d0de82b464048426c21013a98b94b226

This model is a fine-tuned version of distilbert/distilbert-base-uncased-distilled-squad on the nyu-mll/glue [stsb] dataset. It achieves the following results on the evaluation set:

  • Loss: 0.5263
  • Data Size: 1.0
  • Epoch Runtime: 6.2360
  • Mse: 0.5266
  • Mae: 0.5632
  • R2: 0.7644

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: 5e-05
  • train_batch_size: 8
  • eval_batch_size: 8
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 4
  • total_train_batch_size: 32
  • total_eval_batch_size: 32
  • optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: constant
  • num_epochs: 50

Training results

Training Loss Epoch Step Validation Loss Data Size Epoch Runtime Mse Mae R2
No log 0 0 7.1630 0 0.9963 7.1642 2.2442 -2.2048
No log 1 179 4.1232 0.0078 1.3503 4.1242 1.6843 -0.8449
No log 2 358 2.3754 0.0156 1.2906 2.3763 1.3290 -0.0630
No log 3 537 1.7820 0.0312 1.5784 1.7826 1.1047 0.2026
No log 4 716 1.2457 0.0625 1.7373 1.2458 0.9017 0.4427
No log 5 895 0.6916 0.125 1.9072 0.6918 0.6535 0.6905
0.0938 6 1074 0.8121 0.25 2.5930 0.8121 0.6988 0.6367
0.5726 7 1253 0.5908 0.5 3.6150 0.5911 0.6119 0.7356
0.4264 8.0 1432 0.6125 1.0 6.1224 0.6126 0.5930 0.7260
0.2704 9.0 1611 0.6132 1.0 6.1703 0.6135 0.6034 0.7256
0.2095 10.0 1790 0.5552 1.0 6.1675 0.5555 0.5592 0.7515
0.1598 11.0 1969 0.6219 1.0 6.1374 0.6221 0.5942 0.7217
0.1313 12.0 2148 0.5631 1.0 6.4765 0.5633 0.5739 0.7480
0.107 13.0 2327 0.5578 1.0 6.0775 0.5581 0.5739 0.7503
0.1065 14.0 2506 0.5416 1.0 6.0314 0.5419 0.5551 0.7576
0.1038 15.0 2685 0.5477 1.0 6.1429 0.5480 0.5622 0.7549
0.0869 16.0 2864 0.5574 1.0 6.1720 0.5576 0.5697 0.7506
0.0699 17.0 3043 0.5085 1.0 6.0534 0.5088 0.5404 0.7724
0.0737 18.0 3222 0.6007 1.0 6.0803 0.6010 0.5945 0.7312
0.0612 19.0 3401 0.5205 1.0 6.0794 0.5208 0.5545 0.7670
0.0654 20.0 3580 0.5380 1.0 6.1629 0.5383 0.5534 0.7592
0.0567 21.0 3759 0.5263 1.0 6.2360 0.5266 0.5632 0.7644

Framework versions

  • Transformers 4.57.0
  • Pytorch 2.8.0+cu128
  • Datasets 4.3.0
  • Tokenizers 0.22.1
Downloads last month
2
Safetensors
Model size
67M params
Tensor type
F32
ยท
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for contemmcm/d0de82b464048426c21013a98b94b226

Finetuned
(62)
this model