yusuke111/myBit-Llama2-jp-127M-2B4TLike-102-1024-asc

myBit-Llama2-jp-127M-2B4TLike

This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.7946

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: 0.0024
• train_batch_size: 12
• eval_batch_size: 12
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 96
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.95) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 750
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
5.5592	0.0555	500	4.6971
3.9123	0.1111	1000	4.2204
3.5329	0.1666	1500	3.8957
3.3627	0.2222	2000	3.7229
3.2594	0.2777	2500	3.5721
3.1944	0.3333	3000	3.3995
3.1391	0.3888	3500	3.2851
3.1046	0.4443	4000	3.1923
3.0648	0.4999	4500	3.1243
3.0325	0.5554	5000	3.0711
2.9877	0.6110	5500	3.0184
2.9752	0.6665	6000	2.9568
2.9901	0.7221	6500	2.9294
2.9769	0.7776	7000	2.9101
2.9432	0.8331	7500	2.8855
2.9199	0.8887	8000	2.8612
2.8869	0.9442	8500	2.8358
2.8574	0.9998	9000	2.7946

Framework versions

• Transformers 4.47.1
• Pytorch 2.6.0+cu124
• Datasets 3.5.1
• Tokenizers 0.21.1