Add distilbert SST2 NNCF model
Browse files- README.md +38 -0
- config.json +33 -0
- ov_config.json +49 -0
- ov_model.bin +3 -0
- ov_model.xml +0 -0
- special_tokens_map.json +7 -0
- tokenizer.json +0 -0
- tokenizer_config.json +16 -0
- vocab.txt +0 -0
README.md
ADDED
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
language: en
|
3 |
+
license: apache-2.0
|
4 |
+
datasets:
|
5 |
+
- sst2
|
6 |
+
- glue
|
7 |
+
tags:
|
8 |
+
- openvino
|
9 |
+
---
|
10 |
+
|
11 |
+
## distilbert-base-uncased-finetuned-sst-2-english
|
12 |
+
|
13 |
+
[distilbert-base-uncased-finetuned-sst-2-english](https://huggingface.co/distilbert-base-uncased-finetuned-sst-2-english) quantized with NNCF PTQ and exported to the OpenVINO IR.
|
14 |
+
|
15 |
+
**Model Description:** l is This model reaches an accuracy of 90.0 on the validation set. See [ov\_config.json](./ov_config.json) for the quantization config.
|
16 |
+
|
17 |
+
## Usage example
|
18 |
+
|
19 |
+
You can use this model with Transformers *pipeline*.
|
20 |
+
|
21 |
+
```python
|
22 |
+
from transformers import AutoTokenizer, pipeline
|
23 |
+
from optimum.intel.openvino import OVModelForSequenceClassification
|
24 |
+
|
25 |
+
model_id = "helenai/distilbert-base-uncased-finetuned-sst-2-english-ov-int8"
|
26 |
+
model = OVModelForSequenceClassification.from_pretrained(model_id)
|
27 |
+
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
28 |
+
cls_pipe = pipeline("text-classification", model=model, tokenizer=tokenizer)
|
29 |
+
text = "He's a dreadful magician."
|
30 |
+
outputs = cls_pipe(text)
|
31 |
+
print(outputs)
|
32 |
+
```
|
33 |
+
|
34 |
+
Example output:
|
35 |
+
|
36 |
+
```bash
|
37 |
+
[{'label': 'NEGATIVE', 'score': 0.9929909706115723}]
|
38 |
+
```
|
config.json
ADDED
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "distilbert-base-uncased-finetuned-sst-2-english",
|
3 |
+
"activation": "gelu",
|
4 |
+
"architectures": [
|
5 |
+
"DistilBertForSequenceClassification"
|
6 |
+
],
|
7 |
+
"attention_dropout": 0.1,
|
8 |
+
"dim": 768,
|
9 |
+
"dropout": 0.1,
|
10 |
+
"finetuning_task": "sst-2",
|
11 |
+
"hidden_dim": 3072,
|
12 |
+
"id2label": {
|
13 |
+
"0": "NEGATIVE",
|
14 |
+
"1": "POSITIVE"
|
15 |
+
},
|
16 |
+
"initializer_range": 0.02,
|
17 |
+
"label2id": {
|
18 |
+
"NEGATIVE": 0,
|
19 |
+
"POSITIVE": 1
|
20 |
+
},
|
21 |
+
"max_position_embeddings": 512,
|
22 |
+
"model_type": "distilbert",
|
23 |
+
"n_heads": 12,
|
24 |
+
"n_layers": 6,
|
25 |
+
"output_past": true,
|
26 |
+
"pad_token_id": 0,
|
27 |
+
"qa_dropout": 0.1,
|
28 |
+
"seq_classif_dropout": 0.2,
|
29 |
+
"sinusoidal_pos_embds": false,
|
30 |
+
"tie_weights_": true,
|
31 |
+
"transformers_version": "4.22.2",
|
32 |
+
"vocab_size": 30522
|
33 |
+
}
|
ov_config.json
ADDED
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_commit_hash": null,
|
3 |
+
"compression": {
|
4 |
+
"algorithm": "quantization",
|
5 |
+
"ignored_scopes": [
|
6 |
+
"{re}.*Embeddings.*",
|
7 |
+
"{re}.*__add___[0-1]",
|
8 |
+
"{re}.*layer_norm_0",
|
9 |
+
"{re}.*matmul_1",
|
10 |
+
"{re}.*__truediv__*"
|
11 |
+
],
|
12 |
+
"initializer": {
|
13 |
+
"batchnorm_adaptation": {
|
14 |
+
"num_bn_adaptation_samples": 0
|
15 |
+
},
|
16 |
+
"range": {
|
17 |
+
"num_init_samples": 300,
|
18 |
+
"type": "mean_min_max"
|
19 |
+
}
|
20 |
+
},
|
21 |
+
"overflow_fix": "disable",
|
22 |
+
"preset": "mixed",
|
23 |
+
"scope_overrides": {
|
24 |
+
"activations": {
|
25 |
+
"{re}.*matmul_0": {
|
26 |
+
"mode": "symmetric"
|
27 |
+
}
|
28 |
+
}
|
29 |
+
}
|
30 |
+
},
|
31 |
+
"input_info": [
|
32 |
+
{
|
33 |
+
"sample_size": [
|
34 |
+
8,
|
35 |
+
256
|
36 |
+
],
|
37 |
+
"type": "long"
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"sample_size": [
|
41 |
+
8,
|
42 |
+
256
|
43 |
+
],
|
44 |
+
"type": "long"
|
45 |
+
}
|
46 |
+
],
|
47 |
+
"optimum_version": "1.4.0",
|
48 |
+
"transformers_version": "4.22.2"
|
49 |
+
}
|
ov_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e5051f5e3c6a1c8cd9f51755c8691299c9938daa2ef94def5b709950c630677c
|
3 |
+
size 138817460
|
ov_model.xml
ADDED
The diff for this file is too large to render.
See raw diff
|
special_tokens_map.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": "[CLS]",
|
3 |
+
"mask_token": "[MASK]",
|
4 |
+
"pad_token": "[PAD]",
|
5 |
+
"sep_token": "[SEP]",
|
6 |
+
"unk_token": "[UNK]"
|
7 |
+
}
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
tokenizer_config.json
ADDED
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": "[CLS]",
|
3 |
+
"do_basic_tokenize": true,
|
4 |
+
"do_lower_case": true,
|
5 |
+
"mask_token": "[MASK]",
|
6 |
+
"model_max_length": 512,
|
7 |
+
"name_or_path": "distilbert-base-uncased-finetuned-sst-2-english",
|
8 |
+
"never_split": null,
|
9 |
+
"pad_token": "[PAD]",
|
10 |
+
"sep_token": "[SEP]",
|
11 |
+
"special_tokens_map_file": null,
|
12 |
+
"strip_accents": null,
|
13 |
+
"tokenize_chinese_chars": true,
|
14 |
+
"tokenizer_class": "DistilBertTokenizer",
|
15 |
+
"unk_token": "[UNK]"
|
16 |
+
}
|
vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|