Xuandong commited on
Commit
c5f1ffa
1 Parent(s): d977059
1_Pooling/config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "word_embedding_dimension": 384,
3
+ "pooling_mode_cls_token": false,
4
+ "pooling_mode_mean_tokens": true,
5
+ "pooling_mode_max_tokens": false,
6
+ "pooling_mode_mean_sqrt_len_tokens": false
7
+ }
2_Dense/config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"in_features": 384, "out_features": 128, "bias": true, "activation_function": "torch.nn.modules.activation.Tanh"}
2_Dense/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:98a29078d83e39ae97f72162b048432ee1a0b2023d8cd12606c85a31742f8c29
3
+ size 198183
README.md CHANGED
@@ -1,3 +1,78 @@
1
  ---
2
  license: apache-2.0
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: apache-2.0
3
  ---
4
+
5
+ # HPD-MiniLM-F128
6
+
7
+ This repository contains the pre-trained models for our paper [Compressing Sentence Representation for Semantic Retrieval via Homomorphic Projective Distillation](https://arxiv.org/abs/2203.07687).
8
+
9
+ ## Overview
10
+
11
+ We propose **H**omomorphic **P**rojective **D**istillation (HPD) to learn compressed sentence embeddings. Our method augments a small Transformer encoder model with learnable projection layers to produce compact representations while mimicking a large pre-trained language model to retain the sentence representation quality.
12
+
13
+ ## Details
14
+
15
+ This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 128 dimensional dense vector space and can be used for tasks like clustering or semantic search.
16
+
17
+ The teacher model is [`princeton-nlp/sup-simcse-roberta-large`](https://huggingface.co/princeton-nlp/sup-simcse-bert-base-uncased) and the student model is [`nreimers/MiniLM-L6-H384-uncased`](https://huggingface.co/nreimers/MiniLM-L6-H384-uncased).
18
+
19
+ ## Usage
20
+
21
+ Using this model becomes easy when you have [sentence-transformers](https://www.SBERT.net) installed:
22
+
23
+ ```
24
+ pip install -U sentence-transformers
25
+ ```
26
+
27
+ After installing the package, you can simply load our model
28
+ ```python
29
+ from sentence_transformers import SentenceTransformer
30
+ model = SentenceTransformer('Xuandong/HPD-MiniLM-F128')
31
+ ```
32
+
33
+ Then you can use our model for **encoding sentences into embeddings**
34
+ ```python
35
+ sentences = ['He plays guitar.', 'A street vendor is outside.']
36
+ sentence_embeddings = model.encode(sentences)
37
+
38
+ for sentence, embedding in zip(sentences, sentence_embeddings):
39
+ print("Sentence:", sentence)
40
+ print("Embedding:", embedding)
41
+ print("")
42
+ ```
43
+
44
+ ## Evaluation Results
45
+
46
+ We evaluate our model on semantic textual similarity (STS) tasks. The results are:
47
+
48
+ | STS12 | STS13 | STS14 | STS15 | STS16 | STS-B | SICK-R | Avg. |
49
+ |-------|-------|-------|-------|-------|--------------|-----------------|-------|
50
+ | 74.94 | 84.52 | 80.25 | 84.87 | 81.90 | 84.98 | 81.15 | 81.80 |
51
+
52
+
53
+ ## Training
54
+
55
+ Please refer to the github repo (https://github.com/XuandongZhao/HPD) for the details about the training.
56
+
57
+
58
+ ## Full Model Architecture
59
+ ```
60
+ SentenceTransformer(
61
+ (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel
62
+ (1): Pooling({'word_embedding_dimension': 384, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False})
63
+ (2): Dense({'in_features': 384, 'out_features': 128, 'bias': True, 'activation_function': 'torch.nn.modules.activation.Tanh'})
64
+ )
65
+ ```
66
+
67
+ ## Citation
68
+
69
+ Please cite our paper if you use HPD in your work:
70
+
71
+ ```bibtex
72
+ @article{zhao2022compressing,
73
+ title={Compressing Sentence Representation for Semantic Retrieval via Homomorphic Projective Distillation},
74
+ author={Zhao, Xuandong and Yu, Zhiguo and Wu, Ming and Li, Lei},
75
+ journal={arXiv preprint arXiv:2203.07687},
76
+ year={2022}
77
+ }
78
+ ```
config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "Xuandong/HPD-MiniLM-F128",
3
+ "architectures": [
4
+ "BertModel"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_dropout": null,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 384,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 1536,
13
+ "layer_norm_eps": 1e-12,
14
+ "max_position_embeddings": 512,
15
+ "model_type": "bert",
16
+ "num_attention_heads": 6,
17
+ "num_hidden_layers": 6,
18
+ "pad_token_id": 0,
19
+ "position_embedding_type": "absolute",
20
+ "torch_dtype": "float32",
21
+ "transformers_version": "4.11.3",
22
+ "type_vocab_size": 2,
23
+ "use_cache": true,
24
+ "vocab_size": 30522
25
+ }
config_sentence_transformers.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "__version__": {
3
+ "sentence_transformers": "2.0.0",
4
+ "transformers": "4.11.3",
5
+ "pytorch": "1.10.0+cu113"
6
+ }
7
+ }
modules.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "idx": 0,
4
+ "name": "0",
5
+ "path": "",
6
+ "type": "sentence_transformers.models.Transformer"
7
+ },
8
+ {
9
+ "idx": 1,
10
+ "name": "1",
11
+ "path": "1_Pooling",
12
+ "type": "sentence_transformers.models.Pooling"
13
+ },
14
+ {
15
+ "idx": 2,
16
+ "name": "2",
17
+ "path": "2_Dense",
18
+ "type": "sentence_transformers.models.Dense"
19
+ }
20
+ ]
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:50a0a61eff4ad95c33b9d69f1fbb3d0c6b3bc8d5d8e79dd554118abac8fc6847
3
+ size 90896561
sentence_bert_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "max_seq_length": 512,
3
+ "do_lower_case": false
4
+ }
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"do_lower_case": true, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "do_basic_tokenize": true, "never_split": null, "model_max_length": 512, "special_tokens_map_file": null, "name_or_path": "Xuandong/HPD-MiniLM-F128", "tokenizer_class": "BertTokenizer"}
vocab.txt ADDED
The diff for this file is too large to render. See raw diff