Phi-4-multimodal-instruct-golf-scorecard

This model is a fine-tuned version of microsoft/Phi-4-multimodal-instruct on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0334

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: 0.0002
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 4
• optimizer: Use OptimizerNames.PAGED_ADAMW_8BIT with betas=(0.9,0.95) and epsilon=1e-07 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 50
• num_epochs: 7

Training results

Training Loss	Epoch	Step	Validation Loss
0.0854	0.1998	50	0.1247
0.1045	0.3996	100	0.0834
0.0539	0.5994	150	0.0630
0.0511	0.7992	200	0.0521
0.0402	0.9990	250	0.0527
0.0346	1.1958	300	0.0540
0.0331	1.3956	350	0.0665
0.0324	1.5954	400	0.0443
0.0279	1.7952	450	0.0421
0.0244	1.9950	500	0.0408
0.0199	2.1918	550	0.0378
0.0204	2.3916	600	0.0408
0.0167	2.5914	650	0.0379
0.0183	2.7912	700	0.0338
0.0138	2.9910	750	0.0313
0.0108	3.1878	800	0.0325
0.0111	3.3876	850	0.0364
0.0092	3.5874	900	0.0301
0.0082	3.7872	950	0.0293
0.0116	3.9870	1000	0.0278
0.006	4.1838	1050	0.0330
0.0053	4.3836	1100	0.0285
0.0067	4.5834	1150	0.0279
0.006	4.7832	1200	0.0285
0.0039	4.9830	1250	0.0277
0.0033	5.1798	1300	0.0290
0.0028	5.3796	1350	0.0314
0.0017	5.5794	1400	0.0316
0.003	5.7792	1450	0.0310
0.0017	5.9790	1500	0.0288
0.0013	6.1758	1550	0.0306
0.0014	6.3756	1600	0.0323
0.0007	6.5754	1650	0.0330
0.0012	6.7752	1700	0.0333
0.0006	6.9750	1750	0.0334

Framework versions

• Transformers 4.48.2
• Pytorch 2.6.0+cu124
• Datasets 3.4.1
• Tokenizers 0.21.1