LSG-BART-Base-4096-MediaSum Open-Source Model - Supports Long-Text Summarization and Handles Sequences of Length 4096

Lsg Bart Base 4096 Mediasum

Developed by ccdv

BART-base model based on LSG technology, fine-tuned for long text summarization tasks on the MediaSum dataset, supporting sequence processing up to 4096 tokens

Text Generation

Transformers

English#Long Text Summarization #Local Sparse Global Attention #4096 Context Window

Downloads 44

Release Time : 5/29/2022

Model Overview

This model employs local-sparse-global attention mechanisms to handle long sequences, making it suitable for long text summarization generation tasks. It is modified based on the BART-base architecture and fine-tuned on the MediaSum dialogue summarization dataset.

Model Features

Long Sequence Processing Capability

Supports input sequences up to 4096 tokens, efficiently processing long texts through local-sparse-global attention mechanisms

Multi-mode Sparse Attention

Offers various sparse attention modes including local/pooling/strided/block-strided/normalized/LSH for flexible selection

Resource Efficiency Optimization

Allows adjustment of block size (32-256) to balance performance and resource consumption, adapting to different hardware conditions

Model Capabilities

Long Text Summarization Generation

Dialogue Content Summarization

Multi-turn Dialogue Understanding

Use Cases

Media Content Processing

Interview Summarization Generation

Condenses lengthy media interview content into concise summaries

Achieves R1=35.16/R2=18.13/RL=31.54 on the MediaSum test set

Meeting Minutes Processing

Automatic Meeting Minutes Generation

Extracts key points from transcribed texts of lengthy meeting recordings

🚀 ccdv/lsg-bart-base-4096-mediasum

This model is a fine - tuned version of ccdv/lsg-bart-base-4096 on the ccdv/mediasum roberta_prepended dataset, designed for text summarization tasks.

🚀 Quick Start

Transformers >= 4.36.1 This model relies on a custom modeling file, you need to add trust_remote_code=True See #13467

LSG ArXiv paper. Github/conversion script is available at this link.

💻 Usage Examples

Basic Usage

from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, pipeline

tokenizer = AutoTokenizer.from_pretrained("ccdv/lsg-bart-base-4096-mediasum", trust_remote_code=True)
model = AutoModelForSeq2SeqLM.from_pretrained("ccdv/lsg-bart-base-4096-mediasum", trust_remote_code=True)

text = "Replace by what you want."
pipe = pipeline("text2text-generation", model=model, tokenizer=tokenizer, device=0)
generated_text = pipe(
  text, 
  truncation=True, 
  max_length=64, 
  no_repeat_ngram_size=7,
  num_beams=2,
  early_stopping=True
  )

📚 Documentation

Model Performance

This model achieves the following results on the test set:

Length	Sparse Type	Block Size	Sparsity	Connexions	R1	R2	RL	RLsum
4096	Local	256	0	768	35.16	18.13	31.54	32.20
4096	Local	128	0	384	34.16	17.61	30.75	31.41
4096	Pooling	128	4	644	34.52	17.71	31.01	31.67
4096	Stride	128	4	644	35.05	18.11	31.47	32.13
4096	Block Stride	128	4	644	34.72	17.81	31.13	31.82
4096	Norm	128	4	644	34.75	17.86	31.10	31.77
4096	LSH	128	4	644	34.54	17.81	31.05	31.71

With smaller block size (lower ressources):

Length	Sparse Type	Block Size	Sparsity	Connexions	R1	R2	RL	RLsum
4096	Local	64	0	192	32.55	16.66	29.36	30.00
4096	Local	32	0	96	30.98	15.41	27.84	28.46
4096	Pooling	32	4	160	31.84	16.02	28.68	29.30
4096	Stride	32	4	160	32.67	16.68	29.47	30.10
4096	Block Stride	32	4	160	32.51	16.64	29.33	29.94
4096	Norm	32	4	160	32.44	16.48	29.20	29.79
4096	LSH	32	4	160	31.79	16.04	28.67	29.31

Model description

The model relies on Local - Sparse - Global attention to handle long sequences: attn

The model has about ~145 millions parameters (6 encoder layers - 6 decoder layers). The model is warm started from BART - base, converted to handle long sequences (encoder only) and fine tuned.

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 8e-05
train_batch_size: 8
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 32
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 6.0

Generate hyperparameters

The following hyperparameters were used during generation:

dataset_name: ccdv/mediasum
dataset_config_name: roberta_prepended
eval_batch_size: 8
eval_samples: 10000
early_stopping: True
ignore_pad_token_for_loss: True
length_penalty: 2.0
max_length: 128
min_length: 3
num_beams: 5
no_repeat_ngram_size: None
seed: 123

Framework versions

Transformers 4.18.0
Pytorch 1.10.1+cu102
Datasets 2.1.0
Tokenizers 0.11.6

🔧 Technical Details

Model Type: Fine - tuned version of ccdv/lsg-bart-base-4096
Training Data: ccdv/mediasum roberta_prepended
Metrics: Rouge

Property	Details
Model Type	Fine - tuned version of ccdv/lsg-bart-base-4096
Training Data	ccdv/mediasum roberta_prepended
Metrics	Rouge

Intended uses & limitations

More information needed

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