llama-3-8b-base-sft-ultrachat-8xh200/README.md

library_name, base_model, tags, datasets, model-index

library_name	base_model	tags	datasets	model-index
transformers	meta-llama/Meta-Llama-3-8B	alignment-handbook generated_from_trainer	HuggingFaceH4/ultrachat_200k	name results llama-3-8b-base-sft-ultrachat-8xh200-20260410-113950

llama-3-8b-base-sft-ultrachat-8xh200-20260410-113950

This model is a fine-tuned version of /scratch/feng.yulu/dynamic-dpo-v4/base_models/Meta-Llama-3-8B on the HuggingFaceH4/ultrachat_200k dataset. It achieves the following results on the evaluation set:

• Loss: 1.0705

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• total_train_batch_size: 128
• total_eval_batch_size: 128
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
1.1453	0.2137	200	1.1343
1.0929	0.4274	400	1.1096
1.0808	0.6410	600	1.0848
1.0529	0.8547	800	1.0705

Framework versions

• Transformers 4.51.0
• Pytorch 2.3.1+cu121
• Datasets 2.21.0
• Tokenizers 0.21.4