Edit model card

Visualize in Weights & Biases

Meta-Llama-3-8B-Base-5e-7

This model is a fine-tuned version of princeton-nlp/Llama-3-Base-8B-SFT on the princeton-nlp/llama3-ultrafeedback dataset. It achieves the following results on the evaluation set:

  • Loss: 1.6913
  • Rewards/chosen: -0.7344
  • Rewards/rejected: -0.8094
  • Rewards/accuracies: 0.5610
  • Rewards/margins: 0.0751
  • Logps/rejected: -0.8094
  • Logps/chosen: -0.7344
  • Logits/rejected: -0.5793
  • Logits/chosen: -0.5922

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: 5e-07
  • train_batch_size: 2
  • eval_batch_size: 4
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 4
  • gradient_accumulation_steps: 16
  • total_train_batch_size: 128
  • total_eval_batch_size: 16
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 1

Training results

Training Loss Epoch Step Validation Loss Rewards/chosen Rewards/rejected Rewards/accuracies Rewards/margins Logps/rejected Logps/chosen Logits/rejected Logits/chosen
1.6906 0.8550 400 1.6913 -0.7344 -0.8094 0.5610 0.0751 -0.8094 -0.7344 -0.5793 -0.5922

Framework versions

  • Transformers 4.42.0
  • Pytorch 2.1.2+cu121
  • Datasets 2.14.6
  • Tokenizers 0.19.1
Downloads last month
0
Safetensors
Model size
8.03B params
Tensor type
BF16
·
Inference API
Unable to determine this model's library. Check the docs .

Model tree for TTTXXX01/Meta-Llama-3-8B-Base-5e-7

Finetuned
(16)
this model

Dataset used to train TTTXXX01/Meta-Llama-3-8B-Base-5e-7