|
--- |
|
datasets: |
|
- imdb |
|
- cornell_movie_dialogue |
|
- MIT Movie |
|
|
|
language: |
|
- English |
|
|
|
thumbnail: |
|
|
|
tags: |
|
- roberta |
|
- roberta-base |
|
- question-answering |
|
- qa |
|
- movies |
|
|
|
license: cc-by-4.0 |
|
|
|
--- |
|
# roberta-base + DAPT + Task Transfer for Domain-Specific QA |
|
|
|
Objective: |
|
This is Roberta Base with Domain Adaptive Pretraining on Movie Corpora --> Then trained for the NER task using MIT Movie Dataset --> Then a changed head to do the SQuAD Task. This makes a QA model capable of answering questions in the movie domain, with additional information coming from a different task (NER - Task Transfer). |
|
https://huggingface.co/thatdramebaazguy/movie-roberta-base was used as the MovieRoberta. |
|
|
|
``` |
|
model_name = "thatdramebaazguy/movie-roberta-MITmovie-squad" |
|
pipeline(model=model_name, tokenizer=model_name, revision="v1.0", task="question-answering") |
|
``` |
|
|
|
## Overview |
|
**Language model:** roberta-base |
|
**Language:** English |
|
**Downstream-task:** NER --> QA |
|
**Training data:** imdb, polarity movie data, cornell_movie_dialogue, 25mlens movie names, MIT Movie, SQuADv1 |
|
**Eval data:** MoviesQA (From https://github.com/ibm-aur-nlp/domain-specific-QA) |
|
**Infrastructure**: 4x Tesla v100 |
|
**Code:** See [example](https://github.com/adityaarunsinghal/Domain-Adaptation/blob/master/scripts/shell_scripts/movieR_NER_squad.sh) |
|
|
|
## Hyperparameters |
|
``` |
|
Num examples = 88567 |
|
Num Epochs = 3 |
|
Instantaneous batch size per device = 32 |
|
Total train batch size (w. parallel, distributed & accumulation) = 128 |
|
|
|
``` |
|
## Performance |
|
|
|
### Eval on SQuADv1 |
|
- eval_samples = 10790 |
|
- exact_match = 83.0274 |
|
- f1 = 90.1615 |
|
|
|
### Eval on MoviesQA |
|
- eval_samples = 5032 |
|
- exact_match = 51.64944 |
|
- f1 = 65.53983 |
|
|
|
Github Repo: |
|
- [Domain-Adaptation Project](https://github.com/adityaarunsinghal/Domain-Adaptation/) |
|
|
|
--- |
|
|
