QQ Browser 2021 AI Algorithm Competition Track 1 1st Place Program

Overview

2021_QQ_AIAC_Tack1_1st

QQ浏览器2021AI算法大赛赛道一 第1名 方案

paper :

环境

python==3.7.10
torch==1.7.1
transformers==4.5.1
pretrain 需要显存>=24GB 内存>=100GB

数据下载

(1) 视频数据集
视频数据集在官网下载 https://algo.browser.qq.com/
预期主办方会开源数据集,开源后会将地址补上
下载后放到 ./input/data 文件夹
tag_list 为标签的 top1w,官方 baseline 中提供,放到同一文件夹

(2) 预训练模型
预训练模型使用了 https://huggingface.co/hfl/chinese-roberta-wwm-ext-large
请使用 python3 -u download_pretrain_model.py 下载

步骤代码

(1) 预训练 + finetune
脚本命令:sh train.sh
时间算力:单模在 1 张 a100 上大约需要 pretrain(2 day),finetune(2 hour)
输出文件:每个单模的 checkpoint 保存在 jobN/model_finetune_1.pth
备注:各个单模间没有前后依赖关系,每个任务需要一张单卡,有多卡可以并行训练各个单模

(2) 代码结构说明
download_pretrain_model.py : 下载预训练模型的脚本
ensemble.py : 融合的脚本
job1-job6 : 六个模型训练任务,其文件结构完全一致,各 job 之间主要差别在预训练设置上
注:job1在赛后额外补充了一些代码注释
jobN/pretrain.py 预训练脚本
jobN/finetune.py finetune脚本
jobN/data 数据预处理部分,包含 dataset、mask token 等
jobN/config 包含 pretrain 与 finetune 的一些超参配置
jobN/qqmodel/qq_uni_model.py 模型定义

简介

简要介绍的 ppt 请参考 Introduction.pdf

模型简介

多模态模型结构与参数量和 Bert-large 一致,
layer=24, hidden_size=1024, num_attention_heads=16。
其输入为[CLS] Video_frame [SEP] Video_title [SEP]。
frame_feature 通过 fc 降维为 1024 维,与 text 的 emb 拼接。
Input_emb -> TransformerEncoder * 24 -> Pooling -> Fc -> Video_emb

预训练

预训练采用了 Tag classify, Mask language model, Mask frame model 三个任务

(1) Video tag classify 任务
tag 为人工标注的视频标签,pointwise 和 pairwise 数据集合中提供。
和官方提供的 baseline 一致,我们采用了出现频率前1w 的tag 做多标签分类任务。
Bert 最后一层的 [CLS] -> fc 得到 tag 的预测标签,与真实标签计算 BCE loss

(2) Mask language model 任务
与常见的自然语言处理 mlm 预训练方法相同,对 text 随机 15% 进行 mask,预测 mask 词。
多模态场景下,结合视频的信息预测 mask 词,可以有效融合多模态信息。

(3) Mask frame model 任务
对 frame 的随机 15% 进行 mask,mask 采用了全 0 的向量填充。
考虑到 frame 为连续的向量,难以类似于 mlm 做分类任务。
借鉴了对比学习思路,希望 mask 的预测帧在整个 batch 内的所有帧范围内与被 mask 的帧尽可能相似。
采用了 Nce loss,最大化 mask 帧和预测帧的互信息

(4) 多任务联合训练
预训练任务的 loss 采用了上述三个任务 loss 的加权和,
L = L(tag) * 1250 / 3 + L(mlm) / 3.75 + L(mfm) / 9
tag 梯度量级比较小,因此乘以了较大的权重。
注:各任务合适的权重对下游 finetune 的效果影响比较大。

(5) 预训练 Setting
初始化:bert 初始化权重来自于在中文语料预训练过的开源模型 https://huggingface.co/hfl/chinese-roberta-wwm-ext-large
数据集:预训练使用了 pointwise 和 pairwise 集合,部分融合模型中加上了 test 集合(只有 mlm 和 mfm 任务)
超参:batch_size=128, epoch=40, learning_rate=5e-5, scheduler=warmup_with_cos_decay, warum_ratio=0.06
注:预训练更多的 epoch 对效果提升比较大,从10 epoch 提升至 20 epoch 对下游任务 finetune 效果提升显著。

Finetune

(1) 下游任务
视频 pair 分别通过 model 得到 256维 embedding,两个 embedding 的 cos 相似度与人工标注标签计算 mse

(2) Finetune header
实验中发现相似度任务中,使用 mean_pooling 或者 attention_pooling 聚合最后一层 emb 接 fc 层降维效果较好。

(3) Label normalize
评估指标为 spearman,考查预测值和实际值 rank 之间的相关性,因此对人工标注 label 做了 rank 归一化。
即 target = scipy.stats.rankdata(target, 'average')

(4) Finetune Setting
数据集:训练集使用了 pairwise 中 (id1%5!=0) | (id2%5 !=0) 的部分约 6.5w,验证集使用了(id1%5==0) & (id2%5==0) 的部分约 2.5k
超参:batch_size=32, epoch=10, learning_rate=1e-5, scheduler=warmup_with_cos_decay, warum_ratio=0.06

Ensemble

(1) 融合的方法
采用了 weighted concat -> svd 降维 方法进行融合。实验中发现这种方法降维效果折损较小。
concat_vec = [np.sqrt(w1) * emb1, np.sqrt(w2) * emb2, np.sqrt(w3) * emb3 ...]
svd_vec = SVD(concat_vec, 256)

(2) 融合的模型
最终的提交融合了六个模型。 模型都使用了 bert-large 这种结构,均为迭代过程中产出的模型,各模型之间只有微小的 diff,各个模型加权权重均为 1/6。
下面表格中列出了各模型的diff部分,验证集mse,验证集spearman

jobid ensemble-weight detail val-spearman val-mse
job1 1/6 base 0.886031 0.028813
job2 1/6 预训练tag分类任务为mean_pooling+fc 0.884257 0.029493
job3 1/6 预训练任务无 mfm 0.883843 0.029248
job4 1/6 预训练数据为 (point + pair)shuf-40epoch => pair-5epoch 0.885397 0.029059
job5 1/6 预训练数据为 (point-shuf => pair-shuf => test-shuf)-32epoch 0.885795 0.028866
job6 1/6 预训练 mlm/mfm mask概率调整为25% 0.886289 0.029039

(3) 单模型的效果与融合的效果
单模的测试集成绩约在 0.836
融合两个模型在 0.845
融合三个模型在 0.849
融合五个模型在 0.852

Code for the paper "SmoothMix: Training Confidence-calibrated Smoothed Classifiers for Certified Robustness" (NeurIPS 2021)

SmoothMix: Training Confidence-calibrated Smoothed Classifiers for Certified Robustness (NeurIPS2021) This repository contains code for the paper "Smo

Jongheon Jeong 17 Dec 27, 2022
BOVText: A Large-Scale, Multidimensional Multilingual Dataset for Video Text Spotting

BOVText: A Large-Scale, Bilingual Open World Dataset for Video Text Spotting Updated on December 10, 2021 (Release all dataset(2021 videos)) Updated o

weijiawu 47 Dec 26, 2022
Deep GPs built on top of TensorFlow/Keras and GPflow

GPflux Documentation | Tutorials | API reference | Slack What does GPflux do? GPflux is a toolbox dedicated to Deep Gaussian processes (DGP), the hier

Secondmind Labs 107 Nov 02, 2022
Code for the CVPR 2021 paper: Understanding Failures of Deep Networks via Robust Feature Extraction

Welcome to Barlow Barlow is a tool for identifying the failure modes for a given neural network. To achieve this, Barlow first creates a group of imag

Sahil Singla 33 Dec 05, 2022
AttentionGAN for Unpaired Image-to-Image Translation & Multi-Domain Image-to-Image Translation

AttentionGAN-v2 for Unpaired Image-to-Image Translation AttentionGAN-v2 Framework The proposed generator learns both foreground and background attenti

Hao Tang 530 Dec 27, 2022
🕵 Artificial Intelligence for social control of public administration

Non-tech crash course into Operação Serenata de Amor Tech crash course into Operação Serenata de Amor Contributing with code and tech skills Supportin

Open Knowledge Brasil - Rede pelo Conhecimento Livre 4.4k Dec 31, 2022
Code for the paper: "On the Bottleneck of Graph Neural Networks and Its Practical Implications"

On the Bottleneck of Graph Neural Networks and its Practical Implications This is the official implementation of the paper: On the Bottleneck of Graph

75 Dec 22, 2022
BYOL for Audio: Self-Supervised Learning for General-Purpose Audio Representation

BYOL for Audio: Self-Supervised Learning for General-Purpose Audio Representation This is a demo implementation of BYOL for Audio (BYOL-A), a self-sup

NTT Communication Science Laboratories 160 Jan 04, 2023
Generate vibrant and detailed images using only text.

CLIP Guided Diffusion From RiversHaveWings. Generate vibrant and detailed images using only text. See captions and more generations in the Gallery See

Clay M. 401 Dec 28, 2022
(ICCV 2021) ProHMR - Probabilistic Modeling for Human Mesh Recovery

ProHMR - Probabilistic Modeling for Human Mesh Recovery Code repository for the paper: Probabilistic Modeling for Human Mesh Recovery Nikos Kolotouros

Nikos Kolotouros 209 Dec 13, 2022
All course materials for the Zero to Mastery Deep Learning with TensorFlow course.

All course materials for the Zero to Mastery Deep Learning with TensorFlow course.

Daniel Bourke 3.4k Jan 07, 2023
This implements the learning and inference/proposal algorithm described in "Learning to Propose Objects, Krähenbühl and Koltun"

Learning to propose objects This implements the learning and inference/proposal algorithm described in "Learning to Propose Objects, Krähenbühl and Ko

Philipp Krähenbühl 90 Sep 10, 2021
LaneAF: Robust Multi-Lane Detection with Affinity Fields

LaneAF: Robust Multi-Lane Detection with Affinity Fields This repository contains Pytorch code for training and testing LaneAF lane detection models i

155 Dec 17, 2022
Accurate 3D Face Reconstruction with Weakly-Supervised Learning: From Single Image to Image Set (CVPRW 2019). A PyTorch implementation.

Accurate 3D Face Reconstruction with Weakly-Supervised Learning: From Single Image to Image Set —— PyTorch implementation This is an unofficial offici

Sicheng Xu 833 Dec 28, 2022
Bib-parser - Convenient script to parse .bib files with the ACM Digital Library like metadata

Bib Parser Convenient script to parse .bib files with the ACM Digital Library li

Mehtab Iqbal (Shahan) 1 Jan 26, 2022
Explaining in Style: Training a GAN to explain a classifier in StyleSpace

Explaining in Style: Official TensorFlow Colab Explaining in Style: Training a GAN to explain a classifier in StyleSpace Oran Lang, Yossi Gandelsman,

Google 197 Nov 08, 2022
Code repository for the paper: Hierarchical Kinematic Probability Distributions for 3D Human Shape and Pose Estimation from Images in the Wild (ICCV 2021)

Hierarchical Kinematic Probability Distributions for 3D Human Shape and Pose Estimation from Images in the Wild Akash Sengupta, Ignas Budvytis, Robert

Akash Sengupta 149 Dec 14, 2022
Phonetic PosteriorGram (PPG)-Based Voice Conversion (VC)

ppg-vc Phonetic PosteriorGram (PPG)-Based Voice Conversion (VC) This repo implements different kinds of PPG-based VC models. Pretrained models. More m

Liu Songxiang 227 Dec 28, 2022
Vit-ImageClassification - Pytorch ViT for Image classification on the CIFAR10 dataset

Vit-ImageClassification Introduction This project uses ViT to perform image clas

Kaicheng Yang 4 Jun 01, 2022
TyXe: Pyro-based BNNs for Pytorch users

TyXe: Pyro-based BNNs for Pytorch users TyXe aims to simplify the process of turning Pytorch neural networks into Bayesian neural networks by leveragi

87 Jan 03, 2023