LongT5-base-global-mediasum Open-source Text Summarization Model - Free to Use, Precise Extraction of Text Points

Longt5 Base Global Mediasum

Developed by nbroad

A text summarization model fine-tuned based on google/long-t5-tglobal-base, performing well on XSum and CNN/DailyMail datasets.

Text Generation

Transformers

#Long text summarization #News summarization #High ROUGE scores

Downloads 19

Release Time : 7/21/2022

Model Overview

This model is a text summarization model based on the LongT5 architecture, specifically optimized for long-text summarization tasks, capable of generating concise and accurate summaries.

Model Features

Long text processing capability

Based on the LongT5 architecture, specifically optimized for handling long text inputs, suitable for document-level summarization tasks.

Performance on multiple datasets

Evaluated and validated on multiple standard summarization datasets including XSum and CNN/DailyMail.

Efficient fine-tuning

Fine-tuned based on the pre-trained LongT5 model, achieving significant performance improvements.

Model Capabilities

Text summarization

Long text processing

English document understanding

Use Cases

News summarization

News article summarization

Automatically compressing long news articles into concise summaries

Achieved a ROUGE-1 score of 21.15 on the CNN/DailyMail test set

Document summarization

Long document summarization

Generating key point summaries for technical documents, reports, and other long texts

🚀 longt5-mediasum

This model, longt5-mediasum, is a fine - tuned version of the google/long - t5 - tglobal - base model. It is designed for summarization tasks and achieves certain performance on evaluation sets.

🚀 Quick Start

This model is a fine - tuned version of [google/long - t5 - tglobal - base](https://huggingface.co/google/long - t5 - tglobal - base) on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 2.0129

📚 Documentation

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 5e - 05
train_batch_size: 12
eval_batch_size: 16
seed: 42
optimizer: Adam with betas=(0.9, 0.999) and epsilon = 1e - 08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 3.0

Training results

Training Loss	Epoch	Step	Validation Loss
2.66	1.0	1667	2.0643
2.472	2.0	3334	2.0241
2.3574	3.0	5001	2.0129

Framework versions

Transformers 4.20.1
Pytorch 1.11.0a0 + 17540c5
Datasets 2.3.2
Tokenizers 0.12.1

📄 License

The model is licensed under CC - BY - NC - SA - 4.0.

📊 Model Index

name: longt5 - mediasum results:
- task: type: summarization name: Summarization dataset: name: xsum type: xsum config: default split: test metrics:
  - type: rouge value: 22.7044 name: ROUGE - 1 verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiMzMyZDQwNWY4M2I2YTM2ZjY1MWFhMWQ4NDAxNTg4N2Y2MDE4NGMyMjJiMTkyYzcwNzg2MWU0Y2YyNGU5MDM4OCIsInZlcnNpb24iOjF9.2MQgk8ReWfDWi98sJyIuEqSXUxumfMjUZl1raOfHiK4TP45FwLKzc5Un - dLbS7yb8L6rEGEKJMGbiFKeapzJBw
  - type: rouge value: 5.616 name: ROUGE - 2 verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiNWIxYTFmNGUzMjMwODA4NTlkZjk4MTllYmVmOGJlYTI3NzA3MTQzY2RkMWJlOTdiM2ZkZWVkNDk0ZjhlMzQxNSIsInZlcnNpb24iOjF9.XwtjOwmYE - NxbH6kOYxDgMpXTNL6XwiSGGi4cdtBu7O0cdjNUh0dnfihvEvJegMM8cnQaQeu47g0zxuLP1LQBQ
  - type: rouge value: 18.0111 name: ROUGE - L verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiYzYzNGExYTc5MmZiMTNhYjM1ZDhjZDNiYTNlMDZhMzdiMjRhNzk2NWI1ZjRkMDlmMGUwOGMxYzhiYzE0MmFhOSIsInZlcnNpb24iOjF9.nPERc8JZ4VaUTSkUZS7vwbZecdWn1uIpIBN6zFTFLQnfuCIP7ZlgDegUGEaehoERJubHVf60Z24I0aYxq2q0DQ
  - type: rouge value: 18.1554 name: ROUGE - LSUM verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiYjMwNTg3MWIxZGQxMzdhMGM0NTg5NDZlYWZhZThjYTUzZjcwODUyMDgwNTY0YWRlOTI4Y2NlOTY3Mzc2ZjU2MCIsInZlcnNpb24iOjF9.syWVW4_SlY4JDqE96kpz3qGXGfhF8wezAj87p4RzQpJ7Fp7ie1 - v5EgTOy4ZFMdtemQCyqiQYAQwcVckttrLBg
  - type: loss value: 2.1656227111816406 name: loss verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiM2UyNWMwNmQwOWU4YjA3MzMzYWQzMTgwYzJjMjNkYTlhMmQ4NTAzMzE1YzlkMWRiZDEzYzJlMGJmMzJkZjEwMyIsInZlcnNpb24iOjF9.8FY9EcQPrS4KXJyJBuX2uLD5VxLlGxRC33qxPuyAg5W9WsXm6XSjb0ZCJ8T - Xl3YlwV9bSL6UMq4MPleEjiNCA
  - type: gen_len value: 18.3527 name: gen_len verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiM2U0Y2E5NzU1MTNhOGRkYmNhNjI2NGQ1NjJiMDc1MmI5NTE2MDJjMDA4NjMxODcyMDJhN2FhZWI0ZThhNzZhNyIsInZlcnNpb24iOjF9.e3yQ--2AjvVzCM - HXzvB - lN4g6eMxqjQ1AyhFqftit8kFmoKnLqoa2ytoyOZmlnXvkCRmpgfSAYq - tnceINLDw
- task: type: summarization name: Summarization dataset: name: cnn_dailymail type: cnn_dailymail config: 3.0.0 split: test metrics:
  - type: rouge value: 21.1522 name: ROUGE - 1 verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiNTkwZWEzYzEyMjdjZTA5MjBlOTM5MDI3NWEzZWJkOWEwNjIwMGNkNjU4NGM1YjVhZjIxNjI2NzJkM2NhZjc4MCIsInZlcnNpb24iOjF9.f9REqBITFW - Y1GjOmq - e1zUI0qK2MQxEsCe9NQZbCFNeygvdhD2CkTslUbTIXpuDcA_bWAJsNiL2VAMSL1ZDAw
  - type: rouge value: 8.1315 name: ROUGE - 2 verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiMjg3MDYxMjVmNGIzYzlmMTBkNjcxODA1MDliNjBlY2I5ODQ4YzQwY2RmMDBlYmYwMTgyOThlMjQ0NjY2MDdhNSIsInZlcnNpb24iOjF9.mJwLahXr - lvwEv1L9PlFfV0dqxN8m3OciTV9wdoENBaj7JLtBnev - LL8oivi - fBRY - WcVpUGpjyx1fQxwLF2Aw
  - type: rouge value: 16.6625 name: ROUGE - L verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiNmUyOTc0MWVmYTQwY2ViMDNjYWE2Mzc3OWFkNGI1YzUxMmU3YjgxMjMyMmY5ZjIzOGRjYmY0OGIxZTMzN2I5MSIsInZlcnNpb24iOjF9.1gGyxK6SIY4AcR_6 - 7An94GyRx4q_Fj4Zmm8zNSE4HqHq3dFUmWfFYGZVgLUOswMr03di63QqmFyVnjW0TqmBw
  - type: rouge value: 19.3603 name: ROUGE - LSUM verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiMGRmMzkxYzRkMTA4ZDYzZWE2MDM4NWU1NjFiYWVjOGVhMjc1N2YzZWI2MmY2OTY3MjVkYmExZGY1ZTcyYjIwOCIsInZlcnNpb24iOjF9.vVMF4qaE5Wg9vZuBVOtcR89L_NldNk49sgEyAFVvx_hhjOx5nZvygseHcwqemMjUlrTzOuRD15x - XbH - 62qhBw
  - type: loss value: 1.899269700050354 name: loss verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiOTU5ZDUzYTc0OTc1YWUyNjY0NDU3NDhlODljNGNlN2E4ZTQzYjNhMWM0NDYyNTE5Yjk4ZDE3ZTNjZDdmYmQyYSIsInZlcnNpb24iOjF9.SuRSjdfuA_NE3WHglrxbIaPZ - Hg9 - F6jB37Cb_MPIME2h3a30qafO9tFmxUsR_JCDNfuILwrGEHT1PJmq4NQDQ
  - type: gen_len value: 17.853 name: gen_len verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiNWU2MDRmMTliNzhkMzdmNmM2NTA2Y2NmYjVhOWJjYmQzZTUxYWM1MGYzMzBjODdiMzY2YjMzYjE4Y2FkNjExZCIsInZlcnNpb24iOjF9.J5NYms5Hg9iLFgTq0YqYXP86H9A0PClmG2eLi09BBs9L8F - PhQhiuzzCeD1HA2fC_6mtLyEv6PPLZCr6Ax7DCQ

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご