Skip to content
This repository was archived by the owner on Nov 16, 2023. It is now read-only.

Daden/bugfix improvement and bug fix based on the bug bash feedback#576

Merged
daden-ms merged 7 commits intostagingfrom
daden/bugfix
Mar 27, 2020
Merged

Daden/bugfix improvement and bug fix based on the bug bash feedback#576
daden-ms merged 7 commits intostagingfrom
daden/bugfix

Conversation

@daden-ms
Copy link
Copy Markdown
Contributor

@daden-ms daden-ms commented Mar 23, 2020

Description

  1. pin requests version
  2. enable cpu finetune and prediction in bertsum extractive and abstractive summarization notebook
  3. fixed bugs in the extractive summarization aml notebook
  4. improvement of summarization_evaluation notebook to avoid users input for setting PYTHONPATH and pyrouge path.
  5. remove the use of __init__.py for abstractive summarization unilm to make it consistent with other transformer models

Related Issues

Checklist:

  • My code follows the code style of this project, as detailed in our contribution guidelines.
  • I have added tests.
  • I have updated the documentation accordingly.

@daden-ms daden-ms requested a review from saidbleik March 23, 2020 22:09
@review-notebook-app
Copy link
Copy Markdown

Check out this pull request on  ReviewNB

You'll be able to see Jupyter notebook diff and discuss changes. Powered by ReviewNB.

@daden-ms daden-ms changed the base branch from master to staging March 23, 2020 22:09
"source": [
"## Before you start\n",
"Set QUICK_RUN = True to run the notebook on a small subset of data and a smaller number of steps. If QUICK_RUN = False, the notebook takes about 5 hours to run on a VM with 4 16GB NVIDIA V100 GPUs. Finetuning costs around 1.5 hours and inferecing costs around 3.5 hour. Better performance can be achieved by increasing the MAX_STEPS.\n",
"The notebook is only tested on GPU machines. It's also recommended to run this notebook on GPU machines as it's very computational intensive. Set QUICK_RUN = True to run the notebook on a small subset of data and a smaller number of steps. If QUICK_RUN = False, the notebook takes about 5 hours to run on a VM with 4 16GB NVIDIA V100 GPUs. Finetuning costs around 1.5 hours and inferecing costs around 3.5 hour. Better performance can be achieved by increasing the MAX_STEPS.\n",
Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

computationally intensive

@daden-ms daden-ms merged commit 2bc3203 into staging Mar 27, 2020
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants