Acadys/PointCon-Vigogne33B

PointCon-vigogne-33b-instruct-3

This model is a fine-tuned version of bofenghuang/vigogne-33b-instruct on the .CON french satirical corpus. It achieves the following results on the evaluation set:

• Loss: 1.8266

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: 5e-05
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 2
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
2.0831	0.24	30	1.9738
1.9472	0.48	60	1.8989
1.8874	0.73	90	1.8626
1.8311	0.97	120	1.8403
1.7394	1.21	150	1.8423
1.6894	1.45	180	1.8373
1.6351	1.69	210	1.8295
1.7245	1.94	240	1.8266

Framework versions

• Transformers 4.35.0
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.14.1