evalscope/docs/en/advanced_guides/custom_dataset/llm.md

138 lines
6.4 KiB
Markdown

# Large Language Model
This framework supports multiple-choice questions and question-answering questions, with two predefined dataset formats. The usage process is as follows:
## Multiple-Choice Question Format (MCQ)
Suitable for scenarios where users need multiple-choice questions. The evaluation metric is accuracy.
### 1. Data Preparation
Prepare a CSV file in the multiple-choice question format. The directory structure is as follows:
```text
mcq/
├── example_dev.csv # (Optional) Filename should be `{subset_name}_dev.csv`, used for few-shot evaluation
└── example_val.csv # Filename should be `{subset_name}_val.csv`, used for actual evaluation data
```
The CSV file needs to be in the following format:
```text
id,question,A,B,C,D,answer
1,The amino acids that make up animal proteins typically include ____,4 types,22 types,20 types,19 types,C
2,Among the following substances present in the blood, which is not a metabolic end product?____,urea,uric acid,pyruvic acid,carbon dioxide,C
```
Where:
- `id` is the sequence number (optional)
- `question` is the question
- `A`, `B`, `C`, `D`, etc., are the options, with a maximum of 10 options supported
- `answer` is the correct option
### 2. Configuration Task
Run the following code to start the evaluation:
```python
from evalscope import TaskConfig, run_task
task_cfg = TaskConfig(
model='Qwen/Qwen2-0.5B-Instruct',
datasets=['general_mcq'], # Data format, fixed as 'general_mcq' for multiple-choice format
dataset_args={
'general_mcq': {
"local_path": "custom_eval/text/mcq", # Custom dataset path
"subset_list": [
"example" # Evaluation dataset name, mentioned subset_name
]
}
},
)
run_task(task_cfg=task_cfg)
```
Results:
```text
+---------------------+-------------+-----------------+----------+-------+---------+---------+
| Model | Dataset | Metric | Subset | Num | Score | Cat.0 |
+=====================+=============+=================+==========+=======+=========+=========+
| Qwen2-0.5B-Instruct | general_mcq | AverageAccuracy | example | 12 | 0.5833 | default |
+---------------------+-------------+-----------------+----------+-------+---------+---------+
```
## Question-Answering Format (QA)
Suitable for scenarios where users need question-answering questions. The evaluation metrics are `ROUGE` and `BLEU`.
### 1. Data Preparation
Prepare a JSON lines file for the question-answering format. The directory contains a single file:
```text
qa/
└── example.jsonl
```
The JSON lines file needs to be in the following format:
```json
{"system": "You are a geographer", "query": "What is the capital of China?", "response": "The capital of China is Beijing"}
{"query": "What is the highest mountain in the world?", "response": "It's Mount Everest"}
{"query": "Why can't penguins be seen in the Arctic?", "response": "Because penguins mostly live in Antarctica"}
```
Where:
- `system` is the system prompt (optional field)
- `query` is the question (required)
- `response` is the correct answer (required)
### 2. Configuration Task
Run the following code to start the evaluation:
```python
from evalscope import TaskConfig, run_task
task_cfg = TaskConfig(
model='qwen/Qwen2-0.5B-Instruct',
datasets=['general_qa'], # Data format, fixed as 'general_qa' for question-answering format
dataset_args={
'general_qa': {
"local_path": "custom_eval/text/qa", # Custom dataset path
"subset_list": [
"example" # Evaluation dataset name, corresponding to * in the above *.jsonl
]
}
},
)
run_task(task_cfg=task_cfg)
```
Results:
```text
+---------------------+-------------+-----------------+----------+-------+---------+---------+
| Model | Dataset | Metric | Subset | Num | Score | Cat.0 |
+=====================+=============+=================+==========+=======+=========+=========+
| Qwen2-0.5B-Instruct | general_qa | bleu-1 | example | 12 | 0.2324 | default |
+---------------------+-------------+-----------------+----------+-------+---------+---------+
| Qwen2-0.5B-Instruct | general_qa | bleu-2 | example | 12 | 0.1451 | default |
+---------------------+-------------+-----------------+----------+-------+---------+---------+
| Qwen2-0.5B-Instruct | general_qa | bleu-3 | example | 12 | 0.0625 | default |
+---------------------+-------------+-----------------+----------+-------+---------+---------+
| Qwen2-0.5B-Instruct | general_qa | bleu-4 | example | 12 | 0.0556 | default |
+---------------------+-------------+-----------------+----------+-------+---------+---------+
| Qwen2-0.5B-Instruct | general_qa | rouge-1-f | example | 12 | 0.3441 | default |
+---------------------+-------------+-----------------+----------+-------+---------+---------+
| Qwen2-0.5B-Instruct | general_qa | rouge-1-p | example | 12 | 0.2393 | default |
+---------------------+-------------+-----------------+----------+-------+---------+---------+
| Qwen2-0.5B-Instruct | general_qa | rouge-1-r | example | 12 | 0.8889 | default |
+---------------------+-------------+-----------------+----------+-------+---------+---------+
| Qwen2-0.5B-Instruct | general_qa | rouge-2-f | example | 12 | 0.2062 | default |
+---------------------+-------------+-----------------+----------+-------+---------+---------+
| Qwen2-0.5B-Instruct | general_qa | rouge-2-p | example | 12 | 0.1453 | default |
+---------------------+-------------+-----------------+----------+-------+---------+---------+
| Qwen2-0.5B-Instruct | general_qa | rouge-2-r | example | 12 | 0.6167 | default |
+---------------------+-------------+-----------------+----------+-------+---------+---------+
| Qwen2-0.5B-Instruct | general_qa | rouge-l-f | example | 12 | 0.333 | default |
+---------------------+-------------+-----------------+----------+-------+---------+---------+
| Qwen2-0.5B-Instruct | general_qa | rouge-l-p | example | 12 | 0.2324 | default |
+---------------------+-------------+-----------------+----------+-------+---------+---------+
| Qwen2-0.5B-Instruct | general_qa | rouge-l-r | example | 12 | 0.8889 | default |
+---------------------+-------------+-----------------+----------+-------+---------+---------+
```