salesforce/awd-lstm-lm
A PyTorch toolkit for training LSTM and QRNN-based neural language models on word and character levels.
Not currently ranked — collecting fresh signals.
star history
This repository implements word-level and character-level language models using LSTM and Quasi-Recurrent Neural Network (QRNN) architectures in PyTorch. It supports training on standard benchmarks including Penn Treebank, WikiText-2, WikiText-103, and the Hutter Prize dataset (enwik8). The toolkit includes techniques for regularization, optimization, finetuning, and optional continuous cache pointer mechanisms to improve model performance.
Frequently asked
- What is salesforce/awd-lstm-lm?
- A PyTorch toolkit for training LSTM and QRNN-based neural language models on word and character levels.
- Is awd-lstm-lm open source?
- Yes — salesforce/awd-lstm-lm is open source, released under the BSD-3-Clause license.
- What language is awd-lstm-lm written in?
- salesforce/awd-lstm-lm is primarily written in Python.
- How popular is awd-lstm-lm?
- salesforce/awd-lstm-lm has 2k stars on GitHub.
- Where can I find awd-lstm-lm?
- salesforce/awd-lstm-lm is on GitHub at https://github.com/salesforce/awd-lstm-lm.