All Projects → liuaiting → Hip Hop Seq2seq

liuaiting / Hip Hop Seq2seq

DeeCamp 2018,AI有嘻哈 自动写歌词

Programming Languages

python
139335 projects - #7 most used programming language

Labels

Projects that are alternatives of or similar to Hip Hop Seq2seq

Neural sp
End-to-end ASR/LM implementation with PyTorch
Stars: ✭ 408 (+827.27%)
Mutual labels:  seq2seq
Seq2seq Couplet
Play couplet with seq2seq model. 用深度学习对对联。
Stars: ✭ 5,149 (+11602.27%)
Mutual labels:  seq2seq
Seq2seq Signal Prediction
Signal forecasting with a Sequence-to-Sequence (seq2seq) Recurrent Neural Network (RNN) model in TensorFlow - Guillaume Chevalier
Stars: ✭ 890 (+1922.73%)
Mutual labels:  seq2seq
Joeynmt
Minimalist NMT for educational purposes
Stars: ✭ 420 (+854.55%)
Mutual labels:  seq2seq
Text Summarization Tensorflow
Tensorflow seq2seq Implementation of Text Summarization.
Stars: ✭ 527 (+1097.73%)
Mutual labels:  seq2seq
Seq2seq Pytorch
Sequence to Sequence Models with PyTorch
Stars: ✭ 678 (+1440.91%)
Mutual labels:  seq2seq
Nmtpytorch
Sequence-to-Sequence Framework in PyTorch
Stars: ✭ 392 (+790.91%)
Mutual labels:  seq2seq
Augmented seq2seq
enhance seq2seq model for open ended dialog generation
Stars: ✭ 29 (-34.09%)
Mutual labels:  seq2seq
Seq2seq
Minimal Seq2Seq model with Attention for Neural Machine Translation in PyTorch
Stars: ✭ 552 (+1154.55%)
Mutual labels:  seq2seq
Neuralconvo
Neural conversational model in Torch
Stars: ✭ 773 (+1656.82%)
Mutual labels:  seq2seq
Practical Pytorch
Go to https://github.com/pytorch/tutorials - this repo is deprecated and no longer maintained
Stars: ✭ 4,329 (+9738.64%)
Mutual labels:  seq2seq
Seq2seq.pytorch
Sequence-to-Sequence learning using PyTorch
Stars: ✭ 514 (+1068.18%)
Mutual labels:  seq2seq
Deeplearningmugenknock
でぃーぷらーにんぐを無限にやってディープラーニングでDeepLearningするための実装CheatSheet
Stars: ✭ 684 (+1454.55%)
Mutual labels:  seq2seq
Tf seq2seq chatbot
[unmaintained]
Stars: ✭ 420 (+854.55%)
Mutual labels:  seq2seq
Attention Ocr
A Tensorflow model for text recognition (CNN + seq2seq with visual attention) available as a Python package and compatible with Google Cloud ML Engine.
Stars: ✭ 844 (+1818.18%)
Mutual labels:  seq2seq
Nlp Tutorials
Simple implementations of NLP models. Tutorials are written in Chinese on my website https://mofanpy.com
Stars: ✭ 394 (+795.45%)
Mutual labels:  seq2seq
Practical seq2seq
A simple, minimal wrapper for tensorflow's seq2seq module, for experimenting with datasets rapidly
Stars: ✭ 563 (+1179.55%)
Mutual labels:  seq2seq
Sockeye
Sequence-to-sequence framework with a focus on Neural Machine Translation based on Apache MXNet
Stars: ✭ 990 (+2150%)
Mutual labels:  seq2seq
Keras Question And Answering Web Api
Question answering system developed using seq2seq and memory network model in Keras
Stars: ✭ 21 (-52.27%)
Mutual labels:  seq2seq
Cluener2020
CLUENER2020 中文细粒度命名实体识别 Fine Grained Named Entity Recognition
Stars: ✭ 689 (+1465.91%)
Mutual labels:  seq2seq

Hip-Hop-Seq2seq

Projects of DeeCamp2018, hip-hop lyrics generation using Seq2Seq.

Setup

  • python 3.6
  • tensorflow==1.4
  • jieba

Model

基于nmt模型,在inference部分进行了一些修改。

decoder

设计了两种hook的生成方式:

  1. hook中每一句的第一个词一样
和你我不说累一切都无所谓
和你我不后悔和你我不流泪
和你我不说累一切都无所谓
和你我不后悔和你我不流泪
  1. 生成的句子与原始句子押韵(inference时将不押韵的词的概率置0)
你感受冰冷的风
也曾经感受过梦
像一场不敢醒来不痛的梦
你试过苦涩的痛
没想过郁郁而终

Training and Evaluate

在项目主目录下,运行下面的命令训练第一种hook生成模型:

python3 -m seq2seq.train\
        --source_train_data="seq2seq/data/v2/train.src"\
        --target_train_data="seq2seq/data/v2/train.tgt"\
        --source_dev_data="seq2seq/data/v2/dev.src"\
        --target_dev_data="seq2seq/data/v2/dev.tgt"\
        --src_vocab_file="seq2seq/data/v2/vocab.tgt"\
        --tgt_vocab_file="seq2seq/data/v2/vocab.tgt"\
        --src_vocab_size=43836\
        --tgt_vocab_size=43836\
        --share_vocab=True\
        --out_dir="seq2seq/model1"\
        --decoder_rule="samefirst"\
        --num_epochs=60

运行下面的命令训练第二种hook生成模型:

python3 -m seq2seq.train\
        --source_train_data="seq2seq/data/v3/train.src"\
        --target_train_data="seq2seq/data/v3/train.tgt"\
        --source_dev_data="seq2seq/data/v3/dev.src"\
        --target_dev_data="seq2seq/data/v3/dev.tgt"\
        --src_vocab_file="seq2seq/data/v3/vocab.tgt"\
        --tgt_vocab_file="seq2seq/data/v3/vocab.tgt"\
        --src_vocab_size=23442\
        --tgt_vocab_size=23442\
        --share_vocab=True\
        --out_dir="seq2seq/model3"\
        --decoder_rule="rhyme"\
        --num_epochs=60

Inference

训练好模型后,运行下面的命令进行第一种hook的inference:

python3 -m seq2seq.test_inference\
        --src_vocab_file="seq2seq/data/v2/vocab.tgt"\
        --tgt_vocab_file="seq2seq/data/v2/vocab.tgt"\
        --src_vocab_size=43836\
        --tgt_vocab_size=43836\
        --share_vocab=True\
        --out_dir="seq2seq/model1"\
        --inference_input_file="seq2seq/data/plana/1.0.txt"\
        --inference_output_file="seq2seq/model1/output"\
        --decoder_rule="samefirst"

第二种hook的inference:

python3 -m seq2seq.test_inference\
        --src_vocab_file="seq2seq/data/v3/vocab.tgt"\
        --tgt_vocab_file="seq2seq/data/v3/vocab.tgt"\
        --src_vocab_size=23442\
        --tgt_vocab_size=23442\
        --share_vocab=True\
        --out_dir="seq2seq/model3"\
        --inference_input_file="seq2seq/data/plana/1.0.txt"\
        --inference_output_file="seq2seq/model3/output"\
        --rhyme_table_file="seq2seq/data/v3/table_23442.npy"\
        --decoder_rule="rhyme"

Run Server

在项目主目录下运行下面的命令开启服务。

bash server.sh

References

News

Homie快来听,AI的嘻哈也很酷 | DeeCamp Show

填词作曲斗图搞艺术,AI统统都可以

2分钟写出一首嘻哈歌曲:95后AI创新团队叫板“中国新说唱”

Demo

视频展示:YouTube

Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].