pytest nltk tqdm progressbar pynvml
 RUN pip --no-cache-dir --no-cache install git+https://github.com/NVIDIA/dllogger wget
 
 WORKDIR /workspace
diff --git a/TensorFlow/LanguageModeling/BERT/requirements.txt b/TensorFlow/LanguageModeling/BERT/requirements.txt
index 04809c0d..fef53507 100644
--- a/TensorFlow/LanguageModeling/BERT/requirements.txt
+++ b/TensorFlow/LanguageModeling/BERT/requirements.txt
@@ -5,5 +5,4 @@ networkx
 pytest
 nltk
 tqdm
-html2text
 progressbar
diff --git a/TensorFlow2/LanguageModeling/BERT/Dockerfile b/TensorFlow2/LanguageModeling/BERT/Dockerfile
index 12129d5b..f5d419cf 100644
--- a/TensorFlow2/LanguageModeling/BERT/Dockerfile
+++ b/TensorFlow2/LanguageModeling/BERT/Dockerfile
@@ -41,7 +41,6 @@ RUN pip3 install \
   tensorflow_hub \
   pynvml \
   wget \
-  html2text \
   progressbar \
   git+https://github.com/NVIDIA/dllogger
 
diff --git a/TensorFlow2/LanguageModeling/BERT/run.sub b/TensorFlow2/LanguageModeling/BERT/run.sub
index 733dead8..029b45be 100644
--- a/TensorFlow2/LanguageModeling/BERT/run.sub
+++ b/TensorFlow2/LanguageModeling/BERT/run.sub
@@ -67,7 +67,6 @@ PIP_CMD="pip3 install \
   tensorflow_hub \
   pynvml \
   wget \
-  html2text \
   progressbar \
   git+https://github.com/NVIDIA/dllogger"
 
diff --git a/TensorFlow2/LanguageModeling/ELECTRA/Dockerfile b/TensorFlow2/LanguageModeling/ELECTRA/Dockerfile
index 92c71f7b..836324f0 100644
--- a/TensorFlow2/LanguageModeling/ELECTRA/Dockerfile
+++ b/TensorFlow2/LanguageModeling/ELECTRA/Dockerfile
@@ -24,7 +24,7 @@ RUN git clone https://github.com/soskek/bookcorpus.git
 WORKDIR /workspace/electra
 RUN pip install --upgrade --no-cache-dir pip \
  && pip install --no-cache-dir \
- tqdm boto3 requests six ipdb h5py html2text nltk progressbar filelock tokenizers==0.7.0 \
+ tqdm boto3 requests six ipdb h5py nltk progressbar filelock tokenizers==0.7.0 \
  git+https://github.com/NVIDIA/dllogger \
  nvidia-ml-py3==7.352.0
 
diff --git a/TensorFlow2/LanguageModeling/ELECTRA/scripts/docker/launch.sh b/TensorFlow2/LanguageModeling/ELECTRA/scripts/docker/launch.sh
index 190dff7a..b5bd7d60 100644
--- a/TensorFlow2/LanguageModeling/ELECTRA/scripts/docker/launch.sh
+++ b/TensorFlow2/LanguageModeling/ELECTRA/scripts/docker/launch.sh
@@ -26,4 +26,4 @@ docker run -it --rm \
   --privileged \
   -e LD_LIBRARY_PATH='/workspace/install/lib/' \
   -v $PWD:/workspace/electra \
-  electra $CMD
\ No newline at end of file
+  electra $CMD