This model is a fine-tuned version of /gemini/space/pretrained_models/Qwen3-8B on the web_sft_0115 dataset. It achieves the following results on the evaluation set:

Loss: 0.2433

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: 1e-05
train_batch_size: 1
eval_batch_size: 1
seed: 42
distributed_type: multi-GPU
num_devices: 4
gradient_accumulation_steps: 4
total_train_batch_size: 16
total_eval_batch_size: 4
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_ratio: 0.1
num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss
0.5327	0.1013	50	0.5031
0.4048	0.2026	100	0.4044
0.3865	0.3040	150	0.3753
0.3676	0.4053	200	0.3531
0.3331	0.5066	250	0.3346
0.3249	0.6079	300	0.3211
0.3115	0.7092	350	0.3067
0.3037	0.8105	400	0.2933
0.2856	0.9119	450	0.2809
0.2305	1.0122	500	0.2770
0.1847	1.1135	550	0.2729
0.1671	1.2148	600	0.2706
0.168	1.3161	650	0.2625
0.1665	1.4174	700	0.2583
0.1607	1.5187	750	0.2528
0.1578	1.6201	800	0.2481
0.1652	1.7214	850	0.2454
0.1624	1.8227	900	0.2441
0.1436	1.9240	950	0.2435

Framework versions

Transformers 4.57.1
Pytorch 2.9.1+cu128
Datasets 4.0.0
Tokenizers 0.22.2

Rain-air/Qwen3-8B-web-sft_0115

sft

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Rain-air/
Qwen3-8B-web-sft_0115