<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
20230825050330
This model is a fine-tuned version of bert-large-cased on the super_glue dataset. It achieves the following results on the evaluation set:
- Loss: 0.4125
- Accuracy: 0.7545
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: 0.005
- train_batch_size: 16
- eval_batch_size: 8
- seed: 11
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 80.0
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
No log | 1.0 | 156 | 0.6296 | 0.5235 |
No log | 2.0 | 312 | 0.8911 | 0.4729 |
No log | 3.0 | 468 | 0.6506 | 0.5162 |
0.9735 | 4.0 | 624 | 0.8420 | 0.4765 |
0.9735 | 5.0 | 780 | 0.6867 | 0.5307 |
0.9735 | 6.0 | 936 | 0.6089 | 0.5884 |
0.8059 | 7.0 | 1092 | 0.5402 | 0.6209 |
0.8059 | 8.0 | 1248 | 0.5361 | 0.6245 |
0.8059 | 9.0 | 1404 | 0.7656 | 0.5271 |
0.7362 | 10.0 | 1560 | 0.4564 | 0.7148 |
0.7362 | 11.0 | 1716 | 0.6545 | 0.6209 |
0.7362 | 12.0 | 1872 | 0.5022 | 0.6787 |
0.6634 | 13.0 | 2028 | 0.4531 | 0.7148 |
0.6634 | 14.0 | 2184 | 0.4135 | 0.7256 |
0.6634 | 15.0 | 2340 | 0.4077 | 0.7112 |
0.6634 | 16.0 | 2496 | 0.4737 | 0.6823 |
0.7064 | 17.0 | 2652 | 0.4535 | 0.7509 |
0.7064 | 18.0 | 2808 | 0.4382 | 0.6931 |
0.7064 | 19.0 | 2964 | 0.3934 | 0.7184 |
0.6064 | 20.0 | 3120 | 0.6309 | 0.6498 |
0.6064 | 21.0 | 3276 | 0.4171 | 0.7112 |
0.6064 | 22.0 | 3432 | 0.4834 | 0.7076 |
0.5413 | 23.0 | 3588 | 0.3896 | 0.7184 |
0.5413 | 24.0 | 3744 | 0.4087 | 0.7473 |
0.5413 | 25.0 | 3900 | 0.4606 | 0.7004 |
0.543 | 26.0 | 4056 | 0.4200 | 0.7184 |
0.543 | 27.0 | 4212 | 0.3754 | 0.7365 |
0.543 | 28.0 | 4368 | 0.4998 | 0.7112 |
0.4837 | 29.0 | 4524 | 0.4105 | 0.7329 |
0.4837 | 30.0 | 4680 | 0.4650 | 0.7365 |
0.4837 | 31.0 | 4836 | 0.5529 | 0.6895 |
0.4837 | 32.0 | 4992 | 0.4410 | 0.7617 |
0.4667 | 33.0 | 5148 | 0.4656 | 0.7184 |
0.4667 | 34.0 | 5304 | 0.4388 | 0.7256 |
0.4667 | 35.0 | 5460 | 0.4408 | 0.7473 |
0.429 | 36.0 | 5616 | 0.3913 | 0.7256 |
0.429 | 37.0 | 5772 | 0.5831 | 0.7292 |
0.429 | 38.0 | 5928 | 0.3963 | 0.7220 |
0.4242 | 39.0 | 6084 | 0.4154 | 0.7329 |
0.4242 | 40.0 | 6240 | 0.6203 | 0.6931 |
0.4242 | 41.0 | 6396 | 0.4409 | 0.7653 |
0.3911 | 42.0 | 6552 | 0.4201 | 0.7437 |
0.3911 | 43.0 | 6708 | 0.4647 | 0.7329 |
0.3911 | 44.0 | 6864 | 0.4582 | 0.7292 |
0.3749 | 45.0 | 7020 | 0.4134 | 0.7329 |
0.3749 | 46.0 | 7176 | 0.4969 | 0.7292 |
0.3749 | 47.0 | 7332 | 0.5039 | 0.7545 |
0.3749 | 48.0 | 7488 | 0.4449 | 0.7365 |
0.3442 | 49.0 | 7644 | 0.4380 | 0.7365 |
0.3442 | 50.0 | 7800 | 0.4263 | 0.7473 |
0.3442 | 51.0 | 7956 | 0.4364 | 0.7437 |
0.3162 | 52.0 | 8112 | 0.4414 | 0.7617 |
0.3162 | 53.0 | 8268 | 0.5216 | 0.7329 |
0.3162 | 54.0 | 8424 | 0.5045 | 0.7256 |
0.3127 | 55.0 | 8580 | 0.3838 | 0.7509 |
0.3127 | 56.0 | 8736 | 0.4114 | 0.7437 |
0.3127 | 57.0 | 8892 | 0.5307 | 0.7256 |
0.3128 | 58.0 | 9048 | 0.4306 | 0.7509 |
0.3128 | 59.0 | 9204 | 0.5574 | 0.7184 |
0.3128 | 60.0 | 9360 | 0.3862 | 0.7653 |
0.2924 | 61.0 | 9516 | 0.4506 | 0.7329 |
0.2924 | 62.0 | 9672 | 0.4236 | 0.7473 |
0.2924 | 63.0 | 9828 | 0.4497 | 0.7726 |
0.2924 | 64.0 | 9984 | 0.4775 | 0.7437 |
0.2812 | 65.0 | 10140 | 0.4406 | 0.7509 |
0.2812 | 66.0 | 10296 | 0.4474 | 0.7581 |
0.2812 | 67.0 | 10452 | 0.4287 | 0.7545 |
0.2738 | 68.0 | 10608 | 0.4214 | 0.7581 |
0.2738 | 69.0 | 10764 | 0.4300 | 0.7509 |
0.2738 | 70.0 | 10920 | 0.4047 | 0.7581 |
0.2596 | 71.0 | 11076 | 0.4289 | 0.7509 |
0.2596 | 72.0 | 11232 | 0.4178 | 0.7762 |
0.2596 | 73.0 | 11388 | 0.4273 | 0.7581 |
0.2437 | 74.0 | 11544 | 0.4166 | 0.7653 |
0.2437 | 75.0 | 11700 | 0.4299 | 0.7581 |
0.2437 | 76.0 | 11856 | 0.3947 | 0.7545 |
0.2513 | 77.0 | 12012 | 0.4163 | 0.7581 |
0.2513 | 78.0 | 12168 | 0.4200 | 0.7545 |
0.2513 | 79.0 | 12324 | 0.4037 | 0.7617 |
0.2513 | 80.0 | 12480 | 0.4125 | 0.7545 |
Framework versions
- Transformers 4.26.1
- Pytorch 2.0.1+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3