learning_rate: 2e-05
train_batch_size: 16
eval_batch_size: 32
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 32
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 300
num_epochs: 6
label_smoothing_factor: 0.1

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.5573	0.6667	250	0.5038	0.8773	0.8781	0.8848	0.8773
0.4445	1.3333	500	0.4608	0.9023	0.9029	0.9091	0.9023
0.4207	2.0	750	0.4220	0.9327	0.9332	0.9344	0.9327
0.3699	2.6667	1000	0.4129	0.937	0.9370	0.9383	0.937
0.3446	3.3333	1250	0.4416	0.9367	0.9368	0.9386	0.9367
0.3294	4.0	1500	0.4152	0.9453	0.9453	0.9452	0.9453
0.3118	4.6667	1750	0.4242	0.947	0.9472	0.9477	0.947
0.3044	5.3333	2000	0.4225	0.9467	0.9468	0.9470	0.9467
0.3007	6.0	2250	0.4223	0.9467	0.9468	0.9470	0.9467

Framework versions

Transformers 4.51.3
Pytorch 2.7.0+cu126
Datasets 3.5.1
Tokenizers 0.21.1

Training info:

TrainOutput(global_step=2250, training_loss=0.39833289252387155, metrics={'train_runtime': 2934.7433, 'train_samples_per_second': 24.534, 'train_steps_per_second': 0.767, 'total_flos': 1.8944166076416e+16, 'train_loss': 0.39833289252387155, 'epoch': 6.0})

Ghazi-nak/7_BiasDetctor_HTW_ANG.KI

7_BiasDetctor_HTW_ANG.KI

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Training info:

Ghazi-nak/
7_BiasDetctor_HTW_ANG.KI