distilbert-base-uncased-finetuned-code-snippet-quality-scoring
This model is a fine-tuned version of distilbert-base-uncased on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4070
- Accuracy: 0.8568
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
0.5353 | 0.13 | 1000 | 0.5110 | 0.7574 |
0.4686 | 0.26 | 2000 | 0.4339 | 0.7859 |
0.4517 | 0.39 | 3000 | 0.4240 | 0.8002 |
0.4263 | 0.52 | 4000 | 0.3906 | 0.8169 |
0.4053 | 0.66 | 5000 | 0.3934 | 0.8191 |
0.3867 | 0.79 | 6000 | 0.3859 | 0.8253 |
0.3906 | 0.92 | 7000 | 0.3936 | 0.8335 |
0.3418 | 1.05 | 8000 | 0.3615 | 0.8380 |
0.3418 | 1.18 | 9000 | 0.3585 | 0.8400 |
0.3307 | 1.31 | 10000 | 0.3520 | 0.8432 |
0.3301 | 1.44 | 11000 | 0.3476 | 0.8475 |
0.3275 | 1.57 | 12000 | 0.3511 | 0.8497 |
0.3192 | 1.71 | 13000 | 0.3519 | 0.8540 |
0.3218 | 1.84 | 14000 | 0.3402 | 0.8495 |
0.3199 | 1.97 | 15000 | 0.3375 | 0.8580 |
0.2591 | 2.1 | 16000 | 0.3687 | 0.8568 |
0.2732 | 2.23 | 17000 | 0.3619 | 0.8521 |
0.2681 | 2.36 | 18000 | 0.3574 | 0.8563 |
0.2606 | 2.49 | 19000 | 0.3404 | 0.8581 |
0.2662 | 2.62 | 20000 | 0.3708 | 0.8566 |
0.2685 | 2.76 | 21000 | 0.3743 | 0.8591 |
0.246 | 2.89 | 22000 | 0.3786 | 0.8531 |
0.258 | 3.02 | 23000 | 0.3781 | 0.8578 |
0.2284 | 3.15 | 24000 | 0.3938 | 0.8583 |
0.2206 | 3.28 | 25000 | 0.4121 | 0.8583 |
0.2131 | 3.41 | 26000 | 0.4091 | 0.8575 |
0.2181 | 3.54 | 27000 | 0.4264 | 0.8535 |
0.2289 | 3.67 | 28000 | 0.3998 | 0.8568 |
0.2262 | 3.81 | 29000 | 0.3983 | 0.8580 |
0.2095 | 3.94 | 30000 | 0.4070 | 0.8568 |
Framework versions
- Transformers 4.21.2
- Pytorch 1.12.1+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
- Downloads last month
- 18
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.