CHAE: Fine-Grained Controllable Story Generation with Characters, Actions and Emotions (COLING 2022)
This repository will be continuously improved.
- accelerate >= 0.12.0
- datasets >= 1.8.0
- sentencepiece != 0.1.92
- protobuf
- sacrebleu >= 1.4.12
- py7zr
- torch >= 1.3
The checkpoint can be dowloaded from here.
Inference stories with test data.
sh run_inference.sh
Arguments include:
--device
. The gpu id you want to use.--model_name_or_path
. The checkpoint path would be loaded.--num_beams
. The beam size.--do_sample
. Whether use sampling decoding.--top_k
.--temperature
.
The training steps will be added soon.