metadata

language: en
tags:
  - exbert
license: mit

ColD Fusion BERT uncased model

Finetuned model that aims to be a great base model. It improves over BERT base model (uncased), trained on 35 datasets. Full details at this paper.

Paper Abstract:

Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now, massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources that are only available to well-resourced teams.

In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets, ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.

How to use

Best way to use is to finetune on your own task, but you can also extract features directly. To get the features of a given text in PyTorch:

from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)

and in TensorFlow:

from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)

Evaluation results

Model Recycling

Evaluation on 36 datasets using ibm/ColD-Fusion-bert-base-uncased-itr22-seed0 as a base model yields average score of 75.45 in comparison to 72.20 by bert-base-uncased.

The model is ranked 3rd among all tested models for the bert-base-uncased architecture as of 09/01/2023 Results:

20_newsgroup	ag_news	amazon_reviews_multi	anli	boolq	cb	cola	copa	dbpedia	esnli	financial_phrasebank	imdb	isear	mnli	mrpc	multirc	poem_sentiment	qnli	qqp	rotten_tomatoes	rte	sst2	sst_5bins	stsb	trec_coarse	trec_fine	tweet_ev_emoji	tweet_ev_emotion	tweet_ev_hate	tweet_ev_irony	tweet_ev_offensive	tweet_ev_sentiment	wic	wnli	wsc	yahoo_answers
85.2629	88.8	66.26	47.5	74.2202	78.5714	81.3998	59	78.5333	90.6454	84	92.072	69.7523	84.4081	86.0294	60.7673	82.6923	89.4014	90.3661	89.6811	77.9783	93.6927	55.8824	88.9308	97.2	81	35.884	81.9845	51.2795	69.2602	85.3488	69.4155	65.8307	49.2958	62.5	71.2333

For more information, see: Model Recycling See full evaluation results of this model and many more here When fine-tuned on downstream tasks, this model achieves the following results:

BibTeX entry and citation info

@article{ColDFusion,
  author    = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
  title     = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
  journal   = {CoRR},
  volume    = {abs/2212.01378},
  year      = {2022},
  url       = {https://arxiv.org/abs/2212.01378},
  archivePrefix = {arXiv},
  eprint    = {2212.01378},
}