learning_rate: 0.0001
train_batch_size: 64
eval_batch_size: 64
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 0.1
num_epochs: 3.0

Training Loss	Epoch	Step	Validation Loss
0.8306	0.2075	50	0.7977
0.7036	0.4149	100	0.6887
0.6795	0.6224	150	0.6632
0.6577	0.8299	200	0.6394
0.5876	1.0373	250	0.6330
0.5877	1.2448	300	0.6205
0.5682	1.4523	350	0.6216
0.5731	1.6598	400	0.6096
0.5695	1.8672	450	0.6046
0.4903	2.0747	500	0.6094
0.4938	2.2822	550	0.6101
0.4832	2.4896	600	0.6070
0.4890	2.6971	650	0.6056
0.5012	2.9046	700	0.6048

alanwang2001/qwen3-1.7B-sentiment-lora

sft