hal/tiny_agent.py at main · ROC-HCI/hal · GitHub

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
from openai import OpenAI

class TinyAgent:

    def __init__(self, model, tokenizer=None, debug=False, api_key=None):

        self.model = model
        self.messages = []
        self.max_tokens = 10072
        self.debug = False
        self.reasoning_effort = "low"
        self.temperature = 1
        self.port = 11434
        self.client = OpenAI(api_key=api_key)

    def clear_messages(self):
        self.messages = list()

    def add_message(self,message_type, message):
        self.messages.append({"role": message_type, "content":message})

    def add_system_message(self, message):
        self.add_message("system", message)

    def add_user_message(self, message):
        self.add_message("user", message)

    def add_assistant_message(self, message):
        self.add_message("assistant", message)

    def set_max_tokens(self, max_tokens):
        self.max_tokens = max_tokens

    def set_debug(self, debug):
        self.debug = debug

    def set_reasoning_effort(self, reasoning_effort):
        self.reasoning_effort = reasoning_effort

    # Added
    def set_port(self, port):
        self.port = port

    # Added
    def set_client(self, api_key):
        self.port = port

    def call(self, prompt="", response_type="text", cache=True):
        messages = self.messages.copy()
        if prompt:
            messages.append({"role": "user", "content":prompt})
        if cache:
            self.add_user_message(prompt)

        if "gpt-5" in self.model:
            response = self.client.responses.create(
                model=self.model,
                input=messages,
                reasoning={"effort": self.reasoning_effort},
                text={
                    "format": {
                      "type": response_type
                    },
                    "verbosity": "low"
                  },
            )
            reply = response.output_text

        elif "gpt-4" in self.model or "o3" in self.model or "o4" in self.model:
            response = self.client.responses.create(
              model=self.model,
              input=messages,
              temperature=self.temperature,
              max_output_tokens=self.max_tokens,
              top_p=1,
              text={
                "format": {
                  "type": response_type # "text", "json_object"
                }
              }
            )
            reply = response.output_text

        #Added
        elif "gpt-oss:20b" in self.model or "gpt-oss:120b" in self.model:

            client = OpenAI(
                base_url=f"http://localhost:{self.port}/v1", # Local Ollama API
                api_key="ollama" # Dummy key
            )

            messages.insert(0, {"role": "system", "content": f"Reasoning: {self.reasoning_effort}"})

            response = client.chat.completions.create(
                model=self.model,
                messages=messages
            )

            reply = response.choices[0].message.content

        #Added
        elif "deepseek-r1:70b" in self.model:

            client = OpenAI(
                base_url=f"http://localhost:{self.port}/v1", # Local Ollama API
                api_key="ollama" # Dummy key
            )

            response = client.chat.completions.create(
                model=self.model,
                messages=messages,
            )

            reply = response.choices[0].message.content
            reply = re.sub(r"<think>.*?</think>\n\n", "", reply, flags=re.DOTALL)

        if self.debug:
            print(reply)
        if cache:
            self.add_assistant_message(reply)
        return reply

    def load_json(self,s):
        import json
        try:
            return json.loads(s)
        except json.JSONDecodeError:
            return None

    def call_json(self, prompt=""):
        self.add_system_message("Reply must be JSON format.")
        reply = self.call(prompt=prompt, response_type="json_object")
        if not reply:
            print("Empty reply")
            return None

        reply = reply.strip()
        if reply.startswith("```json"):
            reply = reply[len("```json"):].strip()
            if reply.endswith("```"):
                reply = reply[:-3].strip()

        # Use OR, and guard length
        if not (reply.startswith("{") and reply.endswith("}")):
            print("Not JSON structure")
            return None

        try:
            return self.load_json(reply)
        except Exception:
            print("Error parsing JSON")
            return None