Last active
August 20, 2020 02:17
-
-
Save robgon-art/49096fd0c7f5ceea4f97e9d9c9594f69 to your computer and use it in GitHub Desktop.
Train GPT-2
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
# Use TensorFlow 1.15 | |
%tensorflow_version 1.x | |
# Install GPT-2, download the medium model, and start the session | |
!pip install -q gpt-2-simple | |
import gpt_2_simple as gpt2 | |
model = "774M" # 124M 355M 774M 1558M | |
gpt2.download_gpt2(model_name=model) | |
sess = gpt2.start_tf_sess() | |
# Run the fine-tuning for 10,000 steps | |
gpt2.finetune(sess, dataset='story_plots.txt', model_name=model, steps=10000, | |
restore_from='fresh', run_name='run1', print_every=10, sample_every=1000, | |
save_every=10000) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment