|
--- |
|
library_name: peft |
|
license: mit |
|
datasets: |
|
- yahma/alpaca-cleaned |
|
pipeline_tag: text-generation |
|
--- |
|
|
|
## Quantized LLaMA2 Alpaca |
|
**quantized-llama2-alpaca** is a fine-tuned version of the [LLaMA2](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/) ([Llama-2-7b-hf](https://huggingface.co/meta-llama/Llama-2-7b)) model on the [Alpaca](https://github.com/tatsu-lab/stanford_alpaca) dataset using [QLoRA](https://arxiv.org/abs/2305.14314). |
|
|
|
## Training procedure |
|
|
|
|
|
The following `bitsandbytes` quantization config was used during training: |
|
- load_in_8bit: False |
|
- load_in_4bit: True |
|
- llm_int8_threshold: 6.0 |
|
- llm_int8_skip_modules: None |
|
- llm_int8_enable_fp32_cpu_offload: False |
|
- llm_int8_has_fp16_weight: False |
|
- bnb_4bit_quant_type: nf4 |
|
- bnb_4bit_use_double_quant: True |
|
- bnb_4bit_compute_dtype: bfloat16 |
|
|
|
### Framework versions |
|
|
|
|
|
- PEFT 0.4.0 |