llama-output
This model is a fine-tuned version of meta-llama/Llama-2-7b-chat-hf on the samsum dataset.
Model description
The model is a fine-tuned version of Llama-2-7b-chat-hf using int8 quantization and LoRA. By using this configuration, approximately 6% of parameters are trainable.
Intended uses & limitations
It is intended to improve the summarisation capacities of Llama 2 7B on dialogs, expecting to produce a concise brief. As it is trained on the dataset SamSum, the terms of use are limited to those of the non-commercial licence: CC BY-NC-ND 4.0
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: 1
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 2
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 1
Training results
Framework versions
- Transformers 4.33.1
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.13.3
Model tree for selfmaker/llama2-7B-xsum
Base model
meta-llama/Llama-2-7b-chat-hf