chchen/Llama-3.1-8B-Instruct-KTO-600
017
<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
Llama-3.1-8B-Instruct-KTO-600
This model is a fine-tuned version of meta-llama/Meta-Llama-3.1-8B-Instruct on the bctnoncotkto600 dataset. It achieves the following results on the evaluation set:
- Loss: 0.2017
- Rewards/chosen: 0.0412
- Logps/chosen: -18.3761
- Logits/chosen: -2496719.4921
- Rewards/rejected: -6.6216
- Logps/rejected: -86.0225
- Logits/rejected: -7772195.3684
- Rewards/margins: 6.6628
- Kl: 0.0
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-06
- trainbatchsize: 2
- evalbatchsize: 2
- seed: 42
- gradientaccumulationsteps: 8
- totaltrainbatch_size: 16
- optimizer: Use adamwtorch with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
- lrschedulertype: cosine
- lrschedulerwarmup_ratio: 0.1
- num_epochs: 10.0
Training results
Framework versions
- PEFT 0.12.0
- Transformers 4.46.1
- Pytorch 2.5.1+cu124
- Datasets 3.1.0
- Tokenizers 0.20.3
