IProject-10 commited on
Commit
21a724b
1 Parent(s): bd7b574

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +59 -17
README.md CHANGED
@@ -8,31 +8,70 @@ datasets:
8
  model-index:
9
  - name: albert-base-v2-finetuned-squad2
10
  results: []
 
 
 
 
 
 
11
  ---
12
 
13
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
14
  should probably proofread and complete it, then remove this comment. -->
15
 
16
- # albert-base-v2-finetuned-squad2
17
-
18
- This model is a fine-tuned version of [albert-base-v2](https://huggingface.co/albert-base-v2) on the squad_v2 dataset.
19
- It achieves the following results on the evaluation set:
20
- - Loss: 0.9900
21
-
22
  ## Model description
23
 
24
- More information needed
25
-
26
- ## Intended uses & limitations
27
-
28
- More information needed
29
-
30
- ## Training and evaluation data
31
 
32
- More information needed
 
 
 
 
 
33
 
34
- ## Training procedure
35
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
36
  ### Training hyperparameters
37
 
38
  The following hyperparameters were used during training:
@@ -52,10 +91,13 @@ The following hyperparameters were used during training:
52
  | 0.6593 | 2.0 | 16496 | 0.8548 |
53
  | 0.4314 | 3.0 | 24744 | 0.9900 |
54
 
55
-
 
 
 
56
  ### Framework versions
57
 
58
  - Transformers 4.31.0
59
  - Pytorch 2.0.1+cu118
60
  - Datasets 2.14.3
61
- - Tokenizers 0.13.3
 
8
  model-index:
9
  - name: albert-base-v2-finetuned-squad2
10
  results: []
11
+ language:
12
+ - en
13
+ metrics:
14
+ - exact_match
15
+ - f1
16
+ pipeline_tag: zero-shot-classification
17
  ---
18
 
19
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
20
  should probably proofread and complete it, then remove this comment. -->
21
 
 
 
 
 
 
 
22
  ## Model description
23
 
24
+ ALBERTbase fine-tuned on SQuAD 2.0 : Encoder-based Transformer Language model, pretrained with Parameter Reduction techniques and Sentence Order Prediction<br>
25
+ Suitable for Question-Answering tasks, predicts answer spans within the context provided.<br>
 
 
 
 
 
26
 
27
+ **Language model:** albert-base-v2
28
+ **Language:** English
29
+ **Downstream-task:** Question-Answering
30
+ **Training data:** Train-set SQuAD 2.0
31
+ **Evaluation data:** Evaluation-set SQuAD 2.0
32
+ **Hardware Accelerator used**: GPU Tesla T4
33
 
34
+ ## Intended uses & limitations
35
 
36
+ For Question-Answering -
37
+
38
+ ```python
39
+ !pip install transformers
40
+ from transformers import pipeline
41
+ model_checkpoint = "IProject-10/bert-base-uncased-finetuned-squad2"
42
+ question_answerer = pipeline("question-answering", model=model_checkpoint)
43
+
44
+ context = """
45
+ 🤗 Transformers is backed by the three most popular deep learning libraries — Jax, PyTorch and TensorFlow — with a seamless integration
46
+ between them. It's straightforward to train your models with one before loading them for inference with the other.
47
+ """
48
+
49
+ question = "Which deep learning libraries back 🤗 Transformers?"
50
+ question_answerer(question=question, context=context)
51
+ ```
52
+
53
+ ## Results
54
+
55
+ Evaluation on SQuAD 2.0 validation dataset:
56
+
57
+ ```
58
+ exact: 78.12684241556472,
59
+ f1: 81.54753481344116,
60
+ total: 11873,
61
+ HasAns_exact: 73.80229419703105,
62
+ HasAns_f1: 80.65348867071317,
63
+ HasAns_total: 5928,
64
+ NoAns_exact: 82.4390243902439,
65
+ NoAns_f1: 82.4390243902439,
66
+ NoAns_total: 5945,
67
+ best_exact: 78.12684241556472,
68
+ best_exact_thresh: 0.9990358352661133,
69
+ best_f1: 81.54753481344157,
70
+ best_f1_thresh: 0.9990358352661133,
71
+ total_time_in_seconds: 248.44505145400035,
72
+ samples_per_second: 47.78923923223437,
73
+ latency_in_seconds: 0.020925212789859374
74
+ ```
75
  ### Training hyperparameters
76
 
77
  The following hyperparameters were used during training:
 
91
  | 0.6593 | 2.0 | 16496 | 0.8548 |
92
  | 0.4314 | 3.0 | 24744 | 0.9900 |
93
 
94
+ This model is a fine-tuned version of [albert-base-v2](https://huggingface.co/albert-base-v2) on the squad_v2 dataset.
95
+ It achieves the following results on the evaluation set:
96
+ - Loss: 0.9900
97
+
98
  ### Framework versions
99
 
100
  - Transformers 4.31.0
101
  - Pytorch 2.0.1+cu118
102
  - Datasets 2.14.3
103
+ - Tokenizers 0.13.3