| license: mit | |
| library_name: peft | |
| tags: | |
| - Text Generation | |
| - LLaMA2 | |
| - LLM | |
| datasets: | |
| - yahma/alpaca-cleaned | |
| pipeline_tag: text-generation | |
| base_model: meta-llama/Llama-2-7b-hf | |
| ## Quantized LLaMA2 | |
| **quantized-llama2-alpaca** is a fine-tuned version of the **LLaMA2** (Llama-2-7b-hf) model on the **Alpaca** dataset using **QLoRA**. | |
| ## 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 |