EC2 Default User
commited on
Commit
•
04f5e6c
1
Parent(s):
05bbebd
add model
Browse files- .gitignore +1 -0
- README.md +60 -0
- config.json +25 -0
- log.log +13 -0
- pytorch_model.bin +3 -0
- score.txt +0 -0
- special_tokens_map.json +1 -0
- tokenizer.json +0 -0
- tokenizer_config.json +1 -0
- training_args.bin +3 -0
- vocab.txt +0 -0
.gitignore
ADDED
@@ -0,0 +1 @@
|
|
|
1 |
+
checkpoint-*/
|
README.md
ADDED
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
tags:
|
3 |
+
- generated_from_trainer
|
4 |
+
datasets:
|
5 |
+
- conll2003
|
6 |
+
model_index:
|
7 |
+
- name: distilbert-base-uncased-squad2-with-ner-with-neg
|
8 |
+
results:
|
9 |
+
- task:
|
10 |
+
name: Question Answering
|
11 |
+
type: question-answering
|
12 |
+
dataset:
|
13 |
+
name: conll2003
|
14 |
+
type: conll2003
|
15 |
+
args: conll2003
|
16 |
+
---
|
17 |
+
|
18 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
19 |
+
should probably proofread and complete it, then remove this comment. -->
|
20 |
+
|
21 |
+
# distilbert-base-uncased-squad2-with-ner-with-neg
|
22 |
+
|
23 |
+
This model is a fine-tuned version of [twmkn9/distilbert-base-uncased-squad2](https://huggingface.co/twmkn9/distilbert-base-uncased-squad2) on the conll2003 dataset.
|
24 |
+
|
25 |
+
## Model description
|
26 |
+
|
27 |
+
More information needed
|
28 |
+
|
29 |
+
## Intended uses & limitations
|
30 |
+
|
31 |
+
More information needed
|
32 |
+
|
33 |
+
## Training and evaluation data
|
34 |
+
|
35 |
+
More information needed
|
36 |
+
|
37 |
+
## Training procedure
|
38 |
+
|
39 |
+
### Training hyperparameters
|
40 |
+
|
41 |
+
The following hyperparameters were used during training:
|
42 |
+
- learning_rate: 3e-05
|
43 |
+
- train_batch_size: 32
|
44 |
+
- eval_batch_size: 8
|
45 |
+
- seed: 42
|
46 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
47 |
+
- lr_scheduler_type: linear
|
48 |
+
- lr_scheduler_warmup_steps: 500
|
49 |
+
- num_epochs: 2.0
|
50 |
+
|
51 |
+
### Training results
|
52 |
+
|
53 |
+
|
54 |
+
|
55 |
+
### Framework versions
|
56 |
+
|
57 |
+
- Transformers 4.8.2
|
58 |
+
- Pytorch 1.8.1+cu111
|
59 |
+
- Datasets 1.8.0
|
60 |
+
- Tokenizers 0.10.3
|
config.json
ADDED
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "twmkn9/distilbert-base-uncased-squad2",
|
3 |
+
"activation": "gelu",
|
4 |
+
"architectures": [
|
5 |
+
"DistilBertForQuestionAnswering"
|
6 |
+
],
|
7 |
+
"attention_dropout": 0.1,
|
8 |
+
"dim": 768,
|
9 |
+
"dropout": 0.1,
|
10 |
+
"eos_token_ids": null,
|
11 |
+
"hidden_dim": 3072,
|
12 |
+
"initializer_range": 0.02,
|
13 |
+
"max_position_embeddings": 512,
|
14 |
+
"model_type": "distilbert",
|
15 |
+
"n_heads": 12,
|
16 |
+
"n_layers": 6,
|
17 |
+
"output_past": true,
|
18 |
+
"pad_token_id": 0,
|
19 |
+
"qa_dropout": 0.1,
|
20 |
+
"seq_classif_dropout": 0.2,
|
21 |
+
"sinusoidal_pos_embds": false,
|
22 |
+
"tie_weights_": true,
|
23 |
+
"transformers_version": "4.8.2",
|
24 |
+
"vocab_size": 30522
|
25 |
+
}
|
log.log
ADDED
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
Training dataset length:
|
2 |
+
56164
|
3 |
+
Validation dataset length:
|
4 |
+
13000
|
5 |
+
Test dataset length:
|
6 |
+
13812
|
7 |
+
Current eval performance:
|
8 |
+
Evaluate results:
|
9 |
+
{'eval_exact': 63.38461538461539, 'eval_f1': 65.59405430995848, 'eval_total': 13000, 'eval_HasAns_exact': 42.352654651729175, 'eval_HasAns_f1': 49.34795568179736, 'eval_HasAns_total': 4106, 'eval_NoAns_exact': 73.09422082302676, 'eval_NoAns_f1': 73.09422082302676, 'eval_NoAns_total': 8894, 'eval_best_exact': 68.5, 'eval_best_exact_thresh': 0.0, 'eval_best_f1': 68.52960805860803, 'eval_best_f1_thresh': 0.0, 'eval_samples': 13000}
|
10 |
+
Predict results:
|
11 |
+
{'test_exact': 63.90095569070374, 'test_f1': 66.14471389068335, 'test_total': 13812, 'test_HasAns_exact': 39.21136419299535, 'test_HasAns_f1': 46.80156459909871, 'test_HasAns_total': 4083, 'test_NoAns_exact': 74.26251413300442, 'test_NoAns_f1': 74.26251413300442, 'test_NoAns_total': 9729, 'test_best_exact': 70.55459021141037, 'test_best_exact_thresh': 0.0, 'test_best_f1': 70.6760355196497, 'test_best_f1_thresh': 0.0, 'predict_samples': 13812}
|
12 |
+
Best trial:
|
13 |
+
BestRun(run_id='8', objective=97.27663003662997, hyperparameters={'learning_rate': 5e-05, 'num_train_epochs': 5, 'per_device_train_batch_size': 16, 'warmup_steps': 1000})
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:84fec087a35d64361b8962282795a92e05c679c935ed8dade2aab709590053d5
|
3 |
+
size 265500101
|
score.txt
ADDED
File without changes
|
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, "max_len": 512, "special_tokens_map_file": "/home/ec2-user/.cache/huggingface/transformers/1ca0b152f7e9e7bd279fded5ffd0930ddafcdc395cbebe37fee225d756a796d3.dd8bd9bfd3664b530ea4e645105f557769387b3da9f79bdb55ed556bdd80611d", "name_or_path": "twmkn9/distilbert-base-uncased-squad2", "do_basic_tokenize": true, "never_split": null, "tokenizer_class": "DistilBertTokenizer"}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4366c7e4ace35cec3919d0e9777550a4966b2d8f8a62a6cba073e80b97e87b2c
|
3 |
+
size 2799
|
vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|