codegen-350M-mono-measurement_pred-diamonds-seed1

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

• Loss: 0.4083
• Accuracy: 0.9134
• Accuracy Sensor 0: 0.9153
• Auroc Sensor 0: 0.9651
• Accuracy Sensor 1: 0.9094
• Auroc Sensor 1: 0.9502
• Accuracy Sensor 2: 0.9317
• Auroc Sensor 2: 0.9780
• Accuracy Aggregated: 0.8974
• Auroc Aggregated: 0.9672

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: 2e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 64
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Accuracy Sensor 0	Auroc Sensor 0	Accuracy Sensor 1	Auroc Sensor 1	Accuracy Sensor 2	Auroc Sensor 2	Accuracy Aggregated	Auroc Aggregated
0.2812	0.9997	781	0.2931	0.8747	0.8785	0.9058	0.8806	0.9047	0.8897	0.9331	0.8499	0.9009
0.1938	1.9994	1562	0.2940	0.8844	0.8760	0.9330	0.9017	0.9300	0.9160	0.9574	0.8438	0.9252
0.1202	2.9990	2343	0.2551	0.9080	0.9055	0.9601	0.9119	0.9504	0.9235	0.9757	0.8910	0.9615
0.0779	4.0	3125	0.2902	0.9178	0.9194	0.9667	0.9164	0.9516	0.9309	0.9799	0.9044	0.9680
0.035	4.9984	3905	0.4083	0.9134	0.9153	0.9651	0.9094	0.9502	0.9317	0.9780	0.8974	0.9672

Framework versions

• Transformers 4.41.0
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1