metadata

library_name: peft
license: apache-2.0
base_model: Qwen/Qwen2.5-1.5B-Instruct
tags:
  - generated_from_trainer
model-index:
  - name: qwen_checkpoints
    results: []

qwen_checkpoints

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

Loss: 0.0618
Mse: 0.0618
Mae: 0.1983
R Squared: 0.3107

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: 0.0001
train_batch_size: 64
eval_batch_size: 64
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_ratio: 0.01
num_epochs: 3
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Mae	Mse	R Squared
0.0856	0.1558	100	0.0878	0.2351	0.0878	0.0207
0.0843	0.3115	200	0.0803	0.2314	0.0803	0.1045
0.0851	0.4673	300	0.0882	0.2278	0.0882	0.0168
0.0676	0.6231	400	0.0716	0.2183	0.0716	0.2014
0.0737	0.7788	500	0.0691	0.2164	0.0691	0.2291
0.0694	0.9346	600	0.0696	0.2157	0.0696	0.2242
0.0569	1.0903	700	0.0661	0.2049	0.0661	0.2627
0.0589	1.2461	800	0.0663	0.2045	0.0663	0.2606
0.0648	1.4019	900	0.0649	0.2039	0.0649	0.2764
0.0652	1.5576	1000	0.0644	0.2027	0.0644	0.2813
0.0657	1.7134	1100	0.0649	0.0649	0.2082	0.2763
0.0577	1.8692	1200	0.0639	0.0639	0.2022	0.2869
0.0564	2.0249	1300	0.0636	0.0636	0.2006	0.2902
0.0613	2.1807	1400	0.0633	0.0633	0.1989	0.2939
0.0596	2.3364	1500	0.0624	0.0624	0.1999	0.3036
0.0547	2.4922	1600	0.0621	0.0621	0.1985	0.3076
0.0554	2.6480	1700	0.0620	0.0620	0.1974	0.3087
0.0581	2.8037	1800	0.0618	0.0618	0.1983	0.3107
0.0653	2.9595	1900	0.0618	0.0618	0.1983	0.3107

Framework versions

PEFT 0.13.2
Transformers 4.45.2
Pytorch 2.5.1+cu121
Datasets 3.1.0
Tokenizers 0.20.3