Fix argument 1220 (#242)

modelscope · Dec 20, 2023 · 852f341 · 852f341
1 parent 268b07a
commit 852f341
Show file tree

Hide file tree

Showing 112 changed files with 200 additions and 254 deletions.
diff --git a/docs/source/LLM/LLM微调文档.md b/docs/source/LLM/LLM微调文档.md
@@ -284,21 +284,27 @@ print(f'history: {history}')
 使用**数据集**评估:
 ```bash
 # 直接推理
-CUDA_VISIBLE_DEVICES=0 swift infer --ckpt_dir 'xxx/vx_xxx/checkpoint-xxx'
+CUDA_VISIBLE_DEVICES=0 \
+swift infer \
+ --ckpt_dir 'xxx/vx_xxx/checkpoint-xxx' \
+ --load_dataset_config true \
 
 # Merge LoRA增量权重并推理
 swift merge-lora --ckpt_dir 'xxx/vx_xxx/checkpoint-xxx'
-CUDA_VISIBLE_DEVICES=0 swift infer --ckpt_dir 'xxx/vx_xxx/checkpoint-xxx-merged'
+CUDA_VISIBLE_DEVICES=0 \
+swift infer \
+ --ckpt_dir 'xxx/vx_xxx/checkpoint-xxx-merged' \
+ --load_dataset_config true \
 ```
 
 **人工**评估:
 ```bash
 # 直接推理
-CUDA_VISIBLE_DEVICES=0 swift infer --ckpt_dir 'xxx/vx_xxx/checkpoint-xxx' --eval_human true
+CUDA_VISIBLE_DEVICES=0 swift infer --ckpt_dir 'xxx/vx_xxx/checkpoint-xxx'
 
 # Merge LoRA增量权重并推理
 swift merge-lora --ckpt_dir 'xxx/vx_xxx/checkpoint-xxx'
-CUDA_VISIBLE_DEVICES=0 swift infer --ckpt_dir 'xxx/vx_xxx/checkpoint-xxx-merged' --eval_human true
+CUDA_VISIBLE_DEVICES=0 swift infer --ckpt_dir 'xxx/vx_xxx/checkpoint-xxx-merged'
 ```
 
 ## Web-UI

diff --git a/docs/source/LLM/VLLM推理加速与部署.md b/docs/source/LLM/VLLM推理加速与部署.md
@@ -204,13 +204,17 @@ print(f"history: {resp['history']}")
 swift merge-lora --ckpt_dir 'xxx/vx_xxx/checkpoint-xxx'
 
 # 使用数据集评估
-CUDA_VISIBLE_DEVICES=0 swift infer --ckpt_dir 'xxx/vx_xxx/checkpoint-xxx-merged' --infer_backend vllm
+CUDA_VISIBLE_DEVICES=0 \
+swift infer \
+ --ckpt_dir 'xxx/vx_xxx/checkpoint-xxx-merged' \
+ --infer_backend vllm \
+ --load_dataset_config true \
+
 # 人工评估
 CUDA_VISIBLE_DEVICES=0 \
 swift infer \
  --ckpt_dir 'xxx/vx_xxx/checkpoint-xxx-merged' \
  --infer_backend vllm \
- --eval_human true \
 ```
 
 ## Web-UI加速

diff --git a/docs/source/LLM/命令行参数.md b/docs/source/LLM/命令行参数.md
@@ -18,7 +18,7 @@
 - `--seed`: 全局的seed, 默认使用`42`. 用于复现训练效果.
 - `--resume_from_checkpoint`: 用于断点续训, 默认为`None`. 你可以将其设置为checkpoint的路径, 例如: `'output/qwen-7b-chat/vx_xxx/checkpoint-xxx'`, 来进行断点续训.
 - `--dtype`: 基模型载入时的torch_dtype, 默认为`'AUTO'`, 即智能选择dtype: 如果机器不支持bf16, 则使用fp16, 如果`MODEL_MAPPING`中对应模型有指定torch_dtype, 则使用其对应dtype, 否则使用bf16. 你可以选择的值包括: 'bf16', 'fp16', 'fp32'.
-- `--dataset`: 用于选择训练的数据集, 默认为`'blossom-math-zh'`. 可以选择的数据集可以查看`DATASET_MAPPING.keys()`. 如果需要使用多个数据集进行训练, 你可以使用','或者' '进行分割, 例如: `alpaca-en,alpaca-zh` or `alpaca-en alpaca-zh`.
+- `--dataset`: 用于选择训练的数据集, 默认为`None`. 可以选择的数据集可以查看`DATASET_MAPPING.keys()`. 如果需要使用多个数据集进行训练, 你可以使用','或者' '进行分割, 例如: `alpaca-en,alpaca-zh` or `alpaca-en alpaca-zh`.
 - `--dataset_seed`: 用于指定数据集处理的seed, 默认为`42`. 以random_state形式存在, 不影响全局seed.
 - `--dataset_test_ratio`: 用于指定子数据集切分成训练集和验证集的比例, 默认为`0.01`. 如果子数据集已经进行了训练集和验证集的切分, 则此参数无效.
 - `--train_dataset_sample`: 对训练集进行采样, 默认是`20000`, 用于加快训练的速度. 该参数是为了避免数据集过大, 单个epoch训练时间过长的问题. LoRA的收敛通常较快, 不需要很多数据样本的微调. 如果你指定为`-1`, 则使用完整的训练集进行训练, 该情况一般出现在全参数微调的设置下.
@@ -93,15 +93,15 @@
 - `--template_type`: 默认值为`'AUTO'`, 具体的参数介绍可以在`sft.sh命令行参数`中查看.
 - `--infer_backend`: 你可以选择'AUTO', 'vllm', 'pt'. 默认使用'AUTO', 进行智能选择, 即如果没有传入`ckpt_dir`或使用全参数微调, 并且安装了vllm且模型支持vllm则使用vllm引擎, 否则使用原生torch进行推理. vllm环境准备可以参考[VLLM推理加速与部署](./VLLM推理加速与部署.md#环境准备), vllm支持的模型可以查看[支持的模型](./支持的模型和数据集.md#模型).
 - `--ckpt_dir`: 必填项, 值为SFT阶段保存的checkpoint路径, e.g. `'/path/to/your/vx_xxx/checkpoint-xxx'`.
-- `--load_args_from_ckpt_dir`: 是否从`ckpt_dir`的`sft_args.json`文件中读取配置信息. 默认是`True`.
-- `--load_dataset_config`: 该参数只有在`--load_args_from_ckpt_dir true`时才生效. 即是否从`ckpt_dir`的`sft_args.json`文件中读取数据集相关的配置信息. 默认为`True`.
-- `--eval_human`: 使用数据集中的验证集部分进行评估还是使用人工的方式评估. 默认值为`None`, 如有传入数据集, 则设置为True, 否则设置为False.
+- `--load_args_from_ckpt_dir`: 是否从`ckpt_dir`的`sft_args.json`文件中读取模型配置信息. 默认是`True`.
+- `--load_dataset_config`: 该参数只有在`--load_args_from_ckpt_dir true`时才生效. 即是否从`ckpt_dir`的`sft_args.json`文件中读取数据集相关的配置信息. 默认为`False`.
+- `--eval_human`: 使用数据集中的验证集部分进行评估还是使用人工的方式评估. 默认值为`None`, 进行智能选择, 如果没有任何数据集(含自定义数据集)传入, 则会使用人工评估的方式. 如果有数据集传入, 则会使用数据集方式评估.
 - `--seed`: 默认值为`42`, 具体的参数介绍可以在`sft.sh命令行参数`中查看.
 - `--dtype`: 默认值为`'AUTO`, 具体的参数介绍可以在`sft.sh命令行参数`中查看.
-- `--dataset`: 默认值为`'blossom-math-zh'`, 具体的参数介绍可以在`sft.sh命令行参数`中查看. 该参数在`eval_human`设置为True时不生效.
-- `--dataset_seed`: 默认值为`42`, 具体的参数介绍可以在`sft.sh命令行参数`中查看. 该参数在`eval_human`设置为True时不生效.
-- `--dataset_test_ratio`: 默认值为`0.01`, 具体的参数介绍可以在`sft.sh命令行参数`中查看. 该参数在`eval_human`设置为True时不生效.
-- `--val_dataset_sample`: 表示想要评估和展示的验证集的数量, 默认值为`10`. 该参数在`eval_human`设置为True时不生效.
+- `--dataset`: 默认值为`None`, 具体的参数介绍可以在`sft.sh命令行参数`中查看.
+- `--dataset_seed`: 默认值为`42`, 具体的参数介绍可以在`sft.sh命令行参数`中查看.
+- `--dataset_test_ratio`: 默认值为`0.01`, 具体的参数介绍可以在`sft.sh命令行参数`中查看.
+- `--val_dataset_sample`: 表示想要评估和展示的验证集的数量, 默认值为`10`.
 - `--system`: 默认值为`None`. 具体的参数介绍可以在`sft.sh命令行参数`中查看.
 - `--max_length`: 默认值为`2048`. 具体的参数介绍可以在`sft.sh命令行参数`中查看.
 - `--truncation_strategy`: 默认是`'delete'`. 具体的参数介绍可以在`sft.sh命令行参数`中查看.
@@ -120,11 +120,11 @@
 - `--repetition_penalty`: 默认值为`1.05`.
 - `--use_flash_attn`: 默认值为`None`, 即为'auto'. 具体的参数介绍可以在`sft.sh命令行参数`中查看.
 - `--ignore_args_error`: 默认值为`False`, 具体的参数介绍可以在`sft.sh命令行参数`中查看.
-- `--stream`: 是否使用流式输出, 默认为`True`.
+- `--stream`: 是否使用流式输出, 默认为`True`. 该参数只有在使用数据集评估并且verbose为True时才生效.
 - `--merge_lora_and_save`: 是否将lora权重merge到基模型中, 并保存完整的权重, 默认为`False`. 权重会保存在`ckpt_dir`的同级目录中, e.g. `'/path/to/your/vx_xxx/checkpoint-xxx-merged'`目录下.
-- `--safe_serialization`: 保存成`safetensors`文件还是`bin`文件. 默认为True, 即保存为safetensors文件.
+- `--save_safetensors`: 保存成`safetensors`文件还是`bin`文件. 默认为`True`.
 - `--overwrite_generation_config`: 是否将评估所使用的generation_config保存成`generation_config.json`文件, 默认为`False`. 训练时保存的generation_config文件将被覆盖.
-- `--verbose`: 如果设置为False, 则使用tqdm样式推理. 如果设置为True, 则输出推理的query, response, label. 默认为`None`, 进行自动选择, 即`len(val_dataset) >= 100`时, 设置为False, 否则设置为True. 该参数只有在`--eval_human false`时才生效.
+- `--verbose`: 如果设置为False, 则使用tqdm样式推理. 如果设置为True, 则输出推理的query, response, label. 默认为`None`, 进行自动选择, 即`len(val_dataset) >= 100`时, 设置为False, 否则设置为True. 该参数只有在使用数据集评估时生效.
 - `--share`: 传递给gradio的`demo.queue().launch(...)`函数. 该参数只有在使用`app-ui`时才生效.
 - `--gpu_memory_utilization`: 初始化vllm引擎`EngineArgs`的参数, 默认为`0.9`. 该参数只有在使用vllm时才生效.
 - `--tensor_parallel_size`: 初始化vllm引擎`EngineArgs`的参数, 默认为`1`. 该参数只有在使用vllm时才生效.
diff --git a/docs/source/LLM/自我认知微调最佳实践.md b/docs/source/LLM/自我认知微调最佳实践.md
@@ -247,11 +247,11 @@ My name is Xiao Huang, developed by ModelScope. I am an artificial intelligence
 使用CLI:
 ```bash
 # 直接推理
-CUDA_VISIBLE_DEVICES=0 swift infer --ckpt_dir 'qwen-7b-chat/vx-xxx/checkpoint-xxx' --eval_human true
+CUDA_VISIBLE_DEVICES=0 swift infer --ckpt_dir 'qwen-7b-chat/vx-xxx/checkpoint-xxx'
 
 # Merge LoRA增量权重并推理
 swift merge-lora --ckpt_dir 'xxx/vx_xxx/checkpoint-xxx'
-CUDA_VISIBLE_DEVICES=0 swift infer --ckpt_dir 'qwen-7b-chat/vx-xxx/checkpoint-xxx-merged' --eval_human true
+CUDA_VISIBLE_DEVICES=0 swift infer --ckpt_dir 'qwen-7b-chat/vx-xxx/checkpoint-xxx-merged'
 ```
 
 ## Web-UI

diff --git a/examples/pytorch/animatediff/scripts/full/infer.sh b/examples/pytorch/animatediff/scripts/full/infer.sh
@@ -6,4 +6,3 @@ python animatediff_infer.py \
  --model_id_or_path wyj123456/Realistic_Vision_V5.1_noVAE \
  --sft_type full \
  --ckpt_dir /output/path/like/checkpoints/iter-xxx \
- --eval_human true \
diff --git a/examples/pytorch/animatediff/scripts/lora/infer.sh b/examples/pytorch/animatediff/scripts/lora/infer.sh
@@ -7,4 +7,3 @@ python animatediff_infer.py \
  --motion_adapter_id_or_path Shanghai_AI_Laboratory/animatediff-motion-adapter-v1-5-2 \
  --sft_type lora \
  --ckpt_dir /output/path/like/checkpoints/iter-xxx \
- --eval_human true \
diff --git a/examples/pytorch/llm/scripts/baichuan2_13b_chat/lora_ddp_ds/infer.sh b/examples/pytorch/llm/scripts/baichuan2_13b_chat/lora_ddp_ds/infer.sh
@@ -4,8 +4,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/baichuan2-13b-chat/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 4096 \
  --max_new_tokens 2048 \
  --temperature 0.1 \

diff --git a/examples/pytorch/llm/scripts/baichuan2_13b_chat/lora_mp_ddp/infer.sh b/examples/pytorch/llm/scripts/baichuan2_13b_chat/lora_mp_ddp/infer.sh
@@ -4,8 +4,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0,1 \
 python llm_infer.py \
  --ckpt_dir "output/baichuan2-13b-chat/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 2048 \
  --max_new_tokens 2048 \
  --temperature 0.1 \

diff --git a/examples/pytorch/llm/scripts/baichuan2_13b_chat/qlora_ddp_ds/infer.sh b/examples/pytorch/llm/scripts/baichuan2_13b_chat/qlora_ddp_ds/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/baichuan2-13b-chat/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 4096 \
  --max_new_tokens 2048 \
  --temperature 0.1 \

diff --git a/examples/pytorch/llm/scripts/baichuan2_13b_chat_int4/qlora_ddp_ds/infer.sh b/examples/pytorch/llm/scripts/baichuan2_13b_chat_int4/qlora_ddp_ds/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/baichuan2-13b-chat-int4/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 4096 \
  --max_new_tokens 2048 \
  --temperature 0.1 \

diff --git a/examples/pytorch/llm/scripts/baichuan2_7b/qlora/infer.sh b/examples/pytorch/llm/scripts/baichuan2_7b/qlora/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/baichuan2-7b/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 2048 \
  --max_new_tokens 2048 \
  --temperature 0.7 \

diff --git a/examples/pytorch/llm/scripts/baichuan2_7b_chat/lora_ddp/infer.sh b/examples/pytorch/llm/scripts/baichuan2_7b_chat/lora_ddp/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/baichuan2-7b-chat/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 4096 \
  --max_new_tokens 2048 \
  --temperature 0.1 \

diff --git a/examples/pytorch/llm/scripts/baichuan2_7b_chat/lora_ddp_ds/infer.sh b/examples/pytorch/llm/scripts/baichuan2_7b_chat/lora_ddp_ds/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/baichuan2-7b-chat/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 4096 \
  --max_new_tokens 2048 \
  --temperature 0.1 \

diff --git a/examples/pytorch/llm/scripts/baichuan2_7b_chat/qlora_ddp_ds/infer.sh b/examples/pytorch/llm/scripts/baichuan2_7b_chat/qlora_ddp_ds/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/baichuan2-7b-chat/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 4096 \
  --max_new_tokens 2048 \
  --temperature 0.1 \

diff --git a/examples/pytorch/llm/scripts/baichuan2_7b_chat_int4/qlora_ddp_ds/infer.sh b/examples/pytorch/llm/scripts/baichuan2_7b_chat_int4/qlora_ddp_ds/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/baichuan2-7b-chat-int4/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 4096 \
  --max_new_tokens 2048 \
  --temperature 0.1 \

diff --git a/examples/pytorch/llm/scripts/baichuan_13b_chat/qlora_ddp_ds/infer.sh b/examples/pytorch/llm/scripts/baichuan_13b_chat/qlora_ddp_ds/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/baichuan-13b-chat/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 2048 \
  --max_new_tokens 2048 \
  --temperature 0.1 \

diff --git a/examples/pytorch/llm/scripts/bluelm_7b_chat/lora/infer.sh b/examples/pytorch/llm/scripts/bluelm_7b_chat/lora/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/bluelm-7b-chat/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 2048 \
  --max_new_tokens 2048 \
  --temperature 0.1 \

diff --git a/examples/pytorch/llm/scripts/chatglm2_6b/lora_ddp_ds/infer.sh b/examples/pytorch/llm/scripts/chatglm2_6b/lora_ddp_ds/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/chatglm2-6b/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 4096 \
  --max_new_tokens 2048 \
  --temperature 0.1 \

diff --git a/examples/pytorch/llm/scripts/chatglm3_6b/lora/infer.sh b/examples/pytorch/llm/scripts/chatglm3_6b/lora/infer.sh
@@ -4,8 +4,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/chatglm3-6b/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 2048 \
  --max_new_tokens 2048 \
  --temperature 0.1 \

diff --git a/examples/pytorch/llm/scripts/chatglm3_6b/lora_ddp_ds/infer.sh b/examples/pytorch/llm/scripts/chatglm3_6b/lora_ddp_ds/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/chatglm3-6b/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 4096 \
  --max_new_tokens 2048 \
  --temperature 0.1 \

diff --git a/examples/pytorch/llm/scripts/chatglm3_6b_32k/qlora/infer.sh b/examples/pytorch/llm/scripts/chatglm3_6b_32k/qlora/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/chatglm3-6b-32k/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 4096 \
  --max_new_tokens 2048 \
  --temperature 0.1 \

diff --git a/examples/pytorch/llm/scripts/chatglm3_6b_32k/rome.sh b/examples/pytorch/llm/scripts/chatglm3_6b_32k/rome.sh
@@ -6,7 +6,6 @@ python rome_infer.py \
  --model_revision master \
  --template_type chatglm3 \
  --dtype AUTO \
- --eval_human true \
  --max_new_tokens 128 \
  --temperature 0.1 \
  --top_p 0.7 \

diff --git a/examples/pytorch/llm/scripts/chatglm3_6b_base/lora/infer.sh b/examples/pytorch/llm/scripts/chatglm3_6b_base/lora/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/chatglm3-6b-base/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 2048 \
  --max_new_tokens 2048 \
  --temperature 0.7 \

diff --git a/examples/pytorch/llm/scripts/chatglm3_6b_base/lora_ddp_ds/infer.sh b/examples/pytorch/llm/scripts/chatglm3_6b_base/lora_ddp_ds/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/chatglm3-6b-base/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 2048 \
  --max_new_tokens 2048 \
  --temperature 0.7 \

diff --git a/examples/pytorch/llm/scripts/codefuse_codellama_34b/lora/infer.sh b/examples/pytorch/llm/scripts/codefuse_codellama_34b/lora/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/codefuse-codellama-34b-chat/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 4096 \
  --use_flash_attn true \
  --max_new_tokens 2048 \

diff --git a/examples/pytorch/llm/scripts/custom/tigerbot_13b_chat/qlora_ddp_ds/infer.sh b/examples/pytorch/llm/scripts/custom/tigerbot_13b_chat/qlora_ddp_ds/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/tigerbot-13b-chat/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 2048 \
  --max_new_tokens 2048 \
  --temperature 0.3 \

diff --git a/examples/pytorch/llm/scripts/custom/tigerbot_7b/lora_ddp_ds/infer.sh b/examples/pytorch/llm/scripts/custom/tigerbot_7b/lora_ddp_ds/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/tigerbot-13b/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 2048 \
  --max_new_tokens 2048 \
  --temperature 0.3 \

diff --git a/examples/pytorch/llm/scripts/internlm_20b/lora_ddp/infer.sh b/examples/pytorch/llm/scripts/internlm_20b/lora_ddp/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/internlm-20b/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 2048 \
  --max_new_tokens 2048 \
  --temperature 0.3 \

diff --git a/examples/pytorch/llm/scripts/internlm_20b/qlora/infer.sh b/examples/pytorch/llm/scripts/internlm_20b/qlora/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/internlm-20b/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 2048 \
  --max_new_tokens 2048 \
  --temperature 0.7 \

diff --git a/examples/pytorch/llm/scripts/internlm_20b_chat/lora_ddp/infer.sh b/examples/pytorch/llm/scripts/internlm_20b_chat/lora_ddp/infer.sh
@@ -4,8 +4,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/internlm-20b-chat/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 2048 \
  --max_new_tokens 2048 \
  --temperature 0.1 \

diff --git a/examples/pytorch/llm/scripts/internlm_20b_chat/qlora/infer.sh b/examples/pytorch/llm/scripts/internlm_20b_chat/qlora/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/internlm-20b-chat/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 2048 \
  --max_new_tokens 2048 \
  --temperature 0.1 \

diff --git a/examples/pytorch/llm/scripts/internlm_20b_chat/qlora_ddp/infer.sh b/examples/pytorch/llm/scripts/internlm_20b_chat/qlora_ddp/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/internlm-20b-chat/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 2048 \
  --max_new_tokens 2048 \
  --temperature 0.1 \

diff --git a/examples/pytorch/llm/scripts/llama2_13b_chat/qlora_ddp_ds/infer.sh b/examples/pytorch/llm/scripts/llama2_13b_chat/qlora_ddp_ds/infer.sh
@@ -3,8 +3,7 @@ PYTHONPATH=../../.. \
 CUDA_VISIBLE_DEVICES=0 \
 python llm_infer.py \
  --ckpt_dir "output/llama2-13b-chat/vx_xxx/checkpoint-xxx" \
- --load_args_from_ckpt_dir true \
- --eval_human false \
+ --load_dataset_config true \
  --max_length 4096 \
  --max_new_tokens 2048 \
  --temperature 0.1 \

diff --git a/examples/pytorch/llm/scripts/llama2_13b_chat/rome.sh b/examples/pytorch/llm/scripts/llama2_13b_chat/rome.sh
@@ -6,7 +6,6 @@ python rome_infer.py \
  --model_revision master \
  --template_type llama \
  --dtype AUTO \
- --eval_human true \
  --max_new_tokens 128 \
  --temperature 0.1 \
  --top_p 0.7 \