learning_rate: 0.0001
train_batch_size: 8
eval_batch_size: 8
seed: 42
distributed_type: multi-GPU
num_devices: 2
gradient_accumulation_steps: 2
total_train_batch_size: 32
total_eval_batch_size: 16
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: cosine
lr_scheduler_warmup_ratio: 0.03
num_epochs: 3
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss
0.2573	0.2623	200	0.2500
0.2427	0.5246	400	0.2366
0.2298	0.7869	600	0.2261
0.1805	1.0492	800	0.2246
0.1834	1.3115	1000	0.2222
0.162	1.5738	1200	0.2188
0.1604	1.8361	1400	0.2148
0.1187	2.0984	1600	0.2219
0.1236	2.3607	1800	0.2252
0.1156	2.6230	2000	0.2258
0.1127	2.8852	2200	0.2253

marcuschill1823/incar-nlu-qwen1.5b-v12

incar-nlu-qwen1.5b-v12