Upload README.md

README.md

@@ -1,12 +1,63 @@
----
-language:
-- en
-- de
-- fr
-- es
-- tr
-tags:
-- nsp
-- next-sentence-prediction
-- gpt
----
+# mGPT-nsp
+
+mGPT-nsp is fine-tuned for Next Sentence Prediction task on the [wikipedia dataset](https://huggingface.co/datasets/wikipedia) using [multilingual GPT](https://huggingface.co/THUMT/mGPT) model. It was introduced in this [paper](https://arxiv.org/abs/2307.07331) and first released on this page.
+
+## Model description
+
+mGPT-nsp is a Transformer-based model which fine-tuned for Next Sentence Prediction task on 11000 English and 11000 German Wikipedia articles. Similar to GPT-2, It was pre-trained on the raw texts only, with no human labeling. We use the same tokenization and vocabulary as the [mT5 model](https://huggingface.co/google/mt5-base).
+
+## Intended uses
+
+- Apply Next Sentence Prediction tasks. (compare the results with BERT models since BERT natively supports this task)
+- See how to fine-tune a mGPT2 model using our [code](https://github.com/slds-lmu/stereotypes-multi/tree/main)
+- Check our [paper](https://arxiv.org/abs/2307.07331) to see its results
+
+## How to use
+
+You can use this model directly with a pipeline for next sentence prediction.  Here is how to use this model in PyTorch:
+
+### Necessary Initialization
+```python
+from transformers import MT5Tokenizer, GPT2Model
+import torch
+from huggingface_hub import hf_hub_download
+
+class ModelNSP(torch.nn.Module):
+    def __init__(self, pretrained_model="THUMT/mGPT"):
+        super(ModelNSP, self).__init__()
+        self.core_model = GPT2Model.from_pretrained(pretrained_model)
+        hidden_size = self.core_model.config.hidden_size
+        self.nsp_head = torch.nn.Sequential(torch.nn.Linear(hidden_size, 300), torch.nn.Linear(300, 300), torch.nn.Linear(300, 2))
+
+    def forward(self, input_ids, attention_mask=None):
+        core_model_outputs = self.core_model(input_ids, attention_mask=attention_mask)[0].mean(dim=1)
+        return self.nsp_head(core_model_outputs).softmax(dim=-1)
+
+weights = torch.load(hf_hub_download(repo_id="tolga-ozturk/mGPT-nsp", filename="model_weights.bin"))
+model = torch.nn.DataParallel(ModelNSP().eval())
+model.load_state_dict(weights)
+tokenizer = MT5Tokenizer.from_pretrained("tolga-ozturk/mGPT-nsp")
+```
+
+### Inference
+```python
+encoded_dict = tokenizer.batch_encode_plus(batch_text_or_text_pairs=[("In Italy, pizza is presented unsliced.", "The sky is blue."),
+                                                                     ("In Italy, pizza is presented unsliced.", "However, it is served sliced in Turkey.")], truncation="longest_first", padding=True, return_tensors="pt", return_attention_mask=True, max_length=256)
+outputs = model(encoded_dict.input_ids, attention_mask=encoded_dict.attention_mask)
+print(outputs)
+print(torch.argmax(outputs, dim=-1))
+```
+
+## BibTeX entry and citation info
+
+```bibtex
+@misc{title={How Different Is Stereotypical Bias Across Languages?}, 
+      author={Ibrahim Tolga Öztürk and Rostislav Nedelchev and Christian Heumann and Esteban Garces Arias and Marius Roger and Bernd Bischl and Matthias Aßenmacher},
+      year={2023},
+      eprint={2307.07331},
+      archivePrefix={arXiv},
+      primaryClass={cs.CL}
+}
+```
+
+The work is done with Ludwig-Maximilians-Universität Statistics group, don't forget to check out [their huggingface page](https://huggingface.co/misoda) for other interesting works!