File size: 6,679 Bytes
e3bb7ae
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
import requests
import json
import time
import sys

BASE_URL = "http://localhost:8000/v1"
MODEL_NAME = "RWKV-GLM-4.7-Flash-Preview-v0.1"

# ==========================================================
# Utility
# ==========================================================
def print_section(title):
    print("\n" + "=" * 60)
    print(title)
    print("=" * 60)


def safe_json(resp):
    try:
        return resp.json()
    except:
        print("❌ JSON decode failed")
        print(resp.text)
        sys.exit(1)


# ==========================================================
# 1️⃣ Models API
# ==========================================================
def test_models():
    print_section("TEST: /v1/models")

    resp = requests.get(f"{BASE_URL}/models")
    assert resp.status_code == 200, "Models API failed"

    data = safe_json(resp)

    assert "data" in data, "No model list returned"
    assert len(data["data"]) > 0, "Empty model list"

    print("✅ Models endpoint OK")
    print("Available models:", [m["id"] for m in data["data"]])


# ==========================================================
# 2️⃣ Non-stream basic
# ==========================================================
def test_basic_completion():
    print_section("TEST: Basic Non-Streaming Completion")

    payload = {
        "model": MODEL_NAME,
        "messages": [{"role": "user", "content": "Say hello."}],
        "max_tokens": 30,
        "stream": False
    }

    resp = requests.post(
        f"{BASE_URL}/chat/completions",
        headers={"Content-Type": "application/json"},
        data=json.dumps(payload)
    )

    assert resp.status_code == 200, "Completion failed"

    data = safe_json(resp)

    assert "choices" in data, "No choices returned"
    assert "usage" in data, "No usage returned"

    print("Assistant:", data["choices"][0]["message"]["content"])
    print("Usage:", data["usage"])
    print("✅ Basic completion OK")


# ==========================================================
# 3️⃣ Streaming
# ==========================================================
def test_streaming():
    print_section("TEST: Streaming Completion")

    payload = {
        "model": MODEL_NAME,
        "messages": [{"role": "user", "content": "Count from 1 to 5."}],
        "max_tokens": 50,
        "stream": True
    }

    full_text = ""

    with requests.post(
        f"{BASE_URL}/chat/completions",
        headers={"Content-Type": "application/json"},
        data=json.dumps(payload),
        stream=True
    ) as resp:

        assert resp.status_code == 200, "Streaming failed"

        for line in resp.iter_lines():
            if line:
                decoded = line.decode("utf-8")

                if decoded.startswith("data: "):
                    content = decoded[len("data: "):]

                    if content == "[DONE]":
                        break

                    chunk = json.loads(content)
                    delta = chunk["choices"][0]["delta"]

                    if "content" in delta:
                        print(delta["content"], end="", flush=True)
                        full_text += delta["content"]

    print("\n\n✅ Streaming OK")
    assert len(full_text) > 0, "Streaming returned empty"


# ==========================================================
# 4️⃣ Sampling Variations
# ==========================================================
def test_sampling_variations():
    print_section("TEST: Sampling Variations")

    base_payload = {
        "model": MODEL_NAME,
        "messages": [{"role": "user", "content": "Write a creative sentence about AI."}],
        "max_tokens": 50,
        "stream": False
    }

    configs = [
        {"temperature": 0.0},
        {"temperature": 0.7},
        {"top_p": 0.8},
        {"top_k": 20},
        {"repetition_penalty": 1.2},
        {"presence_penalty": 0.5},
        {"frequency_penalty": 0.5}
    ]

    for cfg in configs:
        payload = base_payload.copy()
        payload.update(cfg)

        resp = requests.post(
            f"{BASE_URL}/chat/completions",
            headers={"Content-Type": "application/json"},
            data=json.dumps(payload)
        )

        assert resp.status_code == 200, f"Sampling failed: {cfg}"

        data = safe_json(resp)

        text = data["choices"][0]["message"]["content"]

        print(f"\nConfig: {cfg}")
        print("Output:", text[:120], "...")

    print("\n✅ Sampling parameter variations OK")


# ==========================================================
# 5️⃣ Deterministic Check (temperature=0)
# ==========================================================
def test_deterministic():
    print_section("TEST: Deterministic Mode (temperature=0)")

    payload = {
        "model": MODEL_NAME,
        "messages": [{"role": "user", "content": "Define gravity in one sentence."}],
        "temperature": 0.0,
        "max_tokens": 50,
        "stream": False
    }

    resp1 = requests.post(f"{BASE_URL}/chat/completions",
                          headers={"Content-Type": "application/json"},
                          data=json.dumps(payload))
    resp2 = requests.post(f"{BASE_URL}/chat/completions",
                          headers={"Content-Type": "application/json"},
                          data=json.dumps(payload))

    out1 = safe_json(resp1)["choices"][0]["message"]["content"]
    out2 = safe_json(resp2)["choices"][0]["message"]["content"]

    print("Run1:", out1)
    print("Run2:", out2)

    assert out1 == out2, "❌ Deterministic mode not deterministic"
    print("✅ Deterministic check OK")


# ==========================================================
# 6️⃣ Error Handling
# ==========================================================
def test_error_handling():
    print_section("TEST: Error Handling")

    payload = {
        "model": MODEL_NAME,
        # missing messages intentionally
    }

    resp = requests.post(
        f"{BASE_URL}/chat/completions",
        headers={"Content-Type": "application/json"},
        data=json.dumps(payload)
    )

    if resp.status_code != 200:
        print("✅ Server correctly handled bad request")
    else:
        print("⚠️ Warning: server did not reject bad request")


# ==========================================================
# Main
# ==========================================================
if __name__ == "__main__":
    start = time.time()

    test_models()
    test_basic_completion()
    test_streaming()
    test_sampling_variations()
    test_deterministic()
    test_error_handling()

    print_section("ALL TESTS PASSED")
    print(f"Total time: {round(time.time() - start, 2)} sec")