9bfecc547b76d8fecaa6c32be2e54343

This model is a fine-tuned version of google/gemma-2b on the nyu-mll/glue [qnli] dataset. It achieves the following results on the evaluation set:

• Loss: 2.7438
• Data Size: 0.125
• Epoch Runtime: 89.7536
• Accuracy: 0.5557
• F1 Macro: 0.5555
• Rouge1: 0.5561
• Rouge2: 0.0
• Rougel: 0.5559
• Rougelsum: 0.5557

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 32
• total_eval_batch_size: 32
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Accuracy	F1 Macro	Rouge1	Rouge2	Rougel	Rougelsum
No log	0	0	3.6654	0	8.5171	0.5121	0.4545	0.5119	0.0	0.5121	0.5125
No log	1	3273	2.1014	0.0078	12.8466	0.7307	0.7224	0.7306	0.0	0.7311	0.7305
0.0499	2	6546	2.2790	0.0156	19.0899	0.7232	0.7094	0.7232	0.0	0.7237	0.7232
2.2861	3	9819	2.2956	0.0312	30.4848	0.7384	0.7273	0.7388	0.0	0.7388	0.7382
3.2397	4	13092	3.9930	0.0625	51.7418	0.5057	0.3359	0.5053	0.0	0.5057	0.5056
2.7864	5	16365	2.7438	0.125	89.7536	0.5557	0.5555	0.5561	0.0	0.5559	0.5557

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.3.0
• Tokenizers 0.22.1