peft-flan-t5-mc-question-generation-eduqg
Browse files
README.md
ADDED
@@ -0,0 +1,105 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
base_model: google/flan-t5-base
|
4 |
+
tags:
|
5 |
+
- generated_from_trainer
|
6 |
+
model-index:
|
7 |
+
- name: peft-flan-t5-mc-question-generation-eduqg
|
8 |
+
results: []
|
9 |
+
---
|
10 |
+
|
11 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
12 |
+
should probably proofread and complete it, then remove this comment. -->
|
13 |
+
|
14 |
+
# peft-flan-t5-mc-question-generation-eduqg
|
15 |
+
|
16 |
+
This model is a fine-tuned version of [google/flan-t5-base](https://huggingface.co/google/flan-t5-base) on an unknown dataset.
|
17 |
+
It achieves the following results on the evaluation set:
|
18 |
+
- Loss: 0.9867
|
19 |
+
|
20 |
+
## Model description
|
21 |
+
|
22 |
+
More information needed
|
23 |
+
|
24 |
+
## Intended uses & limitations
|
25 |
+
|
26 |
+
More information needed
|
27 |
+
|
28 |
+
## Training and evaluation data
|
29 |
+
|
30 |
+
More information needed
|
31 |
+
|
32 |
+
## Training procedure
|
33 |
+
|
34 |
+
### Training hyperparameters
|
35 |
+
|
36 |
+
The following hyperparameters were used during training:
|
37 |
+
- learning_rate: 5e-05
|
38 |
+
- train_batch_size: 1
|
39 |
+
- eval_batch_size: 1
|
40 |
+
- seed: 42
|
41 |
+
- gradient_accumulation_steps: 4
|
42 |
+
- total_train_batch_size: 4
|
43 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
44 |
+
- lr_scheduler_type: linear
|
45 |
+
- num_epochs: 7
|
46 |
+
|
47 |
+
### Training results
|
48 |
+
|
49 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
50 |
+
|:-------------:|:-----:|:----:|:---------------:|
|
51 |
+
| 2.548 | 0.15 | 100 | 1.9606 |
|
52 |
+
| 2.0409 | 0.29 | 200 | 1.5282 |
|
53 |
+
| 1.6969 | 0.44 | 300 | 1.2851 |
|
54 |
+
| 1.4654 | 0.59 | 400 | 1.1794 |
|
55 |
+
| 1.3643 | 0.73 | 500 | 1.1358 |
|
56 |
+
| 1.3305 | 0.88 | 600 | 1.1068 |
|
57 |
+
| 1.2952 | 1.03 | 700 | 1.0873 |
|
58 |
+
| 1.211 | 1.17 | 800 | 1.0703 |
|
59 |
+
| 1.2123 | 1.32 | 900 | 1.0590 |
|
60 |
+
| 1.229 | 1.47 | 1000 | 1.0505 |
|
61 |
+
| 1.2136 | 1.61 | 1100 | 1.0464 |
|
62 |
+
| 1.1835 | 1.76 | 1200 | 1.0402 |
|
63 |
+
| 1.217 | 1.91 | 1300 | 1.0319 |
|
64 |
+
| 1.1986 | 2.05 | 1400 | 1.0300 |
|
65 |
+
| 1.1586 | 2.2 | 1500 | 1.0257 |
|
66 |
+
| 1.1634 | 2.35 | 1600 | 1.0214 |
|
67 |
+
| 1.1661 | 2.49 | 1700 | 1.0170 |
|
68 |
+
| 1.166 | 2.64 | 1800 | 1.0157 |
|
69 |
+
| 1.1663 | 2.79 | 1900 | 1.0126 |
|
70 |
+
| 1.1367 | 2.93 | 2000 | 1.0107 |
|
71 |
+
| 1.1576 | 3.08 | 2100 | 1.0100 |
|
72 |
+
| 1.1314 | 3.23 | 2200 | 1.0070 |
|
73 |
+
| 1.1198 | 3.37 | 2300 | 1.0049 |
|
74 |
+
| 1.1712 | 3.52 | 2400 | 1.0019 |
|
75 |
+
| 1.1369 | 3.67 | 2500 | 1.0028 |
|
76 |
+
| 1.1604 | 3.82 | 2600 | 1.0005 |
|
77 |
+
| 1.1101 | 3.96 | 2700 | 0.9977 |
|
78 |
+
| 1.1227 | 4.11 | 2800 | 0.9968 |
|
79 |
+
| 1.1247 | 4.26 | 2900 | 0.9963 |
|
80 |
+
| 1.1211 | 4.4 | 3000 | 0.9952 |
|
81 |
+
| 1.1398 | 4.55 | 3100 | 0.9939 |
|
82 |
+
| 1.1333 | 4.7 | 3200 | 0.9928 |
|
83 |
+
| 1.1367 | 4.84 | 3300 | 0.9918 |
|
84 |
+
| 1.1124 | 4.99 | 3400 | 0.9919 |
|
85 |
+
| 1.1298 | 5.14 | 3500 | 0.9905 |
|
86 |
+
| 1.117 | 5.28 | 3600 | 0.9904 |
|
87 |
+
| 1.1237 | 5.43 | 3700 | 0.9886 |
|
88 |
+
| 1.1016 | 5.58 | 3800 | 0.9889 |
|
89 |
+
| 1.1492 | 5.72 | 3900 | 0.9887 |
|
90 |
+
| 1.1032 | 5.87 | 4000 | 0.9882 |
|
91 |
+
| 1.0838 | 6.02 | 4100 | 0.9879 |
|
92 |
+
| 1.0807 | 6.16 | 4200 | 0.9880 |
|
93 |
+
| 1.1062 | 6.31 | 4300 | 0.9879 |
|
94 |
+
| 1.131 | 6.46 | 4400 | 0.9873 |
|
95 |
+
| 1.1159 | 6.6 | 4500 | 0.9872 |
|
96 |
+
| 1.1361 | 6.75 | 4600 | 0.9868 |
|
97 |
+
| 1.1051 | 6.9 | 4700 | 0.9867 |
|
98 |
+
|
99 |
+
|
100 |
+
### Framework versions
|
101 |
+
|
102 |
+
- Transformers 4.33.2
|
103 |
+
- Pytorch 2.0.0
|
104 |
+
- Datasets 2.1.0
|
105 |
+
- Tokenizers 0.13.3
|