core-350

Task	Version	Metric	Value		Stderr
arc_challenge	0	acc	0.2048	±	0.0118
		acc_norm	0.2509	±	0.0127
arc_easy	0	acc	0.4247	±	0.0101
		acc_norm	0.3965	±	0.0100
boolq	1	acc	0.5468	±	0.0087
hellaswag	0	acc	0.2844	±	0.0045
		acc_norm	0.3031	±	0.0046
openbookqa	0	acc	0.1560	±	0.0162
		acc_norm	0.2660	±	0.0198
piqa	0	acc	0.5854	±	0.0115
		acc_norm	0.5762	±	0.0115
winogrande	0	acc	0.4909	±	0.0141

This model is a fine-tuned version of ./core-350 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.8128
• Accuracy: 0.8237

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: 2
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 32
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant
• num_epochs: 10.0

Training results

Framework versions

• Transformers 4.35.0.dev0
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.14.1