Hugging Face's logo

ygory
/
roberta-base-on-cuad-finetuned-squad

Question Answering
Transformers
TensorBoard
Safetensors
roberta
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
ygory commited on
Commit
ea16a9b
•
1 Parent(s): 97484de

Model save

Browse files
Files changed (3) hide show
  1. README.md +28 -28
  2. model.safetensors +1 -1
  3. tokenizer.json +11 -2
README.md CHANGED
@@ -15,7 +15,7 @@ should probably proofread and complete it, then remove this comment. -->
15
 
16
  This model is a fine-tuned version of [Rakib/roberta-base-on-cuad](https://huggingface.co/Rakib/roberta-base-on-cuad) on the None dataset.
17
  It achieves the following results on the evaluation set:
18
- - Loss: 0.1095
19
 
20
  ## Model description
21
 
@@ -34,7 +34,7 @@ More information needed
34
  ### Training hyperparameters
35
 
36
  The following hyperparameters were used during training:
37
- - learning_rate: 0.0002
38
  - train_batch_size: 16
39
  - eval_batch_size: 16
40
  - seed: 42
@@ -46,32 +46,32 @@ The following hyperparameters were used during training:
46
 
47
  | Training Loss | Epoch | Step | Validation Loss |
48
  |:-------------:|:------:|:----:|:---------------:|
49
- | 0.5857 | 0.1142 | 50 | 0.1486 |
50
- | 0.6676 | 0.2283 | 100 | 0.1320 |
51
- | 0.7345 | 0.3425 | 150 | 4.5831 |
52
- | 1.3841 | 0.4566 | 200 | 0.1313 |
53
- | 0.1602 | 0.5708 | 250 | 0.1352 |
54
- | 0.1043 | 0.6849 | 300 | 0.1465 |
55
- | 0.1937 | 0.7991 | 350 | 0.1348 |
56
- | 0.1971 | 0.9132 | 400 | 0.1370 |
57
- | 0.1321 | 1.0274 | 450 | 0.1337 |
58
- | 0.2461 | 1.1416 | 500 | 0.1327 |
59
- | 0.2117 | 1.2557 | 550 | 0.1305 |
60
- | 0.1829 | 1.3699 | 600 | 0.1306 |
61
- | 0.1973 | 1.4840 | 650 | 0.1304 |
62
- | 0.1966 | 1.5982 | 700 | 0.1367 |
63
- | 0.1698 | 1.7123 | 750 | 0.1335 |
64
- | 0.1037 | 1.8265 | 800 | 0.1379 |
65
- | 0.0902 | 1.9406 | 850 | 0.1397 |
66
- | 0.2125 | 2.0548 | 900 | 0.1311 |
67
- | 0.0898 | 2.1689 | 950 | 0.1369 |
68
- | 0.1593 | 2.2831 | 1000 | 0.1322 |
69
- | 0.2223 | 2.3973 | 1050 | 0.1300 |
70
- | 0.2266 | 2.5114 | 1100 | 0.1211 |
71
- | 0.1182 | 2.6256 | 1150 | 0.1325 |
72
- | 0.1429 | 2.7397 | 1200 | 0.1280 |
73
- | 0.1822 | 2.8539 | 1250 | 0.1179 |
74
- | 0.1932 | 2.9680 | 1300 | 0.1095 |
75
 
76
 
77
  ### Framework versions
 
15
 
16
  This model is a fine-tuned version of [Rakib/roberta-base-on-cuad](https://huggingface.co/Rakib/roberta-base-on-cuad) on the None dataset.
17
  It achieves the following results on the evaluation set:
18
+ - Loss: 0.0793
19
 
20
  ## Model description
21
 
 
34
  ### Training hyperparameters
35
 
36
  The following hyperparameters were used during training:
37
+ - learning_rate: 2e-05
38
  - train_batch_size: 16
39
  - eval_batch_size: 16
40
  - seed: 42
 
46
 
47
  | Training Loss | Epoch | Step | Validation Loss |
48
  |:-------------:|:------:|:----:|:---------------:|
49
+ | 0.1503 | 0.1142 | 50 | 0.1216 |
50
+ | 0.1305 | 0.2283 | 100 | 0.1138 |
51
+ | 0.1693 | 0.3425 | 150 | 0.1135 |
52
+ | 0.1986 | 0.4566 | 200 | 0.1063 |
53
+ | 0.1089 | 0.5708 | 250 | 0.0963 |
54
+ | 0.0799 | 0.6849 | 300 | 0.1018 |
55
+ | 0.1527 | 0.7991 | 350 | 0.0986 |
56
+ | 0.1387 | 0.9132 | 400 | 0.1064 |
57
+ | 0.0938 | 1.0274 | 450 | 0.0951 |
58
+ | 0.1533 | 1.1416 | 500 | 0.0805 |
59
+ | 0.1329 | 1.2557 | 550 | 0.0800 |
60
+ | 0.1254 | 1.3699 | 600 | 0.0763 |
61
+ | 0.1247 | 1.4840 | 650 | 0.0789 |
62
+ | 0.1185 | 1.5982 | 700 | 0.0817 |
63
+ | 0.0808 | 1.7123 | 750 | 0.0835 |
64
+ | 0.0622 | 1.8265 | 800 | 0.0815 |
65
+ | 0.0455 | 1.9406 | 850 | 0.0809 |
66
+ | 0.0846 | 2.0548 | 900 | 0.0851 |
67
+ | 0.0453 | 2.1689 | 950 | 0.0832 |
68
+ | 0.0808 | 2.2831 | 1000 | 0.0789 |
69
+ | 0.0902 | 2.3973 | 1050 | 0.0793 |
70
+ | 0.0974 | 2.5114 | 1100 | 0.0787 |
71
+ | 0.0508 | 2.6256 | 1150 | 0.0802 |
72
+ | 0.0535 | 2.7397 | 1200 | 0.0835 |
73
+ | 0.0956 | 2.8539 | 1250 | 0.0815 |
74
+ | 0.1126 | 2.9680 | 1300 | 0.0793 |
75
 
76
 
77
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:04f74cd08e66663fe46220637e4c1cfdf733cccfcd1bea1e54ea3de8fafd86ed
3
  size 496250232
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8fd104494a75ea5187d97ac7a7fb15463e7f8776fb16a168a8617127f420ee91
3
  size 496250232
tokenizer.json CHANGED
@@ -4,9 +4,18 @@
4
  "direction": "Right",
5
  "max_length": 384,
6
  "strategy": "OnlySecond",
7
- "stride": 0
 
 
 
 
 
 
 
 
 
 
8
  },
9
- "padding": null,
10
  "added_tokens": [
11
  {
12
  "id": 0,
 
4
  "direction": "Right",
5
  "max_length": 384,
6
  "strategy": "OnlySecond",
7
+ "stride": 128
8
+ },
9
+ "padding": {
10
+ "strategy": {
11
+ "Fixed": 384
12
+ },
13
+ "direction": "Right",
14
+ "pad_to_multiple_of": null,
15
+ "pad_id": 1,
16
+ "pad_type_id": 0,
17
+ "pad_token": "<pad>"
18
  },
 
19
  "added_tokens": [
20
  {
21
  "id": 0,