| """Check actual HTTP endpoint execution, including reported device.""" |
| import argparse |
| import json |
| import urllib.request |
|
|
| p = argparse.ArgumentParser() |
| p.add_argument('--backend', default='cpu-reference') |
| p.add_argument('--url', default='http://127.0.0.1:8080') |
| a = p.parse_args() |
|
|
| def request(path, data=None): |
| body = json.dumps(data).encode() if data is not None else None |
| req = urllib.request.Request(a.url+path, data=body, |
| headers={'Content-Type': 'application/json'} if body else {}) |
| with urllib.request.urlopen(req, timeout=600) as resp: |
| assert resp.status == 200 |
| return json.load(resp) |
|
|
| health = request('/health') |
| assert health['status'] == 'ok', health |
| assert health['backend'] == a.backend, health |
| info = request('/v1/model') |
| assert info['architecture'] == 'falcon_h1' |
| assert info['backend'] == a.backend |
| assert len(request('/v1/logits', {'input_ids': [1,2,3]})['logits']) == 64 |
| assert len(request('/v1/generate', {'input_ids': [1,2], 'max_new_tokens':2})['output_ids']) == 4 |
| print('PASS:', a.backend, 'health, model, logits, generate via HTTP') |
|
|
