Go to file
2024-05-16 01:26:17 +02:00
dev-0 PerplexityHashed: 990 2024-05-16 01:26:17 +02:00
src PerplexityHashed: 990 2024-05-16 01:26:17 +02:00
test-A Final model 2024-04-28 00:56:04 +02:00
train Final model 2024-04-28 00:56:04 +02:00
.gitignore Final model 2024-04-28 00:56:04 +02:00
config.txt Init 2021-08-23 10:55:15 +02:00
geval Final model 2024-04-28 00:56:04 +02:00
in-header.tsv Final model 2024-04-28 00:56:04 +02:00
out-header.tsv Final model 2024-04-28 00:56:04 +02:00
README.md Final model 2024-04-28 00:56:04 +02:00
run.py PerplexityHashed: 990 2024-05-16 01:26:17 +02:00

Challenging America word-gap prediction

This task is to predict the word-gap between two sentences.

Evaluation

PerplexityHashed is the metric so to check the performance of the model. The lower the perplexity, the better the model. To run evaluation run the following command:

./geval --metric PerplexityHashed --test-name dev-0

Perplexity calculated on dev-0 is equal 981.69