Nematus Support

Contact owners and managers

1–30 of 116

Welcome to the Nematus support mailing list, a place for support requests and user discussions for the Nematus NMT toolkit ( https://github.com/rsennrich/nematus )

0 selected

Jul 1

Unable to translate with factored model

Hello, I am reproducing Netamus factored translation pipeline (FR->EN) following these

unread,

Unable to translate with factored model

Hello, I am reproducing Netamus factored translation pipeline (FR->EN) following these

Jul 1

Feb 16

Language Technology Market Study: participate in the supplier survey and share your insights!

Dear Language Technology Supplier, The 2024 Language Technology Supplier Survey is still open!

unread,

Language Technology Market Study: participate in the supplier survey and share your insights!

Dear Language Technology Supplier, The 2024 Language Technology Supplier Survey is still open!

Feb 16

Cameron Jen, Rico Sennrich2

7/14/23

Nematus installation question

Hello Cameron, tornado 5.1.1 seems to be the last tornado version that supports Python 2.7. Note that

unread,

Nematus installation question

Hello Cameron, tornado 5.1.1 seems to be the last tornado version that supports Python 2.7. Note that

7/14/23

Jana Vozárová, Rico Sennrich2

4/7/21

Question regarding accessing output

Hello Jana, you can use the translate.py script to generate translations of any input file. It writes

unread,

Question regarding accessing output

Hello Jana, you can use the translate.py script to generate translations of any input file. It writes

4/7/21

sas cam, … WILLIAMS Philip5

1/7/21

training steps in Nematus

Thank you Philip for the details. On Wed, 6 Jan 2021 at 19:19, WILLIAMS Philip <pwil...@exseed.ed

unread,

training steps in Nematus

Thank you Philip for the details. On Wed, 6 Jan 2021 at 19:19, WILLIAMS Philip <pwil...@exseed.ed

1/7/21

Haruo Con, Rico Sennrich2

10/12/20

Limit translation length and block ngram repeat

Hello Haru, with older versions of Nematus, you can control the translation max length in the model

unread,

Limit translation length and block ngram repeat

Hello Haru, with older versions of Nematus, you can control the translation max length in the model

10/12/20

10/8/20

Using pre-trained word embedding

Hi Rico, As you mentioned in another thread, to train the nematus for 0/1 mini batch and replaced the

unread,

Using pre-trained word embedding

Hi Rico, As you mentioned in another thread, to train the nematus for 0/1 mini batch and replaced the

10/8/20

10/8/20

Validation script not working while training.

Hi Rico, On running the following training script: THEANO_FLAGS=mode=FAST_RUN,floatX=float32,device=$

unread,

Validation script not working while training.

Hi Rico, On running the following training script: THEANO_FLAGS=mode=FAST_RUN,floatX=float32,device=$

10/8/20

sam, … Haruo Con5

10/8/20

how does nematus support pre-trained word embeddings?

Hi Sam, Were you able to include pre-trained embedding in nematus? Thanks, Haru On Friday, April 28,

unread,

how does nematus support pre-trained word embeddings?

Hi Sam, Were you able to include pre-trained embedding in nematus? Thanks, Haru On Friday, April 28,

10/8/20

Shantanu Nath, Rico Sennrich2

6/5/20

how to handle early stop

Hello Shantanu, early stopping means that your last N (by default: 10) validation failed to improve

unread,

how to handle early stop

Hello Shantanu, early stopping means that your last N (by default: 10) validation failed to improve

6/5/20

Shantanu Nath, Rico Sennrich2

5/19/20

I have a question about Building Dictionary

Hello Shantanu, sentencepiece applies its own simple tokenization (basically splitting on whitespace

unread,

I have a question about Building Dictionary

Hello Shantanu, sentencepiece applies its own simple tokenization (basically splitting on whitespace

5/19/20

Shantanu Nath, Rico Sennrich2

5/7/20

UnicodeDecodeError: 'charmap' codec can't decode byte 0x9d in position 187: character maps to <undefined>

Hello Shantanu, From the Traceback, it looks like your environment (powershell) is using the encoding

unread,

UnicodeDecodeError: 'charmap' codec can't decode byte 0x9d in position 187: character maps to <undefined>

Hello Shantanu, From the Traceback, it looks like your environment (powershell) is using the encoding

5/7/20

Shantanu Nath, … Rico Sennrich4

5/3/20

Not showing the output

Thank you for the clarification. On Sun, May 3, 2020 at 1:54 PM Rico Sennrich <rico.sennrich@gmx.

unread,

Not showing the output

Thank you for the clarification. On Sun, May 3, 2020 at 1:54 PM Rico Sennrich <rico.sennrich@gmx.

5/3/20

orto...@gmail.com, privatl...@gmail.com2

4/29/20

TFLiteConverter

This would be quite interesting to me, too. On Saturday, August 3, 2019 at 2:26:57 AM UTC+2, orto...@

unread,

TFLiteConverter

This would be quite interesting to me, too. On Saturday, August 3, 2019 at 2:26:57 AM UTC+2, orto...@

4/29/20

Shantanu Nath, R S2

4/29/20

Errors may have originated from an input operation

Hello Shantanu, could you resolve your problem? In the other thread, you mention that you

unread,

Errors may have originated from an input operation

Hello Shantanu, could you resolve your problem? In the other thread, you mention that you

4/29/20

annette...@rcn.com

4/15/20

unread,

4/15/20

Shantanu Nath, Philip Williams2

2/11/20

Can I use own dictionary without building dikctionary?

Hi, build_dictionary.py is used to create the JSON files that specify the model's vocabulary.

unread,

Can I use own dictionary without building dikctionary?

Hi, build_dictionary.py is used to create the JSON files that specify the model's vocabulary.

2/11/20

Mohammad Mumin, Philip Williams3

1/13/20

layer size for transformer architechture

Thank you Sir for the clarification. Best wishes, Mumin. On Monday, January 13, 2020 at 9:35:40 PM

unread,

layer size for transformer architechture

Thank you Sir for the clarification. Best wishes, Mumin. On Monday, January 13, 2020 at 9:35:40 PM

1/13/20

Zhong Zhou, Rico Sennrich2

12/18/19

Hello Zhong, thanks for the bug report. This should be fixed: https://github.com/EdinburghNLP/nematus

unread,

Hello Zhong, thanks for the bug report. This should be fixed: https://github.com/EdinburghNLP/nematus

12/18/19

Andar, Rico Sennrich8

11/13/19

Scripts for training and evaluation of low-resource NMT

Hi Andar, the BiDeep architecture is an RNN architecture, and the result of enabling two kinds of

unread,

Scripts for training and evaluation of low-resource NMT

Hi Andar, the BiDeep architecture is an RNN architecture, and the result of enabling two kinds of

11/13/19

Mohammad Mumin, Rico Sennrich4

10/28/19

Error at ensemble decoding

Thank you Sir for your reply and suggestions. Best regards, On Monday, October 28, 2019 at 2:31:07 PM

unread,

Error at ensemble decoding

Thank you Sir for your reply and suggestions. Best regards, On Monday, October 28, 2019 at 2:31:07 PM

10/28/19

9/19/19

Re: I have a question about Nematus

Hi Safae, What is the content of en-de/input-test.en? You're saying that you're running

unread,

Re: I have a question about Nematus

Hi Safae, What is the content of en-de/input-test.en? You're saying that you're running

9/19/19

Mohammad Mumin, Rico Sennrich3

8/18/19

OSError: [Errno 12] Cannot allocate memory

Thank you very much Sir for your thoughtful response. Perhaps, the first case is happened for me.

unread,

OSError: [Errno 12] Cannot allocate memory

Thank you very much Sir for your thoughtful response. Perhaps, the first case is happened for me.

8/18/19

Mohammad Mumin, Philip Williams7

7/31/19

UnicodeDecodeError in build_dictionary.py

YES Sir, you have fixed the bug. Thanks a lot sir. Best wishes to Nemaus and its team. On Thursday,

unread,

UnicodeDecodeError in build_dictionary.py

YES Sir, you have fixed the bug. Thanks a lot sir. Best wishes to Nemaus and its team. On Thursday,

7/31/19

Mohammad Mumin, Rico Sennrich4

7/25/19

UnboundLocalError: local variable 'num_sents' referenced before assignment

Thank you sir for your explanation. On Wednesday, July 17, 2019 at 6:44:39 PM UTC+6, Mohammad Mumin

unread,

UnboundLocalError: local variable 'num_sents' referenced before assignment

Thank you sir for your explanation. On Wednesday, July 17, 2019 at 6:44:39 PM UTC+6, Mohammad Mumin

7/25/19

Mohammad Mumin, Philip Williams3

7/22/19

ValueError: too many values to unpack (expected 3)

Sir, Thank you very much for your support. I used fresh clone of the Nematus repository and my

unread,

ValueError: too many values to unpack (expected 3)

Sir, Thank you very much for your support. I used fresh clone of the Nematus repository and my

7/22/19

Mohammad Mumin, Rico Sennrich2

7/15/19

ResourceExhaustedError

Hello Mohammad, I suggest that you continue with the Tensorflow version, since this is better

unread,

ResourceExhaustedError

Hello Mohammad, I suggest that you continue with the Tensorflow version, since this is better

7/15/19

Mohammad Mumin, Rico Sennrich2

7/11/19

Another non-utf8 locales problem

Hello Mohammad, thanks for reporting this. I've pushed a fix that opens all files in UTF-8, which

unread,

Another non-utf8 locales problem

Hello Mohammad, thanks for reporting this. I've pushed a fix that opens all files in UTF-8, which

7/11/19

7/5/19

unread,

7/5/19

vemulapa...@gmail.com, Rico Sennrich5

5/24/19

Translated 0 sentences

The file does exist. What i tried to show is it is giving translated 0 sentences when i give a

unread,

Translated 0 sentences

The file does exist. What i tried to show is it is giving translated 0 sentences when i give a

5/24/19

Search

Clear search

Close search

Google apps

Main menu