From 7fbc470195b1a4e672903298315cc5fb9bad857f Mon Sep 17 00:00:00 2001 From: wql Date: Mon, 26 Aug 2024 15:04:31 +0800 Subject: [PATCH] change: change for inference run --- batch_run.sh | 48 +++++++++++++++++++++++------------------------ src/gpu_status.py | 2 +- 2 files changed, 25 insertions(+), 25 deletions(-) diff --git a/batch_run.sh b/batch_run.sh index 3010aa20..e7976f00 100644 --- a/batch_run.sh +++ b/batch_run.sh @@ -27,32 +27,32 @@ # CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/lora_sft_2/Qwen-7B/Qwen_lora_sft_3_single.yaml | tee results/lora_sft_2/Qwen-7B/Qwen_lora_sft_3_single_log.txt -FORCE_TORCHRUN=1 llamafactory-cli train results/inference/Llama2-7B/llama2_predict_1.yaml | tee results/inference/Llama2-7B/llama2_predict_1_log.txt -FORCE_TORCHRUN=1 llamafactory-cli train results/inference/Llama2-7B/llama2_predict_2.yaml | tee results/inference/Llama2-7B/llama2_predict_2_log.txt -FORCE_TORCHRUN=1 llamafactory-cli train results/inference/Llama2-7B/llama2_predict_3.yaml | tee results/inference/Llama2-7B/llama2_predict_3_log.txt -CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/Llama2-7B/llama2_predict_1_single.yaml | tee results/inference/Llama2-7B/llama2_predict_1_single_log.txt -CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/Llama2-7B/llama2_predict_2_single.yaml | tee results/inference/Llama2-7B/llama2_predict_2_single_log.txt -CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/Llama2-7B/llama2_predict_3_single.yaml | tee results/inference/Llama2-7B/llama2_predict_3_single_log.txt +# FORCE_TORCHRUN=1 llamafactory-cli train results/inference/Llama2-7B/llama2_predict_1.yaml | tee results/inference/Llama2-7B/llama2_predict_1_log.txt +# FORCE_TORCHRUN=1 llamafactory-cli train results/inference/Llama2-7B/llama2_predict_2.yaml | tee results/inference/Llama2-7B/llama2_predict_2_log.txt +# FORCE_TORCHRUN=1 llamafactory-cli train results/inference/Llama2-7B/llama2_predict_3.yaml | tee results/inference/Llama2-7B/llama2_predict_3_log.txt +# CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/Llama2-7B/llama2_predict_1_single.yaml | tee results/inference/Llama2-7B/llama2_predict_1_single_log.txt +# CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/Llama2-7B/llama2_predict_2_single.yaml | tee results/inference/Llama2-7B/llama2_predict_2_single_log.txt +# CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/Llama2-7B/llama2_predict_3_single.yaml | tee results/inference/Llama2-7B/llama2_predict_3_single_log.txt -FORCE_TORCHRUN=1 llamafactory-cli train results/inference/Qwen-7B/Qwen_predict_1.yaml | tee results/inference/Qwen-7B/Qwen_predict_1_log.txt -FORCE_TORCHRUN=1 llamafactory-cli train results/inference/Qwen-7B/Qwen_predict_2.yaml | tee results/inference/Qwen-7B/Qwen_predict_2_log.txt -FORCE_TORCHRUN=1 llamafactory-cli train results/inference/Qwen-7B/Qwen_predict_3.yaml | tee results/inference/Qwen-7B/Qwen_predict_3_log.txt -CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/Qwen-7B/Qwen_predict_1_single.yaml | tee results/inference/Qwen-7B/Qwen_predict_1_single_log.txt -CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/Qwen-7B/Qwen_predict_2_single.yaml | tee results/inference/Qwen-7B/Qwen_predict_2_single_log.txt -CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/Qwen-7B/Qwen_predict_3_single.yaml | tee results/inference/Qwen-7B/Qwen_predict_3_single_log.txt +# FORCE_TORCHRUN=1 llamafactory-cli train results/inference/Qwen-7B/Qwen_predict_1.yaml | tee results/inference/Qwen-7B/Qwen_predict_1_log.txt +# FORCE_TORCHRUN=1 llamafactory-cli train results/inference/Qwen-7B/Qwen_predict_2.yaml | tee results/inference/Qwen-7B/Qwen_predict_2_log.txt +# FORCE_TORCHRUN=1 llamafactory-cli train results/inference/Qwen-7B/Qwen_predict_3.yaml | tee results/inference/Qwen-7B/Qwen_predict_3_log.txt +# CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/Qwen-7B/Qwen_predict_1_single.yaml | tee results/inference/Qwen-7B/Qwen_predict_1_single_log.txt +# CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/Qwen-7B/Qwen_predict_2_single.yaml | tee results/inference/Qwen-7B/Qwen_predict_2_single_log.txt +# CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/Qwen-7B/Qwen_predict_3_single.yaml | tee results/inference/Qwen-7B/Qwen_predict_3_single_log.txt -FORCE_TORCHRUN=1 llamafactory-cli train results/inference/ChatGLM2-6B/ChatGLM2_predict_1.yaml | tee results/inference/ChatGLM2-6B/ChatGLM2_predict_1_log.txt -FORCE_TORCHRUN=1 llamafactory-cli train results/inference/ChatGLM2-6B/ChatGLM2_predict_2.yaml | tee results/inference/ChatGLM2-6B/ChatGLM2_predict_2_log.txt -FORCE_TORCHRUN=1 llamafactory-cli train results/inference/ChatGLM2-6B/ChatGLM2_predict_3.yaml | tee results/inference/ChatGLM2-6B/ChatGLM2_predict_3_log.txt -CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/ChatGLM2-6B/ChatGLM2_predict_1_single.yaml | tee results/inference/ChatGLM2-6B/ChatGLM2_predict_1_single_log.txt -CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/ChatGLM2-6B/ChatGLM2_predict_2_single.yaml | tee results/inference/ChatGLM2-6B/ChatGLM2_predict_2_single_log.txt -CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/ChatGLM2-6B/ChatGLM2_predict_3_single.yaml | tee results/inference/ChatGLM2-6B/ChatGLM2_predict_3_single_log.txt +FORCE_TORCHRUN=1 CUDA_VISIBLE_DEVICES=0,1,4,5,6 llamafactory-cli train results/inference/ChatGLM2-6B/ChatGLM2_predict_1.yaml | tee results/inference/ChatGLM2-6B/ChatGLM2_predict_1_log.txt +FORCE_TORCHRUN=1 CUDA_VISIBLE_DEVICES=0,1,4,5,6 llamafactory-cli train results/inference/ChatGLM2-6B/ChatGLM2_predict_2.yaml | tee results/inference/ChatGLM2-6B/ChatGLM2_predict_2_log.txt +FORCE_TORCHRUN=1 CUDA_VISIBLE_DEVICES=0,1,4,5,6 llamafactory-cli train results/inference/ChatGLM2-6B/ChatGLM2_predict_3.yaml | tee results/inference/ChatGLM2-6B/ChatGLM2_predict_3_log.txt +CUDA_VISIBLE_DEVICES=6 llamafactory-cli train results/inference/ChatGLM2-6B/ChatGLM2_predict_1_single.yaml | tee results/inference/ChatGLM2-6B/ChatGLM2_predict_1_single_log.txt +CUDA_VISIBLE_DEVICES=6 llamafactory-cli train results/inference/ChatGLM2-6B/ChatGLM2_predict_2_single.yaml | tee results/inference/ChatGLM2-6B/ChatGLM2_predict_2_single_log.txt +CUDA_VISIBLE_DEVICES=6 llamafactory-cli train results/inference/ChatGLM2-6B/ChatGLM2_predict_3_single.yaml | tee results/inference/ChatGLM2-6B/ChatGLM2_predict_3_single_log.txt -FORCE_TORCHRUN=1 llamafactory-cli train results/inference/Baichuan2-7B/Baichuan2_predict_1.yaml | tee results/inference/Baichuan2-7B/Baichuan2_predict_1_log.txt -FORCE_TORCHRUN=1 llamafactory-cli train results/inference/Baichuan2-7B/Baichuan2_predict_2.yaml | tee results/inference/Baichuan2-7B/Baichuan2_predict_2_log.txt -FORCE_TORCHRUN=1 llamafactory-cli train results/inference/Baichuan2-7B/Baichuan2_predict_3.yaml | tee results/inference/Baichuan2-7B/Baichuan2_predict_3_log.txt -CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/Baichuan2-7B/Baichuan2_predict_1_single.yaml | tee results/inference/Baichuan2-7B/Baichuan2_predict_1_single_log.txt -CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/Baichuan2-7B/Baichuan2_predict_2_single.yaml | tee results/inference/Baichuan2-7B/Baichuan2_predict_2_single_log.txt -CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/Baichuan2-7B/Baichuan2_predict_3_single.yaml | tee results/inference/Baichuan2-7B/Baichuan2_predict_3_single_log.txt +# FORCE_TORCHRUN=1 llamafactory-cli train results/inference/Baichuan2-7B/Baichuan2_predict_1.yaml | tee results/inference/Baichuan2-7B/Baichuan2_predict_1_log.txt +# FORCE_TORCHRUN=1 llamafactory-cli train results/inference/Baichuan2-7B/Baichuan2_predict_2.yaml | tee results/inference/Baichuan2-7B/Baichuan2_predict_2_log.txt +# FORCE_TORCHRUN=1 llamafactory-cli train results/inference/Baichuan2-7B/Baichuan2_predict_3.yaml | tee results/inference/Baichuan2-7B/Baichuan2_predict_3_log.txt +# CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/Baichuan2-7B/Baichuan2_predict_1_single.yaml | tee results/inference/Baichuan2-7B/Baichuan2_predict_1_single_log.txt +# CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/Baichuan2-7B/Baichuan2_predict_2_single.yaml | tee results/inference/Baichuan2-7B/Baichuan2_predict_2_single_log.txt +# CUDA_VISIBLE_DEVICES=0 llamafactory-cli train results/inference/Baichuan2-7B/Baichuan2_predict_3_single.yaml | tee results/inference/Baichuan2-7B/Baichuan2_predict_3_single_log.txt diff --git a/src/gpu_status.py b/src/gpu_status.py index 82daac46..eb7ad2a8 100644 --- a/src/gpu_status.py +++ b/src/gpu_status.py @@ -44,7 +44,7 @@ def main(): all_processes_status = all_processes_status ) - with open("./results/gpu_status_0826.json", "a", encoding="utf-8") as f: + with open("./results/gpu_status_0826_2.json", "a", encoding="utf-8") as f: f.write(json.dumps(logs) + "\n") print(logs)