2024-03-04 19:16:35 +00:00
|
|
|
#!/bin/bash
|
2024-04-04 06:48:21 +00:00
|
|
|
# DO NOT use quantized model or quantization_bit when merging lora weights
|
2024-03-04 19:16:35 +00:00
|
|
|
|
2024-04-15 14:14:34 +00:00
|
|
|
CUDA_VISIBLE_DEVICES=0 python ../../src/export_model.py \
|
2024-03-04 19:16:35 +00:00
|
|
|
--model_name_or_path meta-llama/Llama-2-7b-hf \
|
|
|
|
--adapter_name_or_path ../../saves/LLaMA2-7B/lora/sft \
|
|
|
|
--template default \
|
|
|
|
--finetuning_type lora \
|
|
|
|
--export_dir ../../models/llama2-7b-sft \
|
|
|
|
--export_size 2 \
|
|
|
|
--export_legacy_format False
|