chessgpt2-small-s

This model is a fine-tuned version of gpt2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.9942

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: 0.0004
• train_batch_size: 32
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.04
• num_epochs: 3
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
2.2324	0.2560	1000	1.6718
1.5685	0.5119	2000	1.3914
1.3827	0.7679	3000	1.2685
1.275	1.0238	4000	1.1896
1.1969	1.2798	5000	1.1338
1.1498	1.5357	6000	1.0901
1.1102	1.7917	7000	1.0563
1.0698	2.0476	8000	1.0300
1.0221	2.3036	9000	1.0122
1.0079	2.5595	10000	1.0003
0.9988	2.8155	11000	0.9942

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.1
• Tokenizers 0.19.1