cola-pixel-handwritten-mean-vatrpp-256-128-4-3e-5-15000-420

This model is a fine-tuned version of noniewiem/pixel-handwritten on the GLUE COLA dataset. It achieves the following results on the evaluation set:

• Loss: 0.6172
• Matthews Correlation: 0.0

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: 3e-05
• train_batch_size: 128
• eval_batch_size: 8
• seed: 420
• gradient_accumulation_steps: 4
• total_train_batch_size: 512
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 200
• training_steps: 15000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Matthews Correlation
0.651	6.24	100	0.6248	0.0
0.6339	12.48	200	0.6229	0.0
0.6347	18.72	300	0.6188	0.0
0.6348	24.96	400	0.6178	0.0
0.6383	31.24	500	0.6179	0.0
0.633	37.48	600	0.6183	0.0
0.633	43.72	700	0.6159	0.0
0.6333	49.96	800	0.6181	0.0
0.6369	56.24	900	0.6213	0.0
0.6359	62.48	1000	0.6191	0.0
0.6348	68.72	1100	0.6172	0.0

Framework versions

• Transformers 4.17.0
• Pytorch 2.3.0+cu121
• Datasets 2.0.0
• Tokenizers 0.13.3