efort
/
glm4
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

fix #23

This commit is contained in:
zR 2024-06-05 15:16:21 +08:00
parent 1b387cc08d
commit 268e774a1d
2 changed files with 76 additions and 72 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

74
finetune_demo/README.md
View File

@ -36,7 +36,7 @@ pip install -r requirements.txt
如果您仅希望微调模型的对话能力，而非工具能力，您应该按照以下格式整理数据。 如果您仅希望微调模型的对话能力，而非工具能力，您应该按照以下格式整理数据。
``` ```json
[ [
{ {
"messages": [ "messages": [
@ -49,45 +49,47 @@ pip install -r requirements.txt
"args": { "args": {
"<arg name>": "<arg value>" "<arg name>": "<arg value>"
} }
},
{
"role": "user",
"content": "<user prompt text>"
},
{
"role": "assistant",
"content": "<assistant response text>"
},
// If Tool Using:
{
"role": "user",
"content": "<user prompt text>"
},
{
"role": "assistant",
"content": "<assistant response text>"
},
{
"role": "observation",
"content": "<observation prompt text>"
},
{
"role": "assistant",
"content": "<assistant response observation>"
},
// Multi_turns:
{
"role": "user",
"content": "<user prompt text>"
},
{
"role": "assistant",
"content": "<assistant response text>"
} }
// Add more tools if needed
] ]
},
{
"role": "user",
"content": "<user prompt text>"
},
{
"role": "assistant",
"content": "<assistant response text>"
},
// If Tool Using
{
"role": "user",
"content": "<user prompt text>"
},
{
"role": "assistant",
"content": "<assistant response text>"
},
{
"role": "observation",
"content": "<observation prompt text>"
},
{
"role": "assistant",
"content": "<assistant response observation>"
},
// Multi_turns
{
"role": "user",
"content": "<user prompt text>"
},
{
"role": "assistant",
"content": "<assistant response text>"
} }
// ...
] ]
}
]
``` ```
这里是一个不带有工具的例子: 这里是一个不带有工具的例子:

74
finetune_demo/README_en.md
View File

@ -32,7 +32,7 @@ The multi-round dialogue fine-tuning example uses the GLM-4 dialogue format conv
For data files, the sample uses the following format: For data files, the sample uses the following format:
``` ```json
[ [
{ {
"messages": [ "messages": [
@ -45,45 +45,47 @@ For data files, the sample uses the following format:
"args": { "args": {
"<arg name>": "<arg value>" "<arg name>": "<arg value>"
} }
},
{
"role": "user",
"content": "<user prompt text>"
},
{
"role": "assistant",
"content": "<assistant response text>"
},
// If Tool Using:
{
"role": "user",
"content": "<user prompt text>"
},
{
"role": "assistant",
"content": "<assistant response text>"
},
{
"role": "observation",
"content": "<observation prompt text>"
},
{
"role": "assistant",
"content": "<assistant response observation>"
},
// Multi_turns:
{
"role": "user",
"content": "<user prompt text>"
},
{
"role": "assistant",
"content": "<assistant response text>"
} }
// Add more tools if needed
] ]
},
{
"role": "user",
"content": "<user prompt text>"
},
{
"role": "assistant",
"content": "<assistant response text>"
},
// If Tool Using
{
"role": "user",
"content": "<user prompt text>"
},
{
"role": "assistant",
"content": "<assistant response text>"
},
{
"role": "observation",
"content": "<observation prompt text>"
},
{
"role": "assistant",
"content": "<assistant response observation>"
},
// Multi_turns
{
"role": "user",
"content": "<user prompt text>"
},
{
"role": "assistant",
"content": "<assistant response text>"
} }
// ...
] ]
}
]
``` ```
This is a sample without tools: This is a sample without tools: