Paper_Imp_Data_Augmented_1

This model is a fine-tuned version of teknium/OpenHermes-2.5-Mistral-7B on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.8094

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: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 200
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss
2.0944	0.08	30	1.6167
1.5308	0.16	60	1.0254
0.9718	0.23	90	0.8645
1.0394	0.31	120	0.8575
0.9591	0.39	150	0.8520
0.7642	0.47	180	0.8452
0.7454	0.54	210	0.8407
0.7823	0.62	240	0.8461
0.8669	0.7	270	0.8369
0.8827	0.78	300	0.8305
0.7243	0.85	330	0.8267
0.9427	0.93	360	0.8180
0.7446	1.01	390	0.8181
0.6774	1.09	420	0.8248
0.7981	1.16	450	0.8191
0.868	1.24	480	0.8147
0.7947	1.32	510	0.8140
0.7324	1.4	540	0.8122
0.8008	1.47	570	0.8114
0.749	1.55	600	0.8114
0.8133	1.63	630	0.8113
0.7606	1.71	660	0.8118
0.7209	1.78	690	0.8103
0.7236	1.86	720	0.8096
0.6766	1.94	750	0.8094

Framework versions

• PEFT 0.7.1
• Transformers 4.36.2
• Pytorch 2.1.2+cu121
• Datasets 2.16.1
• Tokenizers 0.15.0