Se124M10KInfPrompt_endtoken

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

• Loss: 0.6872

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: 0.0005
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 200
• num_epochs: 50
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.8085	1.0	610	0.7760
0.7801	2.0	1220	0.7436
0.7608	3.0	1830	0.7269
0.7438	4.0	2440	0.7199
0.7413	5.0	3050	0.7118
0.7343	6.0	3660	0.7121
0.7332	7.0	4270	0.7089
0.7319	8.0	4880	0.7025
0.7289	9.0	5490	0.7001
0.7236	10.0	6100	0.6965
0.7147	11.0	6710	0.6970
0.7126	12.0	7320	0.6973
0.7167	13.0	7930	0.6935
0.711	14.0	8540	0.6927
0.7057	15.0	9150	0.6940
0.7109	16.0	9760	0.6924
0.7117	17.0	10370	0.6928
0.7086	18.0	10980	0.6882
0.7004	19.0	11590	0.6872
0.7016	20.0	12200	0.6895
0.7027	21.0	12810	0.6884
0.6928	22.0	13420	0.6885
0.7059	23.0	14030	0.6894
0.6916	24.0	14640	0.6875

Framework versions

• PEFT 0.15.1
• Transformers 4.51.3
• Pytorch 2.6.0+cu118
• Datasets 3.5.0
• Tokenizers 0.21.1