hyperparamters to reproduce paper result

[xianxl]

Could you share the full command (and default hyperparameters) you used? For example, I found --word_mask_keep_rand is "0.8,0.1,0.1" in MASS but "0,0,1" in MASS-fairseq.

[AndyELiu]

I also had a hard time to reproduce the unsupervised MT result. I run exactly as suggested in ReadME on en-fr. At epoch 4, my "valid_fr-en_mt_bleu" is only a little above 1, but you had "valid_fr-en_mt_bleu -> 10.55". I run it a few times.

[StillKeepTry]

@xianxl Our implementation based on fairseq contains some different methods from our paper. There will have some slight different experiment settings in fairseq.

[xianxl]

@StillKeepTry thanks for your reply. So what do you recommend to set word_mask_keep_rand in MASS-fairseq implementation? The default is "0,0,1" (which means no mask?) and this arg is not set in the training command you shared.

[caoyuan]

I followed the exactly the same setting as what the git page shows, running on a machine with 8 V100 gpus as the paper describes:

The git page claims that after 4 epochs, even without back translation the unsupervised BLEU should be close to the following numbers:

epoch -> 4
valid_fr-en_mt_bleu -> 10.55
valid_en-fr_mt_bleu -> 7.81
test_fr-en_mt_bleu -> 11.72
test_en-fr_mt_bleu -> 8.80

However this is not what I got, my numbers are much worse at epoch 4:

Could you please let us know if any param is wrong, or there are any hidden recipe that we're not aware of to reproduce the results?

On the other hand, I also loaded your pre-trained en-fr model, and the results are much better. So alternatively, could you share the settings you used to train the pre-trained model?

[caoyuan]

After some investigation, it seems that the suggested epoch size (200000) is really small and not the one used to produce the paper results. Could you confirm on this hypothesis?

[Bachstelze]

Can we conclude that the results are not reproducible?

[tan-xu]

Sorry for late response. We will provide all the scripts, hyper parameters, training logs to help you reproduce the results later today. 在 2019年11月3日，上午1:14，Bachstelze <notifications@github.com<mailto:notifications@github.com>> 写道： Can we conclude that the results are not reproducible? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fmicrosoft%2FMASS%2Fissues%2F32%3Femail_source%3Dnotifications%26email_token%3DALQV7L5L34BIG5JMDBQUBV3QRWYPRA5CNFSM4IBMVTK2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEC5AU2Q%23issuecomment-549063274&data=02%7C01%7CXu.Tan%40microsoft.com%7Ccc48921bb70946d1eb7a08d75fb8212f%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637083116757422671&sdata=lUCfdvSHzyicFpoZdumOLKIEA5jfdc3BQt9zW1VlJYU%3D&reserved=0>, or unsubscribe<https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FALQV7L34VC2MPBEKBUKGEH3QRWYPRANCNFSM4IBMVTKQ&data=02%7C01%7CXu.Tan%40microsoft.com%7Ccc48921bb70946d1eb7a08d75fb8212f%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637083116757422671&sdata=HxzuSmJz0aEESkLwK7DQDpZQzcF2C04odSU9DxH6y7w%3D&reserved=0>.

[StillKeepTry]

First, we used 50M monolingual data for each language during the training, this may be one reason (pre-training usually needs more data). Second, epoch_size=200000 in this code means training 200,000 sentences for one epoch. In other words, nearly 62 epochs (size=200,000) is equal to train on 50M data for once.

To explain this, I have uploaded some logs to this link from my previous experiments when epoch_size = 200000. It can obtain 8.24/5.45 (at 10 epochs), 11.95/8.21 (at 50 epochs), 13.38/9.34 (at 100 epochs), 14.26/10.06 (at 146 epochs). This is just the result of 146 epochs. While we take over 500 epochs in our experiments for pre-training.

And in the latest code, you can try this hyperparameter for pre-training which result in better performance:

DATA_PATH=/data/processed/de-en

export NGPU=8; CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 python -m torch.distributed.launch --nproc_per_node=$NGPU train.py \
        --exp_name unsupMT_ende                              \
        --dump_path ./models/en-de/                          \
        --exp_id test                                        \
        --data_path $DATA_PATH                               \
        --lgs 'en-de'                                        \
        --mass_steps 'en,de'                                 \
        --encoder_only false                                 \
        --emb_dim 1024                                       \
        --n_layers 6                                         \
        --n_heads 8                                          \
        --dropout 0.1                                        \
        --attention_dropout 0.1                              \
        --gelu_activation true                               \
        --tokens_per_batch 3000                              \
        --batch_size 32                                      \
        --bptt 256                                           \
        --optimizer adam_inverse_sqrt,beta1=0.9,beta2=0.98,lr=0.0001 \
        --epoch_size 200000                                  \
        --max_epoch 50                                       \
        --eval_bleu true                                     \
        --word_mass 0.5                                      \
        --min_len 4                                          \
        --save_periodic 10                                   \
        --lambda_span "8"                                    \
        --word_mask_keep_rand '0.8,0.05,0.15'

[k888]

@StillKeepTry could you also please provide the log for bt steps for the en_de model you pretrained?

Also log for pretraining and bt for en_fr would be highly appreciated as well!

[LibertFan]

@k888 have you find the hyperparameters for en_de BT steps?

[Frankszc]

First, we used 50M monolingual data for each language during the training, this may be one reason (pre-training usually needs more data). Second, epoch_size=200000 in this code means training 200,000 sentences for one epoch. In other words, nearly 62 epochs (size=200,000) is equal to train on 50M data for once.

To explain this, I have uploaded some logs to this link from my previous experiments when epoch_size = 200000. It can obtain 8.24/5.45 (at 10 epochs), 11.95/8.21 (at 50 epochs), 13.38/9.34 (at 100 epochs), 14.26/10.06 (at 146 epochs). This is just the result of 146 epochs. While we take over 500 epochs in our experiments for pre-training.

And in the latest code, you can try this hyperparameter for pre-training which result in better performance:

DATA_PATH=/data/processed/de-en

export NGPU=8; CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 python -m torch.distributed.launch --nproc_per_node=$NGPU train.py \
        --exp_name unsupMT_ende                              \
        --dump_path ./models/en-de/                          \
        --exp_id test                                        \
        --data_path $DATA_PATH                               \
        --lgs 'en-de'                                        \
        --mass_steps 'en,de'                                 \
        --encoder_only false                                 \
        --emb_dim 1024                                       \
        --n_layers 6                                         \
        --n_heads 8                                          \
        --dropout 0.1                                        \
        --attention_dropout 0.1                              \
        --gelu_activation true                               \
        --tokens_per_batch 3000                              \
        --batch_size 32                                      \
        --bptt 256                                           \
        --optimizer adam_inverse_sqrt,beta1=0.9,beta2=0.98,lr=0.0001 \
        --epoch_size 200000                                  \
        --max_epoch 50                                       \
        --eval_bleu true                                     \
        --word_mass 0.5                                      \
        --min_len 4                                          \
        --save_periodic 10                                   \
        --lambda_span "8"                                    \
        --word_mask_keep_rand '0.8,0.05,0.15'

Hello, thanks for your great work. And you said that you used 50M monolingual data(50,000,000 sentences) for each language during the training, the epoch_size is 200,000, so why the number of epoch training on 50M data for once is 62? why not 250? @StillKeepTry