Tensor-Transformer(1core)+PN (WT103)
University of California (UC) BerkeleyLanguage modelingOpen weights
Developed by University of California (UC) Berkeley in 2020, Tensor-Transformer(1core)+PN (WT103) is a language modeling model with 85300000.0 parameters with openly available weights.
About Tensor-Transformer(1core)+PN (WT103)
The standard normalization method for neural network (NN) models used in Natural Language Processing (NLP) is layer normalization (LN). This is different than batch normalization (BN), which is widely-adopted in Computer Vision. The preferred use of
Details
- Provider
- University of California (UC) Berkeley
- Task
- Language modeling
- Parameters
- 85300000.0
- Released
- 2020-03-17
- Open weights
- Yes