combined_sft_10000_mcq_2epoch

This model is a fine-tuned version of mistralai/Mistral-Nemo-Instruct-2407 on the combined_10000_mcq dataset. It achieves the following results on the evaluation set:

• Loss: 0.0011

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: 0.0001
• train_batch_size: 10
• eval_batch_size: 10
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 40
• total_eval_batch_size: 40
• 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
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss
0.0044	0.0667	30	0.0043
0.0039	0.1333	60	0.0039
0.0039	0.2	90	0.0038
0.0038	0.2667	120	0.0039
0.0038	0.3333	150	0.0037
0.0036	0.4	180	0.0034
0.0034	0.4667	210	0.0033
0.0028	0.5333	240	0.0027
0.0026	0.6	270	0.0024
0.0022	0.6667	300	0.0022
0.0025	0.7333	330	0.0020
0.0019	0.8	360	0.0019
0.0021	0.8667	390	0.0018
0.0017	0.9333	420	0.0017
0.0017	1.0	450	0.0017
0.0017	1.0667	480	0.0017
0.0015	1.1333	510	0.0016
0.0015	1.2	540	0.0016
0.0014	1.2667	570	0.0016
0.0017	1.3333	600	0.0014
0.0014	1.4	630	0.0014
0.0012	1.4667	660	0.0013
0.0012	1.5333	690	0.0013
0.0012	1.6	720	0.0012
0.0011	1.6667	750	0.0012
0.0009	1.7333	780	0.0011
0.0012	1.8	810	0.0011
0.0012	1.8667	840	0.0011
0.0012	1.9333	870	0.0011
0.001	2.0	900	0.0011

Framework versions

• PEFT 0.12.0
• Transformers 4.46.0
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.20.1