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

Loss: 3.4104
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: 211
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
3.6117	1.0	18593	3.7714	0.3583
3.3802	2.0	37186	3.5869	0.3809
3.2575	3.0	55779	3.4841	0.3924
3.1788	4.0	74372	3.4338	0.3977
3.1253	5.0	92965	3.3785	0.4026
3.0782	6.0	111558	3.3718	0.4042
3.0431	7.0	130151	3.3855	0.4055
3.0174	8.0	148744	3.3763	0.4069
2.9828	9.0	167337	3.3690	0.4076
2.9605	10.0	185930	3.3664	0.4086
2.9367	11.0	204523	3.3552	0.4092
2.9087	12.0	223116	3.3665	0.4094
2.8924	13.0	241709	3.3594	0.4103
2.8696	14.0	260302	3.3819	0.4093
2.8511	15.0	278895	3.3592	0.4111
2.8333	16.0	297488	3.3842	0.4101
2.8145	17.0	316081	3.3862	0.4106
2.7944	18.0	334674	3.4020	0.4102
2.778	19.0	353267	3.4046	0.4103
2.7567	20.0	371860	3.4104	0.4102

Framework versions

Transformers 4.38.0
Pytorch 2.3.1+cu121
Datasets 2.16.1
Tokenizers 0.15.2

kanishka/smolm-autoreg-bpe-counterfactual_babylm_aann_low_variability_numeral-seed_211-1e-3

smolm-autoreg-bpe-counterfactual_babylm_aann_low_variability_numeral_211-1e-3

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

kanishka/
smolm-autoreg-bpe-counterfactual_babylm_aann_low_variability_numeral-seed_211-1e-3