This checkpoint should be loaded into BartForConditionalGeneration.frompretrained. See the BART docs for more information.
Search public pages, research tools, and SAVRN solutions.
Open-weight model · Summarization
by Gabriele Sarti gsarti/it5-base-news-summarization
This repository contains the checkpoint for the IT5 Base model fine-tuned on news summarization on the Fanpage and Il Post corpora as part of the experiments of the paper IT5: Large-scale Text-to-text Pretraining for Italian Language Understanding and…
By Gabriele Sarti, published under apache-2.0, revision 291a73c2d8ea.
This repository contains the checkpoint for the IT5 Base model fine-tuned on news summarization on the Fanpage and Il Post corpora as part of the experiments of the paper IT5: Large-scale Text-to-text Pretraining for Italian Language Understanding and Generation by Gabriele Sarti and Malvina Nissim. A comprehensive overview of other released materials is provided in the gsarti/it5 repository. Refer to the paper for additional details concerning the reported scores and the evaluation approach. Model checkpoints are available for usage in Tensorflow, Pytorch and JAX. They can be used directly with pipelines as: If you use this model in your research, please cite our work as
This repository contains the checkpoint for the IT5 Base model fine-tuned on news summarization on the Fanpage and Il Post corpora as part of the experiments of the paper IT5: Large-scale Text-to-text Pretraining for Italian Language Understanding and Generation by Gabriele Sarti and Malvina Nissim.
A comprehensive overview of other released materials is provided in the gsarti/it5 repository. Refer to the paper for additional details concerning the reported scores and the evaluation approach.
Model checkpoints are available for usage in Tensorflow, Pytorch and JAX. They can be used directly with pipelines as:
from transformers import pipelines
newsum = pipeline("summarization", model='it5/it5-base-news-summarization')
newsum("Dal 31 maggio è infine partita la piattaforma ITsART, a più di un anno da quando – durante il primo lockdown – il ministro della Cultura Dario Franceschini ne aveva parlato come di «una sorta di Netflix della cultura», pensata per «offrire a tutto il mondo la cultura italiana a pagamento». È presto per dare giudizi definitivi sulla piattaforma, e di certo sarà difficile farlo anche più avanti senza numeri precisi. Al momento, l’unica cosa che si può fare è guardare com’è fatto il sito, contare quanti contenuti ci sono (circa 700 “titoli”, tra film, documentari, spettacoli teatrali e musicali e altri eventi) e provare a dare un giudizio sul loro valore e sulla loro varietà. Intanto, una cosa notata da più parti è che diversi contenuti di ITsART sono a pagamento sulla piattaforma sebbene altrove, per esempio su RaiPlay, siano invece disponibili gratuitamente.")
>>> [{"generated_text": "ITsART, la Netflix della cultura italiana, parte da maggio. Film, documentari, spettacoli teatrali e musicali disponibili sul nuovo sito a pagamento."}]
or loaded using autoclasses:
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
tokenizer = AutoTokenizer.from_pretrained("it5/it5-base-news-summarization")
model = AutoModelForSeq2SeqLM.from_pretrained("it5/it5-base-news-summarization")
If you use this model in your research, please cite our work as:
@article{sarti-nissim-2022-it5,
title={{IT5}: Large-scale Text-to-text Pretraining for Italian Language Understanding and Generation},
author={Sarti, Gabriele and Nissim, Malvina},
journal={ArXiv preprint 2203.03759},
url={https://arxiv.org/abs/2203.03759},
year={2022},
month={mar}
}
11 files, 3.0 GB in total. The weights are 3 files totalling 3.0 GB in bin, h5, msgpack.
| File | Type | Size | SHA-256 |
|---|---|---|---|
| flax_model.msgpack | Weights | 990.2 MB | 92cdae6ea47c |
| pytorch_model.bin | Weights | 990.3 MB | 1c0524edc93a |
| tf_model.h5 | Weights | 990.7 MB | c635f7c8d6ad |
| config.json | Configuration | 907 B | — |
| special_tokens_map.json | Configuration | 1.8 KB | — |
| README.md | Documentation | 8.7 KB | — |
| events.out.tfevents.1642260324.t1v-n-9e38d79e-w-0.1294614.0.v2 | Other | 3.3 MB | edaeee56129c |
| .gitattributes | Repository | 1.2 KB | — |
| spiece.model | Tokenizer | 4.3 MB | ef78f86560d8 |
| tokenizer.json | Tokenizer | 1.0 MB | — |
| tokenizer_config.json | Tokenizer | 1.9 KB | — |
Released by Gabriele Sarti through its official repository on Hugging Face. Read the license.
Each result is shown as reported, with the conditions its reporter stated. None is a SAVRN measurement. A comparison lines two results up only when their configuration, unit and setup are all stated and identical.
| Benchmark | Conditions | Result | Reported by | Revision | Date |
|---|---|---|---|---|---|
| NewsSum-IT | Task News SummarizationMetric Test Rouge1Comparison conditions not established | 0.339 | gsarti Publisher reported |
Evaluated revision not stated | — |
| NewsSum-IT | Task News SummarizationMetric Test Rouge2Comparison conditions not established | 0.16 | gsarti Publisher reported |
Evaluated revision not stated | — |
| NewsSum-IT | Task News SummarizationMetric Test RougeLComparison conditions not established | 0.263 | gsarti Publisher reported |
Evaluated revision not stated | — |
| Precision | Weights in memory |
|---|---|
| As published | 3.0 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Yes. it5-base-news-summarization is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.
This checkpoint should be loaded into BartForConditionalGeneration.frompretrained. See the BART docs for more information.
Original TF 1 code here Authors: Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu on Dec 18, 2019 The following is copied from the authors' README. We train a pegasus model with sampled gap sentence ratios on both C4 and HugeNews, and stochastically sample important sentences. The updated the results are reported in this table. The "Mixed & Stochastic" model has the following changes: - trained on both C4 and HugeNews (dataset mixture is weighted by their number of examples). - trained for 1.5M instead of 500k (we observe slower convergence on pretraining perplexity). - the model uniformly sample a gap sentence ratio between 15% and 45%. - importance sentences are sampled using a…
This model was trained using Amazon SageMaker and the new Hugging Face Deep Learning container.
This checkpoint should be loaded into BartForConditionalGeneration.frompretrained. See the BART docs for more information.
This repository contains the mT5 checkpoint finetuned on the 45 languages of XL-Sum dataset. For finetuning details and scripts, see the paper and the official repository. Scores on the XL-Sum test sets are as follows: Language | ROUGE-1 / ROUGE-2 / ROUGE-L Amharic | 20.0485 / 7.4111 / 18.0753 Arabic | 34.9107 / 14.7937 / 29.1623 Azerbaijani | 21.4227 / 9.5214 / 19.3331 Bengali | 29.5653 / 12.1095 / 25.1315 Burmese | 15.9626 / 5.1477 / 14.1819 Chinese (Simplified) | 39.4071 / 17.7913 / 33.406 Chinese (Traditional) | 37.1866 / 17.1432 / 31.6184 English | 37.601 / 15.1536 / 29.8817 French | 35.3398 / 16.1739 / 28.2041 Gujarati | 21.9619 / 7.7417 / 19.86 Hausa | 39.4375 / 17.6786 / 31.6667…