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
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322 | class NVIDIA(OpenAILike, FunctionCallingLLM):
"""NVIDIA's API Catalog Connector."""
_mode: str = PrivateAttr(default="nvidia")
_client: Any = PrivateAttr()
_aclient: Any = PrivateAttr()
_is_hosted: bool = PrivateAttr(True)
def __init__(
self,
model: Optional[str] = None,
nvidia_api_key: Optional[str] = None,
api_key: Optional[str] = None,
base_url: Optional[str] = os.getenv("NVIDIA_BASE_URL", BASE_URL),
max_tokens: Optional[int] = 1024,
**kwargs: Any,
) -> None:
"""
Initialize an instance of the NVIDIA class.
This class provides an interface to the NVIDIA NIM. By default, it connects to a hosted NIM,
but you can switch to an on-premises NIM by providing a `base_url`.
Args:
model (str, optional): The model to use for the NIM.
nvidia_api_key (str, optional): The API key for the NVIDIA NIM. Defaults to None.
api_key (str, optional): An alternative parameter for providing the API key. Defaults to None.
base_url (str, optional): The base URL for the NIM. Use this to switch to an on-premises NIM.
max_tokens (int, optional): The maximum number of tokens to generate. Defaults to 1024.
**kwargs: Additional keyword arguments.
API Keys:
- The recommended way to provide the API key is through the `NVIDIA_API_KEY` environment variable.
Raises:
DeprecationWarning: If an API key is not provided for a hosted NIM, a warning is issued. This will become an error in version 0.2.0.
"""
api_key = get_from_param_or_env(
"api_key",
nvidia_api_key or api_key,
"NVIDIA_API_KEY",
"NO_API_KEY_PROVIDED",
)
base_url = base_url or BASE_URL
super().__init__(
api_key=api_key,
api_base=base_url,
max_tokens=max_tokens,
default_headers={"User-Agent": "llama-index-llms-nvidia"},
**kwargs,
)
self.model = model
self._is_hosted = base_url in KNOWN_URLS or base_url == BASE_URL
if self._is_hosted and api_key == "NO_API_KEY_PROVIDED":
warnings.warn(
"An API key is required for the hosted NIM. This will become an error in 0.2.0.",
)
if not self.model:
if self._is_hosted:
self.model = DEFAULT_MODEL
else:
self.__get_default_model()
if not self.model.startswith("nvdev/"):
self._validate_model(self.model) ## validate model
self.is_chat_model = self._is_chat_model()
self.is_function_calling_model = self._is_function_calling_model()
def __get_default_model(self):
"""Set default model."""
if not self._is_hosted:
valid_models = [
model.id
for model in self.available_models
if not model.base_model or model.base_model == model.id
]
self.model = next(iter(valid_models), None)
if self.model:
warnings.warn(
f"Default model is set as: {self.model}. \n"
"Set model using model parameter. \n"
"To get available models use available_models property.",
UserWarning,
)
else:
raise ValueError("No locally hosted model was found.")
else:
self.model = DEFAULT_MODEL
def _validate_url(self, base_url):
"""
validate the base_url.
if the base_url is not a url, raise an error
if the base_url does not end in /v1, e.g. /completions, /chat/completions,
emit a warning. old documentation told users to pass in the full
inference url, which is incorrect and prevents model listing from working.
normalize base_url to end in /v1.
"""
if base_url is not None:
base_url = base_url.rstrip("/")
parsed = urlparse(base_url)
# Ensure scheme and netloc (domain name) are present
if not (parsed.scheme and parsed.netloc):
expected_format = "Expected format is: http://host:port"
raise ValueError(
f"Invalid base_url format. {expected_format} Got: {base_url}"
)
return base_url
def _validate_model(self, model_name: str) -> None:
"""
Validates compatibility of the hosted model with the client.
Args:
model_name (str): The name of the model.
Raises:
ValueError: If the model is incompatible with the client.
"""
if self._is_hosted:
if model := determine_model(model_name):
if not model.client:
warnings.warn(f"Unable to determine validity of {model.id}")
elif model.client != self.class_name():
raise ValueError(
f"Model {model.id} is incompatible with client {self.class_name()}. "
f"Please check `{self.class_name()}.get_available_models`"
)
if model.endpoint:
self.api_base = model.endpoint
else:
candidates = [
model for model in self.available_models if model.id == model_name
]
assert len(candidates) <= 1, (
f"Multiple candidates for {model_name} "
f"in `available_models`: {candidates}"
)
if candidates:
model = candidates[0]
warnings.warn(
f"Found {model_name} in available_models, but type is "
"unknown and inference may fail."
)
else:
if model_name.startswith("nvdev/"): # assume valid
model = Model(id=model_name)
else:
raise ValueError(
f"Model {model_name} is unknown, "
"check `available_models`"
)
else:
if model_name not in [model.id for model in self.available_models]:
raise ValueError(f"No locally hosted {model_name} was found.")
@property
def available_models(self) -> List[Model]:
models = []
for element in self._get_client().models.list().data:
if not (model := determine_model(element.id)):
model = Model(id=element.id)
models.append(model)
# only exclude models in hosted mode. in non-hosted mode, the administrator has control
# over the model name and may deploy an excluded name that will work.
if self._is_hosted:
exclude = {
"mistralai/mixtral-8x22b-v0.1", # not a /chat/completion endpoint
}
models = [model for model in models if model.id not in exclude]
return models
@classmethod
def class_name(cls) -> str:
return "NVIDIA"
@deprecated(
version="0.1.3",
reason="Will be removed in 0.2. Construct with `base_url` instead.",
)
def mode(
self,
mode: Optional[Literal["nvidia", "nim"]] = "nvidia",
*,
base_url: Optional[str] = None,
model: Optional[str] = None,
api_key: Optional[str] = None,
) -> "NVIDIA":
"""
Deprecated: use NVIDIA(base_url="...") instead.
"""
if mode == "nim":
if not base_url:
raise ValueError("base_url is required for nim mode")
if mode == "nvidia":
api_key = get_from_param_or_env(
"api_key",
api_key,
"NVIDIA_API_KEY",
)
base_url = base_url or BASE_URL
self._mode = mode
if base_url:
self.api_base = base_url
if model:
self.model = model
if api_key:
self.api_key = api_key
return self
def _is_chat_model(self):
model = determine_model(self.model)
return model and model.id in CHAT_MODEL_TABLE
def _is_function_calling_model(self):
model = determine_model(self.model)
return model and model.supports_tools
def _prepare_chat_with_tools(
self,
tools: List["BaseTool"],
user_msg: Optional[Union[str, ChatMessage]] = None,
chat_history: Optional[List[ChatMessage]] = None,
verbose: bool = False,
allow_parallel_tool_calls: bool = False,
**kwargs: Any,
) -> Dict[str, Any]:
"""Prepare the chat with tools."""
# misralai uses the same openai tool format
tool_specs = [
tool.metadata.to_openai_tool(skip_length_check=True) for tool in tools
]
if isinstance(user_msg, str):
user_msg = ChatMessage(role=MessageRole.USER, content=user_msg)
messages = chat_history or []
if user_msg:
messages.append(user_msg)
return {
"messages": messages,
"tools": tool_specs or None,
**kwargs,
}
def get_tool_calls_from_response(
self,
response: "ChatResponse",
error_on_no_tool_call: bool = True,
) -> List[ToolSelection]:
"""Predict and call the tool."""
tool_calls = response.message.additional_kwargs.get("tool_calls", [])
if len(tool_calls) < 1:
if error_on_no_tool_call:
raise ValueError(
f"Expected at least one tool call, but got {len(tool_calls)} tool calls."
)
else:
return []
tool_selections = []
for tool_call in tool_calls:
# if not isinstance(tool_call, ToolCall):
# raise ValueError("Invalid tool_call object")
argument_dict = json.loads(tool_call.function.arguments)
tool_selections.append(
ToolSelection(
tool_id=tool_call.id,
tool_name=tool_call.function.name,
tool_kwargs=argument_dict,
)
)
return tool_selections
|
