OpenNMT Forum

Support

Two short questions on training (2)

OpenNMT pc specs (6)

Opennmt-py demo training failed from time to time (1)

Bad results from DE-EN Pretrained models (12)

How to load pretrained word2vec embedding files with opennmt-tf? (2)

Advice on new GPU (11)

Transformer model error (2)

BLEU was OK at the beginning, but decreased to 0 (6)

DataLossError : Checksum does not match (10)

How to specify input data files for ParallelInputter? (3)

Restoring of source formatting (14)

Cuda OOM for Validation (9)

Early stopping parameters for default transformer config? (3)

Large perplexity (7)

Translating huge number of sentences (3)

Question about using the Pretrained embedding (4)

"Import pyonmttok" failed & "pip install pyonmttok" faileld - Windows 10 (2)

Receiving a "ValueError: best_eval_result cannot be empty or no loss is found in it." while training the transformer model (3)

OpenNMT-tf Fine tuning base model gives worse and decreases BLEU scores (9)

Handle numbers, urls, dates (3)

TransformerBig model - GTX 1080 Ti (11G) - ResourceExhaustedError (9)

Use GloVe with concatenated word features (OpenNMT-py) (1)

Tokenization OpenNMT-tf (4)

Using multiple gpus in training.py (5)

Translation server in OpenNMT-tf (3)

Perplexity in opennmt-tf (2)

Alignment not coming properly (2)

OpenNMT-tf Distributed Training - Processes don't end after training completes (5)

OpenNMT-tf serving with latest prediction_service_pb2_grpc for translation (3)

Hi , I have 130k train src tgt pairs, but the log info of training process shows that there are only 22k train pairs after we preprocess the data (5)