JunxiongWang/Llama3.1-Mamba-8B-distill

Llama-Mamba-3.1-8B-teacher-Llama-3.1-70B-Instruct-kl1.0-ce0.0

This model is a fine-tuned version of meta-llama/Llama-3.1-8B-Instruct on the JunxiongWang/sftdatasetv3 dataset. It achieves the following results on the evaluation set:

• Loss: 242.9601

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: 2
• eval_batch_size: 4
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• total_eval_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.01
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
220.5766	1.0000	51995	242.9601

Framework versions

• Transformers 4.43.1
• Pytorch 2.1.1+cu118
• Datasets 2.20.0
• Tokenizers 0.19.1