.ipynb_checkpoints/README-checkpoint.md

metadata

base_model: ChaiML/reward_models_100_170000000_cp_498032
tags:
  - generated_from_trainer
model-index:
  - name: reward-model
    results: []

reward-model

This model is a fine-tuned version of ChaiML/reward_models_100_170000000_cp_498032 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.5733

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: 3
• eval_batch_size: 3
• seed: 7
• gradient_accumulation_steps: 16
• total_train_batch_size: 48
• 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
0.5918	0.12	200	0.6011
0.5998	0.24	400	0.5911
0.59	0.36	600	0.5863
0.585	0.49	800	0.5823
0.5646	0.61	1000	0.5803
0.6048	0.73	1200	0.5773
0.6041	0.85	1400	0.5784
0.5694	0.97	1600	0.5747
0.5855	1.09	1800	0.5767
0.5764	1.22	2000	0.5764
0.5606	1.34	2200	0.5799
0.5583	1.46	2400	0.5778
0.5656	1.58	2600	0.5731
0.5812	1.7	2800	0.5747
0.5911	1.82	3000	0.5731
0.5771	1.94	3200	0.5733

Framework versions

• Transformers 4.36.0.dev0
• Pytorch 2.0.1+cu117
• Datasets 2.15.0
• Tokenizers 0.15.0