metadata

tags:
  - generated_from_trainer
datasets:
  - kanishka/counterfactual_babylm_aann_high_variability_numeral
metrics:
  - accuracy
model-index:
  - name: >-
      smolm-autoreg-bpe-counterfactual_babylm_aann_high_variability_numeral-seed_1024-1e-3
    results:
      - task:
          name: Causal Language Modeling
          type: text-generation
        dataset:
          name: kanishka/counterfactual_babylm_aann_high_variability_numeral
          type: kanishka/counterfactual_babylm_aann_high_variability_numeral
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.41021489963935376

smolm-autoreg-bpe-counterfactual_babylm_aann_high_variability_numeral-seed_1024-1e-3

This model was trained from scratch on the kanishka/counterfactual_babylm_aann_high_variability_numeral dataset. It achieves the following results on the evaluation set:

Loss: 3.4236
Accuracy: 0.4102

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: 0.001
train_batch_size: 32
eval_batch_size: 64
seed: 1024
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 32000
num_epochs: 20.0
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
3.5932	1.0	18596	3.7701	0.3588
3.3833	2.0	37192	3.5597	0.3819
3.2597	3.0	55788	3.4648	0.3927
3.1741	4.0	74384	3.4191	0.3977
3.1213	5.0	92980	3.3967	0.4009
3.0783	6.0	111576	3.3773	0.4050
3.0456	7.0	130172	3.3826	0.4055
3.0126	8.0	148768	3.3547	0.4077
2.9843	9.0	167364	3.3614	0.4083
2.9592	10.0	185960	3.3779	0.4085
2.9367	11.0	204556	3.3604	0.4099
2.9145	12.0	223152	3.3759	0.4097
2.8924	13.0	241748	3.3856	0.4096
2.8757	14.0	260344	3.3844	0.4105
2.8545	15.0	278940	3.3832	0.4107
2.8339	16.0	297536	3.4079	0.4098
2.8157	17.0	316132	3.3884	0.4104
2.7966	18.0	334728	3.4081	0.4105
2.7807	19.0	353324	3.4181	0.4104
2.7595	20.0	371920	3.4236	0.4102

Framework versions

Transformers 4.41.0
Pytorch 2.2.0+cu121
Datasets 2.16.1
Tokenizers 0.19.1