tian-ft4bit

This model is a fine-tuned version of openthaigpt/openthaigpt-1.0.0-7b-chat on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2898

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: 0.0002
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 2
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
2.8288	0.9524	10	1.1157
0.7351	2.0	21	0.3817
0.3298	2.9524	31	0.2716
0.1567	4.0	42	0.2654
0.091	4.9524	52	0.2830
0.0572	6.0	63	0.2835
0.0542	6.9524	73	0.2912
0.0448	8.0	84	0.2886
0.0464	8.9524	94	0.2890
0.0389	9.5238	100	0.2898

Framework versions

• PEFT 0.10.0
• Transformers 4.40.0
• Pytorch 2.2.2+cu121
• Datasets 2.19.0
• Tokenizers 0.19.1