llama-2-7b-chat-hf-gsm8k-bs32-sharegpt

This model is a fine-tuned version of meta-llama/Llama-2-7b-chat-hf on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.7306

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: 0.0001
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• total_train_batch_size: 32
• total_eval_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.03
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss
0.8458	0.4274	100	0.8449
0.8417	0.8547	200	0.7988
0.8234	1.2821	300	0.7797
0.7774	1.7094	400	0.7670
0.72	2.1368	500	0.7575
0.7188	2.5641	600	0.7498
0.8102	2.9915	700	0.7438
0.7295	3.4188	800	0.7392
0.7591	3.8462	900	0.7351
0.7615	4.2735	1000	0.7323
0.7084	4.7009	1100	0.7306

Framework versions

• PEFT 0.13.2
• Transformers 4.45.2
• Pytorch 2.4.1+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3