v1.0.2

README.md

@@ -19,7 +19,7 @@ llama2算法主要将转换成向量的分词用qkv自相关和全连接层提
 
 ## 环境配置
 ```
-mv TinyLlama_pytorch TinyLlama # 去框架名后缀
+mv tinyllama_pytorch TinyLlama # 去框架名后缀
 ```
 
 ### Docker（方法一）
@@ -27,7 +27,7 @@ mv TinyLlama_pytorch TinyLlama # 去框架名后缀
 docker pull image.sourcefind.cn:5000/dcu/admin/base/pytorch:2.1.0-centos7.6-dtk23.10-py38
 # <your IMAGE ID>为以上拉取的docker的镜像ID替换，本镜像为：ffa1f63239fc
 docker run -it --shm-size=32G -v $PWD/TinyLlama:/home/TinyLlama -v /opt/hyhal:/opt/hyhal --privileged=true --device=/dev/kfd --device=/dev/dri/ --group-add video --name tinyllama <your IMAGE ID> bash
-cd TinyLlama
+cd /home/TinyLlama
 pip install -r requirements.txt
 ```
 ### Dockerfile（方法二）

sft/script.sh

@@ -4,7 +4,7 @@
 # We hope the community can explore on finetuning TinyLlama and come up with better chat models. I will include community-finetuned models in this repo.
 
 # V0.1
-# CUDA_VISIBLE_DEVICES=0 accelerate launch --main_process_port 1234 sft/finetune.py \
+# HIP_VISIBLE_DEVICES=0 accelerate launch --main_process_port 1234 sft/finetune.py \
 CUDA_VISIBLE_DEVICES=0,1,2,3 accelerate launch --multi_gpu --num_processes 4 --main_process_port 1234 sft/finetune.py \
     --model_name_or_path PY007/TinyLlama-1.1B-intermediate-step-240k-503b \
     --output_dir ./output/503B_FT_lr1e-5_ep5 \