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 | class LlamaAPI(CustomLLM):
"""LlamaAPI LLM.
Examples:
`pip install llama-index-llms-llama-api`
```python
from llama_index.llms.llama_api import LlamaAPI
# Obtain an API key from https://www.llama-api.com/
api_key = "your-api-key"
llm = LlamaAPI(api_key=api_key)
# Call the complete method with a prompt
resp = llm.complete("Paul Graham is ")
print(resp)
```
"""
model: str = Field(description="The llama-api model to use.")
temperature: float = Field(description="The temperature to use for sampling.")
max_tokens: int = Field(description="The maximum number of tokens to generate.")
additional_kwargs: Dict[str, Any] = Field(
default_factory=dict, description="Additional kwargs for the llama-api API."
)
_client: Any = PrivateAttr()
def __init__(
self,
model: str = "llama-13b-chat",
temperature: float = 0.1,
max_tokens: int = DEFAULT_NUM_OUTPUTS,
additional_kwargs: Optional[Dict[str, Any]] = None,
api_key: Optional[str] = None,
callback_manager: Optional[CallbackManager] = None,
system_prompt: Optional[str] = None,
messages_to_prompt: Optional[Callable[[Sequence[ChatMessage]], str]] = None,
completion_to_prompt: Optional[Callable[[str], str]] = None,
pydantic_program_mode: PydanticProgramMode = PydanticProgramMode.DEFAULT,
output_parser: Optional[BaseOutputParser] = None,
) -> None:
super().__init__(
model=model,
temperature=temperature,
max_tokens=max_tokens,
additional_kwargs=additional_kwargs or {},
callback_manager=callback_manager,
system_prompt=system_prompt,
messages_to_prompt=messages_to_prompt,
completion_to_prompt=completion_to_prompt,
pydantic_program_mode=pydantic_program_mode,
output_parser=output_parser,
)
self._client = Client(api_key)
@classmethod
def class_name(cls) -> str:
return "llama_api_llm"
@property
def _model_kwargs(self) -> Dict[str, Any]:
base_kwargs = {
"model": self.model,
"temperature": self.temperature,
"max_length": self.max_tokens,
}
return {
**base_kwargs,
**self.additional_kwargs,
}
@property
def metadata(self) -> LLMMetadata:
return LLMMetadata(
context_window=4096,
num_output=DEFAULT_NUM_OUTPUTS,
is_chat_model=True,
is_function_calling_model=True,
model_name="llama-api",
)
@llm_chat_callback()
def chat(self, messages: Sequence[ChatMessage], **kwargs: Any) -> ChatResponse:
message_dicts = to_openai_message_dicts(messages)
json_dict = {
"messages": message_dicts,
**self._model_kwargs,
**kwargs,
}
response = self._client.run(json_dict).json()
message_dict = response["choices"][0]["message"]
message = from_openai_message_dict(message_dict)
return ChatResponse(message=message, raw=response)
@llm_completion_callback()
def complete(
self, prompt: str, formatted: bool = False, **kwargs: Any
) -> CompletionResponse:
complete_fn = chat_to_completion_decorator(self.chat)
return complete_fn(prompt, **kwargs)
@llm_completion_callback()
def stream_complete(
self, prompt: str, formatted: bool = False, **kwargs: Any
) -> CompletionResponseGen:
raise NotImplementedError("stream_complete is not supported for LlamaAPI")
@llm_chat_callback()
def stream_chat(
self, messages: Sequence[ChatMessage], **kwargs: Any
) -> ChatResponseGen:
raise NotImplementedError("stream_chat is not supported for LlamaAPI")
|
