Dialogue_dfm

This model is a fine-tuned version of KennethEnevoldsen/dfm-sentence-encoder-large-exp2-no-lang-align on the None dataset. It achieves the following results on the evaluation set:

• Accuracy: 0.9417
• Precision: 0.9468
• Recall: 0.9417
• F1: 0.9418
• Loss: 0.4894

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• 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	0.9412	8	0.7223	0.7770	0.7223	0.7069	0.8079
No log	2.0	17	0.7821	0.8280	0.7821	0.7670	0.7157
No log	2.9412	25	0.9217	0.9243	0.9217	0.9174	0.3617
No log	4.0	34	0.9283	0.9331	0.9283	0.9272	0.3444
No log	4.9412	42	0.9156	0.9274	0.9156	0.9168	0.4618
No log	6.0	51	0.9271	0.9316	0.9271	0.9277	0.3164
No log	6.9412	59	0.9356	0.9387	0.9356	0.9349	0.3228
No log	8.0	68	0.9329	0.9398	0.9329	0.9334	0.4814
No log	8.9412	76	0.9402	0.9450	0.9402	0.9400	0.4819
No log	10.0	85	0.9409	0.9459	0.9409	0.9409	0.4952
No log	10.9412	93	0.9367	0.9428	0.9367	0.9370	0.5182
No log	12.0	102	0.9409	0.9462	0.9409	0.9411	0.4947
No log	12.9412	110	0.9405	0.9457	0.9405	0.9406	0.4927
No log	14.0	119	0.9409	0.9462	0.9409	0.9411	0.4912
No log	14.9412	127	0.9413	0.9465	0.9413	0.9414	0.4917
No log	16.0	136	0.9413	0.9464	0.9413	0.9415	0.4893
No log	16.9412	144	0.9413	0.9464	0.9413	0.9415	0.4890
No log	18.0	153	0.9417	0.9468	0.9417	0.9418	0.4893
No log	18.8235	160	0.9417	0.9468	0.9417	0.9418	0.4894

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.2
• Tokenizers 0.19.1