llama-3.1-70B-Instruct-rehearsal_playpen_SFT-e3_DABL02_0.82K-steps

This model is a fine-tuned version of meta-llama/Llama-3.1-70B-Instruct on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2147

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: 8
• seed: 7331
• optimizer: Use adamw_8bit with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.03
• lr_scheduler_warmup_steps: 5
• training_steps: 820

Training results

Training Loss	Epoch	Step	Validation Loss
0.5937	0.0499	100	0.3638
0.4825	0.0998	200	0.2560
0.3826	0.1497	300	0.2613
0.3414	0.1996	400	0.2825
0.3106	0.2495	500	0.2232
0.2821	0.2994	600	0.2173
0.2739	0.3493	700	0.2167
0.2221	0.3992	800	0.2147

Framework versions

• PEFT 0.14.0
• Transformers 4.47.1
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.21.0