chexagent-8b-instruct-trl-sft-ChartQA

This model is a fine-tuned version of StanfordAIMI/CheXagent-8b on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2088

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: 1
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 8
• optimizer: Use adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• lr_scheduler_warmup_ratio: 0.03
• num_epochs: 8

Training results

Training Loss	Epoch	Step	Validation Loss
0.211	0.9999	3437	0.2192
0.2113	2.0	6875	0.2123
0.2069	2.9999	10312	0.2075
0.1995	4.0	13750	0.2108
0.2065	4.9999	17187	0.2112
0.192	6.0	20625	0.2070
0.2032	6.9999	24062	0.2104
0.1958	7.9988	27496	0.2088

Framework versions

• PEFT 0.13.2
• Transformers 4.46.3
• Pytorch 2.4.1+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3