Toys_and_Games-sft-dsz2

This model is a fine-tuned version of Qwen/Qwen2.5-1.5B-Instruct on the Toys_and_Games_train dataset. It achieves the following results on the evaluation set:

Loss: 1.6175

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: 0.0003
train_batch_size: 32
eval_batch_size: 32
seed: 42
distributed_type: multi-GPU
num_devices: 8
gradient_accumulation_steps: 8
total_train_batch_size: 2048
total_eval_batch_size: 256
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: 20
num_epochs: 10.0

Training results

Training Loss	Epoch	Step	Validation Loss
2.6194	0.5039	16	3.7493
1.7278	1.0	32	4.8666
1.3161	1.5039	48	3.0621
1.2497	2.0	64	2.8294
1.0487	2.5039	80	2.5177
0.9401	3.0	96	2.0956
0.7792	3.5039	112	1.8423
0.7605	4.0	128	1.6965
0.7075	4.5039	144	1.6534
0.6888	5.0	160	1.6175
0.6084	5.5039	176	1.6297
0.6101	6.0	192	1.6204
0.5102	6.5039	208	1.7053