Update README.md

README.md

@@ -1,15 +1,22 @@
 ---
 language: en
-tags: 
+tags:
 - deberta
 - fill-mask
-thumbnail: https://huggingface.co/front/thumbnails/microsoft.png
 license: mit
+pipeline_tag: text-generation
 ---
 
-## DeBERTa-fixed: Decoding-enhanced BERT with Disentangled Attention
+# DeBERTa-fixed: from paper "BERTs are Generative In-Context Learners"
 
-### Example code
+This is [**deberta-v2-xxlarge**](https://huggingface.co/microsoft/deberta-v2-xxlarge) updated to implement the `AutoModelForCausalLM` class, enabling it to generate text. This implementation is based on our paper "BERTs are Generative In-Context Learners".
+
+This repository also fixes three bugs in the original HF implementation of DeBERTa:
+1. We fixed the incorrect name of the output embedding weights in the checkpoint file;
+2. We fixed the implementation of the enhanced mask decoder (EMD), based on [the original GitHub repository](https://github.com/microsoft/DeBERTa);
+3. We clamp the positional embeddings so that they work with long sequence lengths.
+
+## Example code
 
 ```python
 from transformers import AutoTokenizer, AutoModelForCausalLM
@@ -37,79 +44,16 @@ prediction = tokenizer.decode(prediction).rstrip('\\')
 print(prediction)
 ```
 
-## Old README below:
-
-[DeBERTa](https://arxiv.org/abs/2006.03654) improves the BERT and RoBERTa models using disentangled attention and enhanced mask decoder. It outperforms BERT and RoBERTa on  majority of NLU tasks with 80GB training data. 
-
-Please check the [official repository](https://github.com/microsoft/DeBERTa) for more details and updates.
-
-This is the DeBERTa V2 xxlarge model with 48 layers, 1536 hidden size. The total parameters are 1.5B and it is trained with 160GB raw data.
-
-
-### Fine-tuning on NLU tasks
-
-We present the dev results on SQuAD 1.1/2.0 and several GLUE benchmark tasks.
 
-| Model                     | SQuAD 1.1 | SQuAD 2.0 | MNLI-m/mm   | SST-2 | QNLI | CoLA | RTE    | MRPC  | QQP   |STS-B |
-|---------------------------|-----------|-----------|-------------|-------|------|------|--------|-------|-------|------|
-|                           | F1/EM     | F1/EM     | Acc         | Acc   | Acc  | MCC  | Acc    |Acc/F1 |Acc/F1 |P/S   |
-| BERT-Large                | 90.9/84.1 | 81.8/79.0 | 86.6/-      | 93.2  | 92.3 | 60.6 | 70.4   | 88.0/-       | 91.3/- |90.0/- |
-| RoBERTa-Large             | 94.6/88.9 | 89.4/86.5 | 90.2/-      | 96.4  | 93.9 | 68.0 | 86.6   | 90.9/-       | 92.2/- |92.4/- |
-| XLNet-Large               | 95.1/89.7 | 90.6/87.9 | 90.8/-      | 97.0  | 94.9 | 69.0 | 85.9   | 90.8/-       | 92.3/- |92.5/- |
-| [DeBERTa-Large](https://huggingface.co/microsoft/deberta-large)<sup>1</sup> | 95.5/90.1 | 90.7/88.0 | 91.3/91.1| 96.5|95.3| 69.5| 91.0| 92.6/94.6| 92.3/- |92.8/92.5 |
-| [DeBERTa-XLarge](https://huggingface.co/microsoft/deberta-xlarge)<sup>1</sup> | -/-  | -/-  | 91.5/91.2| 97.0 | - | -    | 93.1   | 92.1/94.3    | -    |92.9/92.7|
-| [DeBERTa-V2-XLarge](https://huggingface.co/microsoft/deberta-v2-xlarge)<sup>1</sup>|95.8/90.8| 91.4/88.9|91.7/91.6| **97.5**| 95.8|71.1|**93.9**|92.0/94.2|92.3/89.8|92.9/92.9|
-|**[DeBERTa-V2-XXLarge](https://huggingface.co/microsoft/deberta-v2-xxlarge)<sup>1,2</sup>**|**96.1/91.4**|**92.2/89.7**|**91.7/91.9**|97.2|**96.0**|**72.0**| 93.5| **93.1/94.9**|**92.7/90.3** |**93.2/93.1** |
---------
-#### Notes.
- - <sup>1</sup> Following RoBERTa, for RTE, MRPC, STS-B, we fine-tune the tasks based on [DeBERTa-Large-MNLI](https://huggingface.co/microsoft/deberta-large-mnli), [DeBERTa-XLarge-MNLI](https://huggingface.co/microsoft/deberta-xlarge-mnli), [DeBERTa-V2-XLarge-MNLI](https://huggingface.co/microsoft/deberta-v2-xlarge-mnli), [DeBERTa-V2-XXLarge-MNLI](https://huggingface.co/microsoft/deberta-v2-xxlarge-mnli). The results of SST-2/QQP/QNLI/SQuADv2 will also be slightly improved when start from MNLI fine-tuned models, however, we only report the numbers fine-tuned from pretrained base models for those 4 tasks.
- - <sup>2</sup> To try the **XXLarge** model with **[HF transformers](https://huggingface.co/transformers/main_classes/trainer.html)**, we recommand using **deepspeed** as it's faster and saves memory.
- 
-Run with `Deepspeed`,
+## Citation
 
-```bash
-pip install datasets
-pip install deepspeed
-
-# Download the deepspeed config file
-wget https://huggingface.co/microsoft/deberta-v2-xxlarge/resolve/main/ds_config.json -O ds_config.json
+If you find DeBERTa useful for your work, please cite the following paper:
 
-export TASK_NAME=mnli
-output_dir="ds_results"
-num_gpus=8
-batch_size=8
-python -m torch.distributed.launch --nproc_per_node=${num_gpus} \\
-  run_glue.py \\
-  --model_name_or_path microsoft/deberta-v2-xxlarge \\
-  --task_name $TASK_NAME \\
-  --do_train \\
-  --do_eval \\
-  --max_seq_length 256 \\
-  --per_device_train_batch_size ${batch_size} \\
-  --learning_rate 3e-6 \\
-  --num_train_epochs 3 \\
-  --output_dir $output_dir \\
-  --overwrite_output_dir \\
-  --logging_steps 10 \\
-  --logging_dir $output_dir \\
-  --deepspeed ds_config.json
-```
+```bibtex
 
-You can also run with `--sharded_ddp`
-```bash  
-cd transformers/examples/text-classification/
-export TASK_NAME=mnli
-python -m torch.distributed.launch --nproc_per_node=8 run_glue.py   --model_name_or_path microsoft/deberta-v2-xxlarge   \\
---task_name $TASK_NAME   --do_train   --do_eval   --max_seq_length 256   --per_device_train_batch_size 8   \\
---learning_rate 3e-6   --num_train_epochs 3   --output_dir /tmp/$TASK_NAME/ --overwrite_output_dir --sharded_ddp --fp16
 ```
 
-
-### Citation
-
-If you find DeBERTa useful for your work, please cite the following paper:
-
-``` latex
+``` bibtex
 @inproceedings{
 he2021deberta,
 title={DEBERTA: DECODING-ENHANCED BERT WITH DISENTANGLED ATTENTION},
@@ -118,4 +62,4 @@ booktitle={International Conference on Learning Representations},
 year={2021},
 url={https://openreview.net/forum?id=XPZIaotutsD}
 }
-```
+```