Train Dense Passage Retriever (DPR) with a single GPU

Overview

Gradient Cached Dense Passage Retrieval

Gradient Cached Dense Passage Retrieval (GC-DPR) - is an extension of the original DPR library.
We introduce Gradient Cache technique which enables scaling batch size of the contrastive form loss and therefore the number of in-batch negatives far beyond GPU RAM limitation. With GC-DPR , you can reproduce the state-of-the-art open Q&A system trained on 8 x 32GB V100 GPUs with a single 11 GB GPU.

To use Gradient Cache in your own project, checkout our GradCache package.

Gradient Cache Technique

  • Retriever training quality depends on its effective batch size.
  • The contrastive form loss (NLL with in-batch negatives) conditions on the entire batch and requires fitting the entire batch into GPU memory.
  • Gradient Cache technique separates back propagation into loss to representation part and representation to encoder model part. It uses an extra forward pass without gradient tracking to compute representations and then store representations' gradients in cache.
  • The cached representation gradients remove data dependency in encoder back propagation and allows updating encoders one sub-batch at a time to fit GPU memory.

Details can be found in our paper Scaling Deep Contrastive Learning Batch Size under Memory Limited Setup , to appeaer at The 6th Workshop on Representation Learning for NLP (RepL4NLP) 2021. Talk to us at ACL if you are interested!

@inproceedings{gao2021scaling,
     title={Scaling Deep Contrastive Learning Batch Size under Memory Limited Setup},
     author={Luyu Gao, Yunyi Zhang, Jiawei Han, Jamie Callan},
     booktitle ={Proceedings of the 6th Workshop on Representation Learning for NLP},
     year={2021},
}

Understanding the Code

One of the goals of this repo is to help clarify our gradient cache technique. The initial commit of the repo is a snapshot of DPR. You can run diff against it to see all changes made in GC-DPR. In general, we have modified train and inference codes. The model code is kept unchanged, and therefore users should be able to use trained checkpoints interchangeably between DPR and GC-DPR.

Additional Features

Several changes were made to speed and/or facilitate easy reproduction of the original result. You may get some speed up using GC-DPR even when not using gradient.

  • Automatic Mix Precision (AMP): GC-DPR replaces apex AMP with Pytorch native AMP. It is supported for both training and inference.
  • Multi-process Data Loading: batches are loaded using Pytorch's DataLoader during training and inference in loader processes, which largely avoids GPU idle waiting for text tokenization and batch creation.
  • Gradient Compensation: Pytorch's Distributed Data Parallel (DDP) average reduces model parameter gradients across GPUs while dense retriever training all-gather representations to compute loss. This means the actual gradient will be divided by the number of GPUs. For reproducibility, we scale/unscale gradient according to the number of available devices to replicate this effect. Another option is to use _register_comm_hook on the DDP object. Scaling loss is opted as it is easier to understand.

The follwoing instructions are adapted from the orignial DPR's.

Installation

You can clone this repo and install with pip.

git clone https://github.com/luyug/GC-DPR.git
cd GC-DPR
pip install .

GC-DPR is tested on Python 3.8.5, PyTorch 1.6.0 and Huggingface transformers 3.0.2.

Resources & Data formats

You can follow the original DPR instructions:

First, you need to prepare data for either retriever or reader training. Each of the DPR components has its own input/output data formats. You can see format descriptions below. DPR provides NQ & Trivia preprocessed datasets (and model checkpoints) to be downloaded from the cloud using our data/download_data.py tool. One needs to specify the resource name to be downloaded. Run 'python data/download_data.py' to see all options.

python data/download_data.py \
   --resource {key from download_data.py's RESOURCES_MAP}  \
   [optional --output_dir {your location}]  

The resource name matching is prefix-based. So if you need to download all data resources, just use --resource data.

Retriever input data format

The data format of the Retriever training data is JSON.
It contains pools of 2 types of negative passages per question, as well as positive passages and some additional information.

[  
  {  
   "question": "....",  
   "answers": ["...", "...", "..."],  
   "positive_ctxs": [{  
      "title": "...",  
      "text": "...."  
   }],  
   "negative_ctxs": ["..."],  
   "hard_negative_ctxs": ["..."]  
  },  
  ...  
]  

Elements' structure for negative_ctxs & hard_negative_ctxs is exactly the same as for positive_ctxs. The preprocessed data available for downloading also contains some extra attributes which may be useful for model modifications (like bm25 scores per passage). Still, they are not currently in use by DPR.

You can download prepared NQ dataset used in the paper by using 'data.retriever.nq' key prefix. Only dev & train subsets are available in this format. We also provide question & answers only CSV data files for all train/dev/test splits. Those are used for the model evaluation since our NQ preprocessing step looses a part of original samples set. Use 'data.retriever.qas.*' resource keys to get respective sets for evaluation.

python data/download_data.py  
   --resource data.retriever  
   [optional --output_dir {your location}]  

Retriever training

In order to start training on one machine:

python train_dense_encoder.py  
   --encoder_model_type {hf_bert | pytext_bert | fairseq_roberta}  
   --pretrained_model_cfg {bert-base-uncased| roberta-base}  
   --train_file {train files glob expression}  
   --dev_file {dev files glob expression}  
   --output_dir {dir to save checkpoints}  
   --grad_cache
   --q_chunk_size {question sub-batch sizes}
   --ctx_chunk_size {context sub-batch sizes}
   --fp16  # if train with mixed precision

We introduce the following parameters to for gradient cached training,

  • --grad_cache activates gradient cached training
  • --q_chunk_size sub-batch size for updating the question encoder, default to 16
  • --ctx_chunk_size sub-batch size for updating context encoder, default to 8

The default settings work for 11GB RTX 2080 ti.

Retriever inference

You can follow the original DPR instructions:

Generating representation vectors for the static documents dataset is a highly parallelizable process which can take up to a few days if computed on a single GPU. You might want to use multiple available GPU servers by running the script on each of them independently and specifying their own shards.

python generate_dense_embeddings.py \
   --model_file {path to biencoder checkpoint} \
   --ctx_file {path to psgs_w100.tsv file} \
   --shard_id {shard_num, 0-based} --num_shards {total number of shards} \
   --out_file ${out files location + name PREFX}  \
   --fp16  # if use mixed precision inference

In addition, we provide the --fp16 flag which will run encoder inference with mixed precision using the autocast context manager. Mixed precision inference gives roughly 2x speed up. The generated embeddings will be converted to full precision before saving for index search on x86 CPU.

Retriever validation against the entire set of documents:

python dense_retriever.py \
   --model_file ${path to biencoder checkpoint} \
   --ctx_file  {path to all documents .tsv file} \
   --qa_file {path to test|dev .csv file} \
   --encoded_ctx_file "{encoded document files glob expression}" \
   --out_file {path to output json file with results} \
  --n-docs 200  

We provide in addition flags --encode_q_and_save, --q_encoding_path, --re_encode_q which allows separation of question encoding and index search. One could therefore opt to run question encoding on a GPU server and search on a CPU server.

The tool writes retrieved results for subsequent reader model training into specified out_file.
It is a json with the following format:

[  
    {  
        "question": "...",  
        "answers": ["...", "...", ... ],  
        "ctxs": [  
            {  
                "id": "...", # passage id from database tsv file  
                "title": "",  
                "text": "....",  
                "score": "...",  # retriever score  
                "has_answer": true|false  
     },  
]  

Results are sorted by their similarity score, from most relevant to least relevant.

Optional reader model input data pre-processing.

Since the reader model uses a specific combination of positive and negative passages for each question and also needs to know the answer span location in the bpe-tokenized form, it is recommended to preprocess and serialize the output from the retriever model before starting the reader training. This saves hours at train time.
If you don't run this preprocessing, the Reader training pipeline checks if the input file(s) extension is .pkl and if not, preprocesses and caches results automatically in the same folder as the original files.

python preprocess_reader_data.py \
   --retriever_results {path to a file with results from dense_retriever.py} \
   --gold_passages {path to gold passages info} \
   --do_lower_case \
   --pretrained_model_cfg {pretrained_cfg} \
   --encoder_model_type {hf_bert | pytext_bert | fairseq_roberta} \
   --out_file {path to for output files} \
   --is_train_set  

Reader model training

python train_reader.py \
   --encoder_model_type {hf_bert | pytext_bert | fairseq_roberta} \
   --pretrained_model_cfg {bert-base-uncased| roberta-base} \
   --train_file "{globe expression for train files from #5 or #6 above}" \
   --dev_file "{globe expression for train files}" \
   --output_dir {path to output dir}  

Notes:

  • if you use pytext_bert or fairseq_roberta, you need to download pre-trained weights and specify --pretrained_file parameter. Specify the dir location of the downloaded files for 'pretrained.fairseq.roberta-base' resource prefix for RoBERTa model or the file path for pytext BERT (resource name 'pretrained.pytext.bert-base.model').
  • Reader training pipeline does model validation every --eval_step batches
  • As the bi-encoder, it saves model checkpoints on every validation
  • Like the bi-encoder, there is no stop condition besides a specified amount of epochs to train.
  • Like the bi-encoder, there is no best checkpoint selection logic, so one needs to select that based on dev set validation performance which is logged in the train process output.
  • Our current code only calculates the Exact Match metric.

Reader model inference

In order to make an inference, run train_reader.py without specifying train_file. Make sure to specify model_file with the path to the checkpoint, passages_per_question_predict with number of passages per question (being used when saving the prediction file), and eval_top_docs with a list of top passages threshold values from which to choose question's answer span (to be printed as logs). The example command line is as follows.

python train_reader.py \
  --prediction_results_file {some dir}/results.json \
  --eval_top_docs 10 20 40 50 80 100 \
  --dev_file {path to data.retriever_results.nq.single.test file} \
  --model_file {path to the reader checkpoint} \
  --dev_batch_size 80 \
  --passages_per_question_predict 100 \
  --sequence_length 350  

Best hyperparameter settings

e2e example with the best settings for NQ dataset. Most of this section follows DPR's instructions. The retriever training session is adjusted for a single 11 GB GPU.

1. Download all retriever training and validation data:

python data/download_data.py --resource data.wikipedia_split.psgs_w100  
python data/download_data.py --resource data.retriever.nq  
python data/download_data.py --resource data.retriever.qas.nq  

2. Biencoder(Retriever) training in single set mode.

Train on a single 11GB RTX 2080 ti GPU,

python train_dense_encoder.py \
   --max_grad_norm 2.0 \
   --encoder_model_type hf_bert \
   --pretrained_model_cfg bert-base-uncased \
   --seed 12345 \
   --sequence_length 256 \
   --warmup_steps 1237 \
   --batch_size 128 \
   --do_lower_case \
   --train_file "{glob expression to train files downloaded as 'data.retriever.nq-train' resource}" \
   --dev_file "{glob expression to dev files downloaded as 'data.retriever.nq-dev' resource}" \
   --output_dir {your output dir} \
   --learning_rate 2e-05 \
   --num_train_epochs 40 \
   --dev_batch_size 16 \
   --val_av_rank_start_epoch 30 \
   --fp16 \
   --grad_cache \
   --global_loss_buf_sz 2097152 \
   --val_av_rank_max_qs 1000

This takes less than 2 days to complete the training for 40 epochs. The best checkpoint for bi-encoder is usually the last, but it should not be so different if you take any after epoch ~ 25. We use smaller value for --val_av_rank_max_qs to speed up validation.

3. Generate embeddings for Wikipedia.

Just use instructions for "Generating representations for large documents set". It takes about 40 minutes to produce 21 mln passages representation vectors on 50 2 GPU servers.

4. Evaluate retrieval accuracy and generate top passage results for each of the train/dev/test datasets.

python dense_retriever.py \
   --model_file {path to checkpoint file from step 1} \
   --ctx_file {path to psgs_w100.tsv file} \
   --qa_file {path to test/dev qas file} \
   --encoded_ctx_file "{glob expression for generated files from step 3}" \
   --out_file {path for output json files} \
   --n-docs 100 \
   --validation_workers 32 \
   --batch_size 64  

Adjust batch_size based on the available number of GPUs, 64 should work for 2 GPU server.

5. Reader training

Here's the DPR's original instruction for reader training. We trained reader model for large datasets using a single 8 GPU x 32 GB server.

python train_reader.py \
   --seed 42 \
   --learning_rate 1e-5 \
   --eval_step 2000 \
   --do_lower_case \
   --eval_top_docs 50 \
   --encoder_model_type hf_bert \
   --pretrained_model_cfg bert-base-uncased \
   --train_file "{glob expression for train output files from step 4}" \
   --dev_file {glob expression for dev output file from step 4} \
   --warmup_steps 0 \
   --sequence_length 350 \
   --batch_size 16 \
   --passages_per_question 24 \
   --num_train_epochs 100000 \
   --dev_batch_size 72 \
   --passages_per_question_predict 50 \
   --output_dir {your save dir path}  

We found that using the learning rate above works best with static schedule, so one needs to stop training manually based on evaluation performance dynamics.
Our best results were achieved on 16-18 training epochs or after ~60k model updates.

We provide all input and intermediate results for e2e pipeline for NQ dataset and most of the similar resources for Trivia.

Misc.

  • TREC validation requires regexp based matching. We support only retriever validation in the regexp mode. See --match parameter option.
  • WebQ validation requires entity normalization, which is not included as of now.

Reference

If you find GC-DPR helpful, please consider citing our paper:

@inproceedings{gao2021scaling,
      title={Scaling Deep Contrastive Learning Batch Size under Memory Limited Setup},
      author={Luyu Gao, Yunyi Zhang, Jiawei Han, Jamie Callan},
      booktitle ={Proceedings of the 6th Workshop on Representation Learning for NLP},
      year={2021},
}

Also consider citing the original DPR paper:

@misc{karpukhin2020dense,  
    title={Dense Passage Retrieval for Open-Domain Question Answering},  
    author={Vladimir Karpukhin and Barlas Oğuz and Sewon Min and Patrick Lewis and Ledell Wu and Sergey Edunov and Danqi Chen and Wen-tau Yih},  
    year={2020},  
    eprint={2004.04906},  
    archivePrefix={arXiv},  
    primaryClass={cs.CL}  
}  

License

GC-DPR inherits DPR's CC-BY-NC 4.0 licensed as of now.

Owner
Luyu Gao
NLP PhD [email protected], CMU
Luyu Gao
Voice assistant - Voice assistant with python

🌐 Python Voice Assistant 🌵 - User's greeting 🌵 - Writing tasks to todo-list ?

PythonToday 10 Dec 26, 2022
This repo provides function call to track multi-objects in videos

Custom Object Tracking Introduction This repo provides function call to track multi-objects in videos with a given trained object detection model and

Jeff Lo 51 Nov 22, 2022
Semi-Supervised Signed Clustering Graph Neural Network (and Implementation of Some Spectral Methods)

SSSNET SSSNET: Semi-Supervised Signed Network Clustering For details, please read our paper. Environment Setup Overview The project has been tested on

Yixuan He 9 Nov 24, 2022
Stochastic gradient descent with model building

Stochastic Model Building (SMB) This repository includes a new fast and robust stochastic optimization algorithm for training deep learning models. Th

S. Ilker Birbil 22 Jan 19, 2022
Code repository for "Stable View Synthesis".

Stable View Synthesis Code repository for "Stable View Synthesis". Setup Install the following Python packages in your Python environment - numpy (1.1

Intelligent Systems Lab Org 195 Dec 24, 2022
MXNet implementation for: Drop an Octave: Reducing Spatial Redundancy in Convolutional Neural Networks with Octave Convolution

Octave Convolution MXNet implementation for: Drop an Octave: Reducing Spatial Redundancy in Convolutional Neural Networks with Octave Convolution Imag

Meta Research 549 Dec 28, 2022
DPT: Deformable Patch-based Transformer for Visual Recognition (ACM MM2021)

DPT This repo is the official implementation of DPT: Deformable Patch-based Transformer for Visual Recognition (ACM MM2021). We provide code and model

CASIA-IVA-Lab 111 Dec 21, 2022
Code for "The Box Size Confidence Bias Harms Your Object Detector"

The Box Size Confidence Bias Harms Your Object Detector - Code Disclaimer: This repository is for research purposes only. It is designed to maintain r

Johannes G. 24 Dec 07, 2022
COVINS -- A Framework for Collaborative Visual-Inertial SLAM and Multi-Agent 3D Mapping

COVINS -- A Framework for Collaborative Visual-Inertial SLAM and Multi-Agent 3D Mapping Version 1.0 COVINS is an accurate, scalable, and versatile vis

ETHZ V4RL 183 Dec 27, 2022
Official Code for VideoLT: Large-scale Long-tailed Video Recognition (ICCV 2021)

Pytorch Code for VideoLT [Website][Paper] Updates [10/29/2021] Features uploaded to Google Drive, for access please send us an e-mail: zhangxing18 at

Skye 26 Sep 18, 2022
Kohei's 5th place solution for xview3 challenge

xview3-kohei-solution Usage This repository assumes that the given data set is stored in the following locations: $ ls data/input/xview3/*.csv data/in

Kohei Ozaki 2 Jan 17, 2022
Transformer based SAR image despeckling

Transformer based SAR image despeckling Using the code: The code is stable while using Python 3.6.13, CUDA =10.1 Clone this repository: git clone htt

27 Nov 13, 2022
Hierarchical Cross-modal Talking Face Generation with Dynamic Pixel-wise Loss (ATVGnet)

Hierarchical Cross-modal Talking Face Generation with Dynamic Pixel-wise Loss (ATVGnet) By Lele Chen , Ross K Maddox, Zhiyao Duan, Chenliang Xu. Unive

Lele Chen 218 Dec 27, 2022
[CVPR2021 Oral] End-to-End Video Instance Segmentation with Transformers

VisTR: End-to-End Video Instance Segmentation with Transformers This is the official implementation of the VisTR paper: Installation We provide instru

Yuqing Wang 687 Jan 07, 2023
Tzer: TVM Implementation of "Coverage-Guided Tensor Compiler Fuzzing with Joint IR-Pass Mutation (OOPSLA'22)“.

Artifact • Reproduce Bugs • Quick Start • Installation • Extend Tzer Coverage-Guided Tensor Compiler Fuzzing with Joint IR-Pass Mutation This is the s

12 Dec 29, 2022
Julia package for contraction of tensor networks, based on the sweep line algorithm outlined in the paper General tensor network decoding of 2D Pauli codes

Julia package for contraction of tensor networks, based on the sweep line algorithm outlined in the paper General tensor network decoding of 2D Pauli codes

Christopher T. Chubb 35 Dec 21, 2022
Spatial color quantization in Rust

rscolorq Rust port of Derrick Coetzee's scolorq, based on the 1998 paper "On spatial quantization of color images" by Jan Puzicha, Markus Held, Jens K

Collyn O'Kane 37 Dec 22, 2022
Make your AirPlay devices as TTS speakers

Apple AirPlayer Home Assistant integration component, make your AirPlay devices as TTS speakers. Before Use 2021.6.X or earlier Apple Airplayer compon

George Zhao 117 Dec 15, 2022
Dogs classification with Deep Metric Learning using some popular losses

Tsinghua Dogs classification with Deep Metric Learning 1. Introduction Tsinghua Dogs dataset Tsinghua Dogs is a fine-grained classification dataset fo

QuocThangNguyen 45 Nov 09, 2022
Implementation of "Unsupervised Domain Adaptive 3D Detection with Multi-Level Consistency"

Unsupervised Domain Adaptive 3D Detection with Multi-Level Consistency (ICCV2021) Paper Link: https://arxiv.org/abs/2107.11355 This implementation bui

32 Nov 17, 2022