OpenNMT Forum

Feature Requests

About the Feature Requests category (1)
Fine-Tuning NMT | how can i test joint learning with Discriminative Mixing? (1)
Fine Tuning | what would be the best way to test L2 regularization towards Out-of-domain model? (1)
Fine Tuning | does NMT support freezing weights in some layers? (1)
Multi-Source Translation ( 2 ) (22)
Multi function REST API (6)
Terminology handling (12)
Transformer model for OpenNMT Lua version (2)
Feature Request List (15)
Implementing SR-NMT in OpenNMT-{py,tf} (2)
Support shard training (3)
Passing arguments to tf.config? (3)
OpenNMT-py Ensemble Decoding (1)
Add user data to checkpoint? (5)
"-add_tokenized_output "option in rest_translation_server.lua (9)
Only save models after X epochs (3)
Combining Fuzzy Matches with NMT (7)
Implement Copy Mechanism (6)
Character-Aware Models (3)
Automating hyperparameter optimization (2)
Control translation length (8)
The Ultimate wish-list for OpenNMT-Lua (2)
English to Tamil and Tamil to English Translation (5)
Batch Normalization / L2 regularization (3)
Improving performance by replacing Softmax (9)
Memory optimization (1)
[OpenNMT-tf] Multitask learning support (2)
Support building vocabs from multiple files (1)
Ensemble decoding (17)
Supporting Tokenize/Detokenize automatically by Translation Server (18)