testner

This model is a fine-tuned version of deepset/gbert-large on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 1.3944
  • Precision: 0.2579
  • Recall: 0.2364
  • F1: 0.2467
  • Accuracy: 0.8626

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: 8
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 100
  • num_epochs: 50

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Accuracy
No log 1.0 229 0.5531 0.0305 0.0133 0.0185 0.8440
No log 2.0 458 0.5008 0.1281 0.1055 0.1157 0.8582
0.5648 3.0 687 0.5616 0.1521 0.1648 0.1582 0.8532
0.5648 4.0 916 0.5269 0.1665 0.2315 0.1937 0.8536
0.2466 5.0 1145 0.6401 0.1885 0.2073 0.1975 0.8551
0.2466 6.0 1374 0.6759 0.1944 0.2036 0.1989 0.8592
0.1155 7.0 1603 0.7172 0.1859 0.2206 0.2018 0.8559
0.1155 8.0 1832 0.8176 0.2 0.2194 0.2092 0.8555
0.0612 9.0 2061 0.8450 0.1904 0.2315 0.2090 0.8519
0.0612 10.0 2290 0.9029 0.1895 0.2048 0.1969 0.8535
0.0376 11.0 2519 0.9917 0.2097 0.2194 0.2145 0.8548
0.0376 12.0 2748 0.9464 0.2346 0.2485 0.2413 0.8609
0.0376 13.0 2977 1.0170 0.2295 0.2412 0.2352 0.8585
0.022 14.0 3206 0.9993 0.2259 0.2242 0.2251 0.8590
0.022 15.0 3435 1.0762 0.2194 0.2473 0.2325 0.8528
0.0152 16.0 3664 1.0343 0.2434 0.2364 0.2399 0.8616
0.0152 17.0 3893 1.0420 0.2241 0.2388 0.2312 0.8570
0.0137 18.0 4122 1.1025 0.2214 0.2206 0.2210 0.8610
0.0137 19.0 4351 1.0975 0.2186 0.2339 0.2260 0.8540
0.0099 20.0 4580 1.1521 0.2281 0.2436 0.2356 0.8592
0.0099 21.0 4809 1.1143 0.2080 0.2461 0.2254 0.8527
0.0084 22.0 5038 1.2333 0.2368 0.24 0.2384 0.8567
0.0084 23.0 5267 1.1713 0.2367 0.2364 0.2365 0.8595
0.0084 24.0 5496 1.2162 0.2599 0.2315 0.2449 0.8643
0.0065 25.0 5725 1.1444 0.2467 0.2473 0.2470 0.8600
0.0065 26.0 5954 1.2645 0.2512 0.2545 0.2529 0.8617
0.0046 27.0 6183 1.2562 0.2252 0.2255 0.2253 0.8610
0.0046 28.0 6412 1.2663 0.2516 0.2327 0.2418 0.8615
0.0043 29.0 6641 1.2686 0.2565 0.2497 0.2531 0.8622
0.0043 30.0 6870 1.2411 0.2342 0.2521 0.2428 0.8586
0.0037 31.0 7099 1.2620 0.2553 0.2485 0.2518 0.8626
0.0037 32.0 7328 1.3049 0.2506 0.24 0.2452 0.8593
0.003 33.0 7557 1.2796 0.2516 0.2339 0.2425 0.8633
0.003 34.0 7786 1.3039 0.2484 0.2339 0.2409 0.8625
0.0025 35.0 8015 1.3241 0.2597 0.2436 0.2514 0.8618
0.0025 36.0 8244 1.3132 0.2475 0.2436 0.2456 0.8613
0.0025 37.0 8473 1.3445 0.25 0.2388 0.2443 0.8620
0.002 38.0 8702 1.3669 0.2556 0.2339 0.2443 0.8635
0.002 39.0 8931 1.3566 0.2623 0.2448 0.2533 0.8622
0.0018 40.0 9160 1.3300 0.2447 0.2388 0.2417 0.8620
0.0018 41.0 9389 1.3311 0.2397 0.24 0.2399 0.8624
0.0019 42.0 9618 1.3368 0.2469 0.2412 0.2440 0.8625
0.0019 43.0 9847 1.3701 0.2430 0.2412 0.2421 0.8624
0.0014 44.0 10076 1.3941 0.2286 0.2327 0.2306 0.8619
0.0014 45.0 10305 1.3842 0.2506 0.2352 0.2427 0.8628
0.0013 46.0 10534 1.3827 0.2443 0.2327 0.2384 0.8619
0.0013 47.0 10763 1.3730 0.2506 0.2376 0.2439 0.8632
0.0013 48.0 10992 1.3936 0.2586 0.2364 0.2470 0.8629
0.0011 49.0 11221 1.3941 0.2634 0.2388 0.2505 0.8627
0.0011 50.0 11450 1.3944 0.2579 0.2364 0.2467 0.8626

Framework versions

  • Transformers 4.48.1
  • Pytorch 2.5.1
  • Datasets 3.2.0
  • Tokenizers 0.21.0
Downloads last month
5
Safetensors
Model size
335M params
Tensor type
F32
·
Inference Providers NEW
This model is not currently available via any of the supported third-party Inference Providers, and the model is not deployed on the HF Inference API.

Model tree for MSLars/nonsense-detection

Finetuned
(15)
this model