llama3_8b_helpsteer_coherence_rm

This model is a fine-tuned version of Jennny/llama3_8b_sft_helpsteer on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.1530
• Accuracy: 0.5918

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: 1e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• gradient_accumulation_steps: 8
• total_train_batch_size: 128
• total_eval_batch_size: 16
• optimizer: Use OptimizerNames.PAGED_ADAMW 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: 0.03
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.436	0.3042	10	0.9576	0.5765
0.14	0.6084	20	1.1787	0.6071
0.146	0.9125	30	0.8044	0.5561
0.0376	1.2433	40	1.5338	0.5714
0.0054	1.5475	50	1.1432	0.6020
0.0353	1.8517	60	1.0350	0.6173
0.0009	2.1825	70	1.1095	0.5867
0.0004	2.4867	80	1.1308	0.5867
0.0005	2.7909	90	1.1530	0.5918

Framework versions

• Transformers 4.49.0
• Pytorch 2.5.1+cu124
• Datasets 3.3.2
• Tokenizers 0.21.1