DeBERTa commited on
Commit
5863130
1 Parent(s): 801bd19

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +23 -13
README.md CHANGED
@@ -11,22 +11,32 @@ Please check the [official repository](https://github.com/microsoft/DeBERTa) for
11
 
12
  This the DeBERTa xlarge model(750M) fine-tuned with mnli task.
13
 
14
- #### Fine-tuning on NLU tasks
15
 
16
  We present the dev results on SQuAD 1.1/2.0 and several GLUE benchmark tasks.
17
 
18
- | Model | SQuAD 1.1 | SQuAD 2.0 | MNLI-m/mm | SST-2 | QNLI | CoLA | RTE | MRPC(acc/f1) | QQP |STS-B|
19
- |---------------------------|-----------|-----------|-------------|-------|------|------|--------|--------------|------|-----|
20
- | 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 |
21
- | 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 |
22
- | 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 |
23
- | [DeBERTa-Large](https://huggingface.co/microsoft/deberta-large) | 95.5/90.1 | 90.7/88.0 | 91.3/91.1 | 96.5 | 95.3 | 69.5 | 86.6 | 92.6/94.6 | 92.3 |92.5 |
24
- | [DeBERTa-XLarge](https://huggingface.co/microsoft/deberta-xlarge) | -/- | -/- | 91.5/91.2 | - | - | - | 89.5 | 92.1/94.3 | - |- |
25
- | [DeBERTa-XLarge-V2](https://huggingface.co/microsoft/deberta-xlarge-v2) | - | - | 91.7/91.6 | - | - | - | - | - | - |- |
26
- |**[DeBERTa-XXLarge-V2](https://huggingface.co/microsoft/deberta-xxlarge-v2)**|**96.1/91.4**|**92.2/89.7**|**91.7/91.9**| - | - | - | - | - | - |- |
27
- |**[DeBERTa-XLarge-V2-MNLI](https://huggingface.co/microsoft/deberta-xlarge-v2-mnli)**| - | - | 91.7/91.6 | - | - | - | 93.9 | - | - |- |
28
- |**[DeBERTa-XXLarge-V2-MNLI](https://huggingface.co/microsoft/deberta-xxlarge-v2-mnli)**| - | - |**91.7/91.9**| - | - | - | 93.5 | - | - |- |
29
-
 
 
 
 
 
 
 
 
 
 
30
 
31
 
32
 
 
11
 
12
  This the DeBERTa xlarge model(750M) fine-tuned with mnli task.
13
 
14
+ ### Fine-tuning on NLU tasks
15
 
16
  We present the dev results on SQuAD 1.1/2.0 and several GLUE benchmark tasks.
17
 
18
+ | Model | SQuAD 1.1 | SQuAD 2.0 | MNLI-m/mm | SST-2 | QNLI | CoLA | RTE | MRPC | QQP |STS-B |
19
+ |---------------------------|-----------|-----------|-------------|-------|------|------|--------|-------|-------|------|
20
+ | | F1/EM | F1/EM | Acc | Acc | Acc | MCC | Acc |Acc/F1 |Acc/F1 |P/S |
21
+ | 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/- |
22
+ | 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/- |
23
+ | 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/- |
24
+ | [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 |
25
+ | [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|
26
+ | [DeBERTa-XLarge-V2](https://huggingface.co/microsoft/deberta-xlarge-v2)<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|
27
+ |**[DeBERTa-XXLarge-V2](https://huggingface.co/microsoft/deberta-xxlarge-v2)<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** |
28
+ --------
29
+ #### Notes.
30
+ - <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-XLarge-V2-MNLI](https://huggingface.co/microsoft/deberta-xlarge-v2-mnli), [DeBERTa-XXLarge-V2-MNLI](https://huggingface.co/microsoft/deberta-xxlarge-v2-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.
31
+ - <sup>2</sup> To try the **XXLarge** model with **[HF transformers](https://huggingface.co/transformers/main_classes/trainer.html)**, you need to specify **--sharded_ddp**
32
+
33
+ ```bash
34
+ cd transformers/examples/text-classification/
35
+ export TASK_NAME=mrpc
36
+ python -m torch.distributed.launch --nproc_per_node=8 run_glue.py --model_name_or_path microsoft/deberta-xxlarge-v2 \
37
+ --task_name $TASK_NAME --do_train --do_eval --max_seq_length 128 --per_device_train_batch_size 4 \
38
+ --learning_rate 3e-6 --num_train_epochs 3 --output_dir /tmp/$TASK_NAME/ --overwrite_output_dir --sharded_ddp --fp16
39
+ ```
40
 
41
 
42