You can deploy this model and start earning money today!

kanishka/
smolm-autoreg-bpe-counterfactual-babylm-aann-no_prototypical-3e-4

Name: smolm-autoreg-bpe-counterfactual-babylm-aann-no_prototypical-3e-4
Availability: InStock

generated_from_trainer

kanishka/counterfactual_babylm_without_prototypical

model-index

autotrain_compatible

text-generation-inference

endpoints_compatible

Views

No views yet

smolm-autoreg-bpe-counterfactual-babylm-aann-no_prototypical-3e-4

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

Loss: 3.4242
Accuracy: 0.4087

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
train_batch_size: 32
eval_batch_size: 64
seed: 42
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
3.738	1.0	18593	3.8623	0.3488
3.4383	2.0	37186	3.6089	0.3755
3.2925	3.0	55779	3.5010	0.3888
3.2055	4.0	74372	3.4100	0.3959
3.1461	5.0	92965	3.4109	0.4001
3.1029	6.0	111558	3.3880	0.4016
3.0643	7.0	130151	3.3745	0.4044
3.0298	8.0	148744	3.3771	0.4053
2.9984	9.0	167337	3.3831	0.4060
2.9747	10.0	185930	3.4011	0.4065
2.9573	11.0	204523	3.3670	0.4077
2.925	12.0	223116	3.3797	0.4083
2.9085	13.0	241709	3.3926	0.4079
2.8902	14.0	260302	3.3865	0.4088
2.8665	15.0	278895	3.3980	0.4086
2.8463	16.0	297488	3.3972	0.4092
2.8272	17.0	316081	3.4006	0.4089
2.8085	18.0	334674	3.4044	0.4090
2.7882	19.0	353267	3.4141	0.4088
2.7754	20.0	371860	3.4242	0.4087

Framework versions

Transformers 4.36.0
Pytorch 2.1.1+cu121
Datasets 2.15.0
Tokenizers 0.15.0

kanishka/smolm-autoreg-bpe-counterfactual-babylm-aann-no_prototypical-3e-4

smolm-autoreg-bpe-counterfactual-babylm-aann-no_prototypical-3e-4

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

kanishka/
smolm-autoreg-bpe-counterfactual-babylm-aann-no_prototypical-3e-4