yuwenz commited on
Commit
15f53e0
1 Parent(s): 2da0fcd

upload int8 onnx model

Browse files

Signed-off-by: yuwenzho <yuwen.zhou@intel.com>

README.md CHANGED
@@ -1,3 +1,38 @@
1
  ---
2
  license: mit
 
 
 
 
 
 
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: mit
3
+ tags:
4
+ - text-classfication
5
+ - int8
6
+ - Intel® Neural Compressor
7
+ - neural-compressor
8
+ - PostTrainingDynamic
9
+ - onnx
10
+ datasets:
11
+ - glue
12
+ metrics:
13
+ - f1
14
  ---
15
+ # INT8 deberta-v3-base-mrpc
16
+
17
+ ## Post-training Dynamic quantization
18
+
19
+ ### ONNX
20
+
21
+ This is an INT8 ONNX model quantized with [Intel® Neural Compressor](https://github.com/intel/neural-compressor).
22
+
23
+ The original fp32 model comes from the fine-tuned model [Intel/deberta-v3-base-mrpc](https://huggingface.co/Intel/deberta-v3-base-mrpc).
24
+
25
+ #### Test result
26
+
27
+ | |INT8|FP32|
28
+ |---|:---:|:---:|
29
+ | **Accuracy (eval-f1)** |0.9205|0.9223|
30
+ | **Model size (MB)** |350|705|
31
+
32
+
33
+ #### Load ONNX model:
34
+
35
+ ```python
36
+ from optimum.onnxruntime import ORTModelForSequenceClassification
37
+ model = ORTModelForSequenceClassification.from_pretrained('Intel/deberta-v3-base-mrpc-int8-dynamic')
38
+ ```
added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "[MASK]": 128000
3
+ }
config.json ADDED
@@ -0,0 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "Intel/deberta-v3-base-mrpc",
3
+ "architectures": [
4
+ "DebertaV2ForSequenceClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "finetuning_task": "mrpc",
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "not_equivalent",
13
+ "1": "equivalent"
14
+ },
15
+ "initializer_range": 0.02,
16
+ "intermediate_size": 3072,
17
+ "label2id": {
18
+ "equivalent": 1,
19
+ "not_equivalent": 0
20
+ },
21
+ "layer_norm_eps": 1e-07,
22
+ "max_position_embeddings": 512,
23
+ "max_relative_positions": -1,
24
+ "model_type": "deberta-v2",
25
+ "norm_rel_ebd": "layer_norm",
26
+ "num_attention_heads": 12,
27
+ "num_hidden_layers": 12,
28
+ "pad_token_id": 0,
29
+ "pooler_dropout": 0,
30
+ "pooler_hidden_act": "gelu",
31
+ "pooler_hidden_size": 768,
32
+ "pos_att_type": [
33
+ "p2c",
34
+ "c2p"
35
+ ],
36
+ "position_biased_input": false,
37
+ "position_buckets": 256,
38
+ "relative_attention": true,
39
+ "share_att_key": true,
40
+ "torch_dtype": "float32",
41
+ "transformers_version": "4.29.1",
42
+ "type_vocab_size": 0,
43
+ "vocab_size": 128100
44
+ }
model.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7e32e2dd88de9085c9c23e3ab289d64288df1030a401bfd2568518c4b91db196
3
+ size 366330774
special_tokens_map.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "[CLS]",
3
+ "cls_token": "[CLS]",
4
+ "eos_token": "[SEP]",
5
+ "mask_token": "[MASK]",
6
+ "pad_token": "[PAD]",
7
+ "sep_token": "[SEP]",
8
+ "unk_token": "[UNK]"
9
+ }
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
+ "bos_token": "[CLS]",
3
+ "clean_up_tokenization_spaces": true,
4
+ "cls_token": "[CLS]",
5
+ "do_lower_case": false,
6
+ "eos_token": "[SEP]",
7
+ "mask_token": "[MASK]",
8
+ "model_max_length": 1000000000000000019884624838656,
9
+ "pad_token": "[PAD]",
10
+ "sep_token": "[SEP]",
11
+ "sp_model_kwargs": {},
12
+ "split_by_punct": false,
13
+ "tokenizer_class": "DebertaV2Tokenizer",
14
+ "unk_token": "[UNK]",
15
+ "vocab_type": "spm"
16
+ }