CoolFace
Modelpublic

kanishka/smolm-autoreg-bpe-counterfactual-babylm-indef-removal-3e-4

sourceHugging Faceupdated 3y agoView on Hugging Face
0likes33downloads
Model Card

<!-- 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. -->

smolm-autoreg-bpe-counterfactual-babylm-indef-removal-3e-4

This model was trained from scratch on the None dataset. It achieves the following results on the evaluation set:

  • —Loss: 3.4106
  • —Accuracy: 0.4085

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: 0.0003
  • —trainbatchsize: 32
  • —evalbatchsize: 64
  • —seed: 42
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: linear
  • —lrschedulerwarmup_steps: 32000
  • —num_epochs: 20.0
  • —mixedprecisiontraining: Native AMP

Training results

Training LossEpochStepValidation LossAccuracy
3.73781.0185953.89440.3462
3.4322.0371903.62050.3749
3.2923.0557853.48840.3891
3.21054.0743803.42250.3963
3.14755.0929753.40460.3996
3.09926.01115703.40070.4017
3.06317.01301653.36530.4045
3.02918.01487603.36180.4058
3.00169.01673553.35730.4066
2.975110.01859503.35440.4072
2.947911.02045453.35450.4072
2.927212.02231403.36690.4073
2.90613.02417353.36450.4082
2.886314.02603303.36300.4088
2.865615.02789253.37190.4092
2.846316.02975203.38680.4083
2.823917.03161153.38300.4089
2.806718.03347103.39370.4089
2.788619.03533053.40570.4085
2.773220.03719003.41060.4085

Framework versions

  • —Transformers 4.35.0
  • —Pytorch 2.1.0+cu121
  • —Datasets 2.12.0
  • —Tokenizers 0.14.1