Search Results for author:
Found 17 papers, 3 papers with code
DiDi’s Machine Translation System for WMT2020
This paper describes the DiDi AI Labs’ submission to the WMT2020 news translation shared task.
MuLTI: Efficient Video-and-Language Understanding with MultiWay-Sampler and Multiple Choice Modeling
Then, we introduce an attention-based adapter to the encoders, which finetunes the shallow features to improve the model's performance with low GPU memory consumption.
Ranked #1 on
TGIF-Transition
on TGIF-QA
Consecutive Question Generation via Dynamic Multitask Learning
In this paper, we propose the task of consecutive question generation (CQG), which generates a set of logically related question-answer pairs to understand a whole passage, with a comprehensive consideration of the aspects including accuracy, coverage, and informativeness.
Why Neural Machine Translation Prefers Empty Outputs
Using different EoS types in target sentences of different lengths exposes and eliminates this implicit smoothing.
DiDi's Machine Translation System for WMT2020
This paper describes DiDi AI Labs' submission to the WMT2020 news translation shared task.
MEEP: An Open-Source Platform for Human-Human Dialog Collection and End-to-End Agent Training
We create a new task-oriented dialog platform (MEEP) where agents are given considerable freedom in terms of utterances and API calls, but are constrained to work within a push-button environment.
One-shot Text Field Labeling using Attention and Belief Propagation for Structure Information Extraction
Existing learning based methods for text labeling task usually require a large amount of labeled examples to train a specific model for each type of document.
FINDINGS OF THE IWSLT 2020 EVALUATION CAMPAIGN
The evaluation campaign of the International Conference on Spoken Language Translation (IWSLT 2020) featured this year six challenge tracks: (i) Simultaneous speech translation, (ii) Video speech translation, (iii) Offline speech translation, (iv) Conversational speech translation, (v) Open domain translation, and (vi) Non-native speech translation.
Fast Locality Sensitive Hashing for Beam Search on GPU
We present a GPU-based Locality Sensitive Hashing (LSH) algorithm to speed up beam search for sequence models.
A Sequential Embedding Approach for Item Recommendation with Heterogeneous Attributes
Our ablation experiments demonstrate the effectiveness of the two components to address heterogeneous attribute challenges including variable lengths and attribute sparseness.
Speeding Up Neural Machine Translation Decoding by Shrinking Run-time Vocabulary
Compared with Locality Sensitive Hashing (LSH), decoding with word alignments is GPU-friendly, orthogonal to existing speedup methods and more robust across language pairs.
Temporal Learning and Sequence Modeling for a Job Recommender System
We present our solution to the job recommendation task for RecSys Challenge 2016.
