0e0c45c5d4a4e0936ddc0c89ed0029a7

This model is a fine-tuned version of google/gemma-2b on the contemmcm/cls_mmlu dataset. It achieves the following results on the evaluation set:

• Loss: 11.7305
• Data Size: 1.0
• Epoch Runtime: 144.5284
• Accuracy: 0.2726
• F1 Macro: 0.2711

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
No log	0	0	8.4322	0	3.8812	0.2613	0.1810
No log	1	438	6.0878	0.0078	4.5330	0.2540	0.2455
No log	2	876	7.0608	0.0156	7.0153	0.2507	0.1053
No log	3	1314	6.0006	0.0312	11.1277	0.2487	0.0996
No log	4	1752	5.8492	0.0625	18.4280	0.2493	0.1140
0.3449	5	2190	5.8626	0.125	26.3440	0.2540	0.1243
0.7592	6	2628	5.6727	0.25	44.5673	0.2453	0.0985
5.6658	7	3066	5.6353	0.5	77.1268	0.2527	0.2088
5.6227	8.0	3504	5.5572	1.0	143.8374	0.2633	0.1723
5.3409	9.0	3942	5.8946	1.0	141.8563	0.2673	0.1616
4.5274	10.0	4380	6.5433	1.0	137.5759	0.2832	0.2522
3.5374	11.0	4818	7.9613	1.0	140.6087	0.2653	0.2576
2.3749	12.0	5256	11.7305	1.0	144.5284	0.2726	0.2711

Framework versions

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