v3c_mistral_lora_lastn

This model is a fine-tuned version of peiyi9979/math-shepherd-mistral-7b-prm on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3067
• Accuracy: 0.8592
• Precision: 0.8580
• Recall: 0.5968
• F1: 0.7040

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: 2e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 765837
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• total_eval_batch_size: 32
• 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_ratio: 0.1
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
No log	0	0	0.6026	0.7339	0.6	0.1542	0.2453
0.5522	0.0495	20	0.5768	0.7395	0.5682	0.2964	0.3896
0.4818	0.0990	40	0.4859	0.7761	0.6835	0.3755	0.4847
0.3892	0.1485	60	0.4218	0.7982	0.6766	0.5375	0.5991
0.2916	0.1980	80	0.3747	0.8237	0.7701	0.5296	0.6276
0.2191	0.2475	100	0.3538	0.8304	0.7778	0.5534	0.6467
0.2189	0.2970	120	0.3754	0.8248	0.88	0.4348	0.5820
0.1841	0.3465	140	0.3427	0.8415	0.8438	0.5336	0.6538
0.2144	0.3960	160	0.3301	0.8404	0.8303	0.5415	0.6555
0.2638	0.4455	180	0.3202	0.8470	0.8485	0.5534	0.6699
0.2032	0.4950	200	0.3125	0.8570	0.8370	0.6087	0.7048
0.1703	0.5446	220	0.3295	0.8337	0.8552	0.4901	0.6231
0.175	0.5941	240	0.3116	0.8503	0.8471	0.5692	0.6809
0.1927	0.6436	260	0.3218	0.8459	0.8654	0.5336	0.6601
0.1848	0.6931	280	0.3069	0.8647	0.8659	0.6126	0.7176
0.222	0.7426	300	0.3036	0.8581	0.8613	0.5889	0.6995
0.1693	0.7921	320	0.3096	0.8525	0.8614	0.5652	0.6826
0.1752	0.8416	340	0.3108	0.8503	0.8554	0.5613	0.6778
0.2353	0.8911	360	0.3072	0.8592	0.8580	0.5968	0.7040
0.1984	0.9406	380	0.3078	0.8603	0.8629	0.5968	0.7056
0.2194	0.9901	400	0.3067	0.8592	0.8580	0.5968	0.7040

Framework versions

• PEFT 0.13.2
• Transformers 4.46.0
• Pytorch 2.5.1+cu124
• Datasets 3.1.0
• Tokenizers 0.20.3