dfm_indirect_speech

This model is a fine-tuned version of KennethEnevoldsen/dfm-sentence-encoder-large-exp2-no-lang-align on an unknown dataset. It achieves the following results on the evaluation set:

• Accuracy: 0.9186
• Precision: 0.9173
• Recall: 0.9186
• F1: 0.9148
• Loss: 0.7077

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: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• 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: linear
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Accuracy	Precision	Recall	F1	Validation Loss
No log	1.0	9	0.5262	0.7356	0.5262	0.4663	0.9806
No log	2.0	18	0.8976	0.8921	0.8976	0.8896	0.3764
No log	3.0	27	0.9241	0.9190	0.9241	0.9210	0.3074
No log	4.0	36	0.9060	0.9066	0.9060	0.9028	0.4696
No log	5.0	45	0.9088	0.9093	0.9088	0.9052	0.5214
No log	6.0	54	0.9132	0.9099	0.9132	0.9093	0.5031
No log	7.0	63	0.9060	0.9090	0.9060	0.9026	0.7231
No log	8.0	72	0.9103	0.9167	0.9103	0.9075	0.6638
No log	9.0	81	0.9125	0.9137	0.9125	0.9088	0.7139
No log	10.0	90	0.9194	0.9169	0.9194	0.9156	0.5924
No log	11.0	99	0.9183	0.9158	0.9183	0.9144	0.6261
No log	12.0	108	0.9157	0.9157	0.9157	0.9120	0.6921
No log	13.0	117	0.9185	0.9176	0.9185	0.9148	0.6814
No log	14.0	126	0.9174	0.9148	0.9174	0.9135	0.6498
No log	15.0	135	0.9181	0.9231	0.9181	0.9145	0.6699
No log	16.0	144	0.9191	0.9240	0.9191	0.9155	0.6835
No log	17.0	153	0.9184	0.9170	0.9184	0.9147	0.7015
No log	17.8235	160	0.9186	0.9173	0.9186	0.9148	0.7077

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu124
• Tokenizers 0.21.0