julien-c HF staff commited on
Commit
0036b36
1 Parent(s): 19c1de8

Migrate model card from transformers-repo

Browse files

Read announcement at https://discuss.huggingface.co/t/announcement-all-model-cards-will-be-migrated-to-hf-co-model-repos/2755
Original file history: https://github.com/huggingface/transformers/commits/master/model_cards/microsoft/MiniLM-L12-H384-uncased/README.md

Files changed (1) hide show
  1. README.md +43 -0
README.md ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ thumbnail: https://huggingface.co/front/thumbnails/microsoft.png
3
+ tags:
4
+ - text-classification
5
+ license: mit
6
+ ---
7
+
8
+ ## MiniLM: Small and Fast Pre-trained Models for Language Understanding and Generation
9
+
10
+ MiniLM is a distilled model from the paper "[MiniLM: Deep Self-Attention Distillation for Task-Agnostic Compression of Pre-Trained Transformers](https://arxiv.org/abs/2002.10957)".
11
+
12
+ Please find the information about preprocessing, training and full details of the MiniLM in the [original MiniLM repository](https://github.com/microsoft/unilm/blob/master/minilm/).
13
+
14
+ Please note: This checkpoint can be an inplace substitution for BERT and it needs to be fine-tuned before use!
15
+
16
+ ### English Pre-trained Models
17
+ We release the **uncased** **12**-layer model with **384** hidden size distilled from an in-house pre-trained [UniLM v2](/unilm) model in BERT-Base size.
18
+
19
+ - MiniLMv1-L12-H384-uncased: 12-layer, 384-hidden, 12-heads, 33M parameters, 2.7x faster than BERT-Base
20
+
21
+ #### Fine-tuning on NLU tasks
22
+
23
+ We present the dev results on SQuAD 2.0 and several GLUE benchmark tasks.
24
+
25
+ | Model | #Param | SQuAD 2.0 | MNLI-m | SST-2 | QNLI | CoLA | RTE | MRPC | QQP |
26
+ |---------------------------------------------------|--------|-----------|--------|-------|------|------|------|------|------|
27
+ | [BERT-Base](https://arxiv.org/pdf/1810.04805.pdf) | 109M | 76.8 | 84.5 | 93.2 | 91.7 | 58.9 | 68.6 | 87.3 | 91.3 |
28
+ | **MiniLM-L12xH384** | 33M | 81.7 | 85.7 | 93.0 | 91.5 | 58.5 | 73.3 | 89.5 | 91.3 |
29
+
30
+ ### Citation
31
+
32
+ If you find MiniLM useful in your research, please cite the following paper:
33
+
34
+ ``` latex
35
+ @misc{wang2020minilm,
36
+ title={MiniLM: Deep Self-Attention Distillation for Task-Agnostic Compression of Pre-Trained Transformers},
37
+ author={Wenhui Wang and Furu Wei and Li Dong and Hangbo Bao and Nan Yang and Ming Zhou},
38
+ year={2020},
39
+ eprint={2002.10957},
40
+ archivePrefix={arXiv},
41
+ primaryClass={cs.CL}
42
+ }
43
+ ```