andylizf's picture
Upload folder using huggingface_hub
5fed0fc verified
## FrontierCS - Algorithmic Problems
### Problem Structure
Each problem in `problems/{id}/` contains:
```
problems/{id}/
β”œβ”€β”€ statement.txt # Problem description
β”œβ”€β”€ tag.txt # Category tag
β”œβ”€β”€ config.yaml # Time/memory limits, test count
β”œβ”€β”€ testdata/ # Test cases (public: 1 per problem)
β”‚ β”œβ”€β”€ 1.in
β”‚ └── 1.ans
└── chk.cc / interactor.cc # Checker or interactor
```
### Quick Start
#### Start Judge Server
```bash
docker-compose up --build -d # First time
docker-compose up -d # Subsequent runs
```
Judge runs at `http://localhost:8081`.
### How It Works
1. **Fetch problem** statement from judge API
2. **Generate solution** via LLM (C++ code)
3. **Submit** to judge server
4. **Poll** for result
5. **Score** based on test case pass rate
The judge sever will save solutions and their detailed judging results under the folder `algorithmic/submissions`.
### Judge API
| Endpoint | Description |
|----------|-------------|
| `GET /problems` | List all problems |
| `GET /problem/{id}/statement` | Get problem statement |
| `POST /submit` | Submit solution |
| `GET /result/{sid}` | Get submission result |
### Python API
```python
from frontier_cs import FrontierCSEvaluator
evaluator = FrontierCSEvaluator()
# Evaluate an algorithmic problem
result = evaluator.evaluate("algorithmic", problem_id=1, code=cpp_code)
print(f"Score: {result.score}")
# Get unbounded score (without clipping)
result = evaluator.evaluate("algorithmic", problem_id=1, code=cpp_code, unbounded=True)
print(f"Score (bounded): {result.score}")
if hasattr(result, 'score_unbounded'):
print(f"Score (unbounded): {result.score_unbounded}")
```
### CLI
```bash
# Evaluate a solution
frontier-eval --algorithmic 1 solution.cpp
# Get unbounded score
frontier-eval --algorithmic 1 solution.cpp --unbounded
```
### Customized Problems
1. Create `problems/{id}/` directory
2. Add required files:
- `statement.txt`: Problem description
- `config.yaml`: Limits and test count
- `testdata/`: Input/output files
- `chk.cc` or `interactor.cc`: Checker/interactor
3. Restart judge to pick up new problems
### Judge Sever Configuration
#### config.yaml
```yaml
time_limit: 1000 # ms
memory_limit: 262144 # KB
test_count: 10
checker: chk.cc # or interactor: interactor.cc
```
#### docker-compose.yml
```yaml
environment:
PORT: "8081" # API port
JUDGE_WORKERS: "8" # Concurrent evaluations
GJ_PARALLELISM: "8" # go-judge parallelism
```