ruadapt_qwen2.5_1.5B_ext_u48_mean_init_libre_instruct

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 3.5833

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: 5e-05
• train_batch_size: 1
• eval_batch_size: 1
• seed: 1337
• distributed_type: multi-GPU
• num_devices: 16
• gradient_accumulation_steps: 128
• total_train_batch_size: 2048
• total_eval_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 16
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss
4.3816	0.9399	16	4.2962
3.5399	1.8798	32	3.5833

Framework versions

• PEFT 0.9.0
• Transformers 4.45.2
• Pytorch 2.3.0a0+6ddf5cf85e.nv24.04
• Datasets 2.18.0
• Tokenizers 0.20.3