simonycl's picture
update model card README.md
eac1aa2
|
raw
history blame
10.6 kB
metadata
license: mit
base_model: roberta-base
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: best_model-yelp_polarity-32-100
    results: []

best_model-yelp_polarity-32-100

This model is a fine-tuned version of roberta-base on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.4649
  • Accuracy: 0.9531

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: 1e-05
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 500
  • num_epochs: 150

Training results

Training Loss Epoch Step Validation Loss Accuracy
No log 1.0 2 0.4546 0.9531
No log 2.0 4 0.4598 0.9531
No log 3.0 6 0.4661 0.9531
No log 4.0 8 0.4814 0.9375
0.5203 5.0 10 0.4985 0.9375
0.5203 6.0 12 0.5179 0.9375
0.5203 7.0 14 0.5372 0.9375
0.5203 8.0 16 0.5624 0.9375
0.5203 9.0 18 0.5780 0.9375
0.5493 10.0 20 0.6019 0.9375
0.5493 11.0 22 0.6239 0.9375
0.5493 12.0 24 0.6582 0.9219
0.5493 13.0 26 0.7018 0.9219
0.5493 14.0 28 0.7868 0.9062
0.4311 15.0 30 0.8397 0.9062
0.4311 16.0 32 0.8642 0.9062
0.4311 17.0 34 0.8456 0.9062
0.4311 18.0 36 0.7841 0.9062
0.4311 19.0 38 0.6959 0.9062
0.3814 20.0 40 0.6684 0.9062
0.3814 21.0 42 0.6086 0.9219
0.3814 22.0 44 0.5737 0.9375
0.3814 23.0 46 0.5216 0.9375
0.3814 24.0 48 0.4856 0.9375
0.3304 25.0 50 0.4508 0.9531
0.3304 26.0 52 0.4121 0.9531
0.3304 27.0 54 0.3536 0.9531
0.3304 28.0 56 0.2920 0.9688
0.3304 29.0 58 0.2699 0.9688
0.2882 30.0 60 0.2532 0.9688
0.2882 31.0 62 0.2417 0.9688
0.2882 32.0 64 0.2335 0.9688
0.2882 33.0 66 0.2233 0.9688
0.2882 34.0 68 0.2204 0.9688
0.0526 35.0 70 0.2195 0.9688
0.0526 36.0 72 0.2246 0.9688
0.0526 37.0 74 0.2375 0.9688
0.0526 38.0 76 0.2515 0.9688
0.0526 39.0 78 0.2652 0.9688
0.0054 40.0 80 0.2865 0.9531
0.0054 41.0 82 0.3170 0.9531
0.0054 42.0 84 0.3356 0.9531
0.0054 43.0 86 0.3346 0.9531
0.0054 44.0 88 0.3329 0.9531
0.0011 45.0 90 0.3320 0.9531
0.0011 46.0 92 0.3160 0.9531
0.0011 47.0 94 0.3016 0.9531
0.0011 48.0 96 0.2909 0.9688
0.0011 49.0 98 0.2851 0.9688
0.0003 50.0 100 0.2829 0.9688
0.0003 51.0 102 0.2822 0.9688
0.0003 52.0 104 0.2822 0.9688
0.0003 53.0 106 0.2827 0.9688
0.0003 54.0 108 0.2836 0.9688
0.0001 55.0 110 0.2852 0.9688
0.0001 56.0 112 0.2871 0.9688
0.0001 57.0 114 0.2892 0.9688
0.0001 58.0 116 0.2920 0.9688
0.0001 59.0 118 0.2965 0.9688
0.0001 60.0 120 0.3036 0.9688
0.0001 61.0 122 0.3120 0.9531
0.0001 62.0 124 0.3212 0.9531
0.0001 63.0 126 0.3298 0.9531
0.0001 64.0 128 0.3377 0.9531
0.0001 65.0 130 0.3450 0.9531
0.0001 66.0 132 0.3513 0.9531
0.0001 67.0 134 0.3585 0.9531
0.0001 68.0 136 0.3646 0.9531
0.0001 69.0 138 0.3696 0.9531
0.0001 70.0 140 0.3741 0.9531
0.0001 71.0 142 0.3783 0.9531
0.0001 72.0 144 0.3819 0.9531
0.0001 73.0 146 0.3852 0.9531
0.0001 74.0 148 0.3873 0.9531
0.0001 75.0 150 0.3896 0.9531
0.0001 76.0 152 0.3912 0.9531
0.0001 77.0 154 0.3921 0.9531
0.0001 78.0 156 0.3928 0.9531
0.0001 79.0 158 0.3933 0.9531
0.0 80.0 160 0.3939 0.9531
0.0 81.0 162 0.3949 0.9531
0.0 82.0 164 0.3961 0.9531
0.0 83.0 166 0.3973 0.9531
0.0 84.0 168 0.3989 0.9531
0.0 85.0 170 0.4004 0.9531
0.0 86.0 172 0.4020 0.9531
0.0 87.0 174 0.4036 0.9531
0.0 88.0 176 0.4052 0.9531
0.0 89.0 178 0.4067 0.9531
0.0 90.0 180 0.4084 0.9531
0.0 91.0 182 0.4101 0.9531
0.0 92.0 184 0.4118 0.9531
0.0 93.0 186 0.4135 0.9531
0.0 94.0 188 0.4149 0.9531
0.0 95.0 190 0.4163 0.9531
0.0 96.0 192 0.4176 0.9531
0.0 97.0 194 0.4189 0.9531
0.0 98.0 196 0.4204 0.9531
0.0 99.0 198 0.4218 0.9531
0.0 100.0 200 0.4232 0.9531
0.0 101.0 202 0.4246 0.9531
0.0 102.0 204 0.4261 0.9531
0.0 103.0 206 0.4277 0.9531
0.0 104.0 208 0.4291 0.9531
0.0 105.0 210 0.4304 0.9531
0.0 106.0 212 0.4315 0.9531
0.0 107.0 214 0.4327 0.9531
0.0 108.0 216 0.4339 0.9531
0.0 109.0 218 0.4350 0.9531
0.0 110.0 220 0.4362 0.9531
0.0 111.0 222 0.4373 0.9531
0.0 112.0 224 0.4381 0.9531
0.0 113.0 226 0.4391 0.9531
0.0 114.0 228 0.4400 0.9531
0.0 115.0 230 0.4410 0.9531
0.0 116.0 232 0.4421 0.9531
0.0 117.0 234 0.4432 0.9531
0.0 118.0 236 0.4443 0.9531
0.0 119.0 238 0.4453 0.9531
0.0 120.0 240 0.4467 0.9531
0.0 121.0 242 0.4479 0.9531
0.0 122.0 244 0.4489 0.9531
0.0 123.0 246 0.4498 0.9531
0.0 124.0 248 0.4507 0.9531
0.0 125.0 250 0.4514 0.9531
0.0 126.0 252 0.4521 0.9531
0.0 127.0 254 0.4528 0.9531
0.0 128.0 256 0.4534 0.9531
0.0 129.0 258 0.4540 0.9531
0.0 130.0 260 0.4547 0.9531
0.0 131.0 262 0.4553 0.9531
0.0 132.0 264 0.4560 0.9531
0.0 133.0 266 0.4567 0.9531
0.0 134.0 268 0.4574 0.9531
0.0 135.0 270 0.4580 0.9531
0.0 136.0 272 0.4584 0.9531
0.0 137.0 274 0.4589 0.9531
0.0 138.0 276 0.4594 0.9531
0.0 139.0 278 0.4597 0.9531
0.0 140.0 280 0.4602 0.9531
0.0 141.0 282 0.4607 0.9531
0.0 142.0 284 0.4612 0.9531
0.0 143.0 286 0.4616 0.9531
0.0 144.0 288 0.4621 0.9531
0.0 145.0 290 0.4625 0.9531
0.0 146.0 292 0.4630 0.9531
0.0 147.0 294 0.4635 0.9531
0.0 148.0 296 0.4640 0.9531
0.0 149.0 298 0.4644 0.9531
0.0 150.0 300 0.4649 0.9531

Framework versions

  • Transformers 4.32.0.dev0
  • Pytorch 2.0.1+cu118
  • Datasets 2.4.0
  • Tokenizers 0.13.3