You can deploy this model and start earning money today!

kanishka/
smolm-autoreg-bpe-counterfactual-babylm-adj_num_freq_balanced-1e-4

Name: smolm-autoreg-bpe-counterfactual-babylm-adj_num_freq_balanced-1e-4
Availability: InStock

generated_from_trainer

kanishka/counterfactual_babylm_aann_excess_adj_removal

model-index

autotrain_compatible

text-generation-inference

endpoints_compatible

Views

No views yet

smolm-autoreg-bpe-counterfactual-babylm-adj_num_freq_balanced-1e-4

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

Loss: 3.4526
Accuracy: 0.4047

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: 0.0001
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
4.0385	1.0	18629	4.2787	0.3095
3.5614	2.0	37258	3.7622	0.3617
3.387	3.0	55887	3.5845	0.3801
3.2838	4.0	74516	3.5198	0.3878
3.2159	5.0	93145	3.4798	0.3934
3.1607	6.0	111774	3.4551	0.3957
3.1171	7.0	130403	3.4418	0.3984
3.0789	8.0	149032	3.4402	0.3994
3.0465	9.0	167661	3.4255	0.4016
3.0191	10.0	186290	3.4157	0.4028
2.9864	11.0	204919	3.4152	0.4030
2.9701	12.0	223548	3.4240	0.4033
2.9439	13.0	242177	3.4312	0.4031
2.9252	14.0	260806	3.4275	0.4035
2.9045	15.0	279435	3.4313	0.4049
2.8853	16.0	298064	3.4346	0.4042
2.8657	17.0	316693	3.4414	0.4045
2.8449	18.0	335322	3.4522	0.4041
2.8304	19.0	353951	3.4497	0.4046
2.8151	20.0	372580	3.4526	0.4047

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-adj_num_freq_balanced-1e-4

smolm-autoreg-bpe-counterfactual-babylm-adj_num_freq_balanced-1e-4

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

kanishka/
smolm-autoreg-bpe-counterfactual-babylm-adj_num_freq_balanced-1e-4