LuxGPT-2 based GER
GPT-2 model for Text Generation in luxembourgish language, trained on 711 MB of text data, consisting of RTL.lu news articles, comments, parlament speeches, the luxembourgish Wikipedia, Newscrawl, Webcrawl and subtitles. Created via transfer learning with an English base model, feature space mapping from LB on Base feature space and gradual layer freezing. The training took place on a 32 GB Nvidia Tesla V100
- with One Cycle policy for the learning rate
- with the help of fastai's LR finder
- for 49.2 hours
- for 18 epochs and 8 cycles
- using the fastai library
Usage
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("laurabernardy/LuxGPT2-basedEN")
model = AutoModelForCausalLM.from_pretrained("laurabernardy/LuxGPT2-basedEN")
Limitations and Biases
See the GPT2 model card for considerations on limitations and bias. See the GPT2 documentation for details on GPT2.
- Downloads last month
- 7
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Evaluation results
- accuracy on Luxembourgish Test Datasetself-reported0.35
- perplexity on Luxembourgish Test Datasetself-reported45.08