chore: Add code interpreter skill and update vehicle status template
Browse files- .vscode/launch.json +15 -0
- kitt/core/model.py +46 -17
- kitt/skills/__init__.py +1 -0
- kitt/skills/interpreter.py +52 -0
- kitt/skills/routing.py +42 -6
- kitt/skills/vehicle.py +3 -7
- main.py +33 -2
.vscode/launch.json
ADDED
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
// Use IntelliSense to learn about possible attributes.
|
3 |
+
// Hover to view descriptions of existing attributes.
|
4 |
+
// For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
|
5 |
+
"version": "0.2.0",
|
6 |
+
"configurations": [
|
7 |
+
{
|
8 |
+
"name": "RUN KITT",
|
9 |
+
"type": "debugpy",
|
10 |
+
"request": "launch",
|
11 |
+
"program": "main.py",
|
12 |
+
"console": "integratedTerminal"
|
13 |
+
}
|
14 |
+
]
|
15 |
+
}
|
kitt/core/model.py
CHANGED
@@ -6,6 +6,7 @@ from langchain.memory import ChatMessageHistory
|
|
6 |
from langchain_core.messages import HumanMessage, AIMessage, ToolMessage
|
7 |
from langchain_core.utils.function_calling import convert_to_openai_function
|
8 |
import ollama
|
|
|
9 |
from pydantic import BaseModel
|
10 |
from loguru import logger
|
11 |
|
@@ -27,12 +28,10 @@ class FunctionCall(BaseModel):
|
|
27 |
|
28 |
|
29 |
schema_json = json.loads(FunctionCall.schema_json())
|
30 |
-
HRMS_SYSTEM_PROMPT = """<|
|
31 |
-
<|im_start|>system
|
32 |
You are a function calling AI agent with self-recursion.
|
33 |
You can call only one function at a time and analyse data you get from function response.
|
34 |
You are provided with function signatures within <tools></tools> XML tags.
|
35 |
-
{car_status}
|
36 |
|
37 |
You may use agentic frameworks for reasoning and planning to help with user query.
|
38 |
Please call a function and wait for function results to be provided to you in the next iteration.
|
@@ -67,8 +66,14 @@ Assistant:
|
|
67 |
{{"arguments": {{"search_query": "Spa"}}, "name": "search_points_of_interests"}}
|
68 |
</tool_call>
|
69 |
|
70 |
-
|
|
|
|
|
|
|
|
|
71 |
|
|
|
|
|
72 |
|
73 |
Use the following pydantic model json schema for each tool call you will make:
|
74 |
{schema}
|
@@ -145,6 +150,8 @@ def parse_tool_calls(text):
|
|
145 |
pattern = r"<tool_call>\s*(\{.*?\})\s*</tool_call>"
|
146 |
|
147 |
if not text.startswith("<tool_call>"):
|
|
|
|
|
148 |
return [], []
|
149 |
|
150 |
matches = re.findall(pattern, text, re.DOTALL)
|
@@ -164,12 +171,22 @@ def parse_tool_calls(text):
|
|
164 |
def process_response(user_query, res, history, tools, depth):
|
165 |
"""Returns True if the response contains tool calls, False otherwise."""
|
166 |
logger.debug(f"Processing response: {res}")
|
167 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
168 |
# TODO: Handle errors
|
169 |
if not tool_calls:
|
170 |
return False, tool_calls, errors
|
171 |
# tool_results = ""
|
172 |
-
|
173 |
for tool_call in tool_calls:
|
174 |
# TODO: Extra Validation
|
175 |
# Call the function
|
@@ -185,12 +202,11 @@ def process_response(user_query, res, history, tools, depth):
|
|
185 |
|
186 |
tool_results = tool_results.strip()
|
187 |
print(f"Tool results: {tool_results}")
|
188 |
-
tool_call_id = uuid.uuid4().hex
|
189 |
history.add_message(ToolMessage(content=tool_results, tool_call_id=tool_call_id))
|
190 |
return True, tool_calls, errors
|
191 |
|
192 |
|
193 |
-
def run_inference_step(history, tools, schema_json, dry_run=False):
|
194 |
# If we decide to call a function, we need to generate the prompt for the model
|
195 |
# based on the history of the conversation so far.
|
196 |
# not break the loop
|
@@ -199,17 +215,26 @@ def run_inference_step(history, tools, schema_json, dry_run=False):
|
|
199 |
print(f"Prompt is:{prompt + AI_PREAMBLE}\n------------------\n")
|
200 |
|
201 |
data = {
|
202 |
-
"prompt": prompt
|
|
|
|
|
203 |
# "streaming": False,
|
204 |
# "model": "smangrul/llama-3-8b-instruct-function-calling",
|
205 |
# "model": "elvee/hermes-2-pro-llama-3:8b-Q5_K_M",
|
206 |
# "model": "NousResearch/Hermes-2-Pro-Llama-3-8B",
|
207 |
-
"model": "interstellarninja/hermes-2-pro-llama-3-8b",
|
|
|
|
|
208 |
"raw": True,
|
209 |
"options": {
|
210 |
"temperature": 0.8,
|
211 |
# "max_tokens": 1500,
|
212 |
"num_predict": 1500,
|
|
|
|
|
|
|
|
|
|
|
213 |
# "num_predict": 1500,
|
214 |
# "max_tokens": 1500,
|
215 |
},
|
@@ -218,8 +243,10 @@ def run_inference_step(history, tools, schema_json, dry_run=False):
|
|
218 |
if dry_run:
|
219 |
print(prompt + AI_PREAMBLE)
|
220 |
return "Didn't really run it."
|
221 |
-
|
222 |
-
|
|
|
|
|
223 |
logger.debug(f"Response from model: {out}")
|
224 |
res = out["response"]
|
225 |
|
@@ -227,18 +254,20 @@ def run_inference_step(history, tools, schema_json, dry_run=False):
|
|
227 |
|
228 |
|
229 |
def process_query(user_query: str, history: ChatMessageHistory, tools):
|
230 |
-
history
|
|
|
|
|
|
|
|
|
231 |
for depth in range(10):
|
232 |
-
out = run_inference_step(history, tools, schema_json)
|
233 |
print(f"Inference step result:\n{out}\n------------------\n")
|
234 |
history.add_message(AIMessage(content=out))
|
235 |
to_continue, tool_calls, errors = process_response(
|
236 |
user_query, out, history, tools, depth
|
237 |
)
|
238 |
if errors:
|
239 |
-
history.add_message(
|
240 |
-
AIMessage(content=f"Errors in tool calls: {errors}")
|
241 |
-
)
|
242 |
|
243 |
if not to_continue:
|
244 |
print(f"This is the answer, no more iterations: {out}")
|
|
|
6 |
from langchain_core.messages import HumanMessage, AIMessage, ToolMessage
|
7 |
from langchain_core.utils.function_calling import convert_to_openai_function
|
8 |
import ollama
|
9 |
+
from ollama import Client
|
10 |
from pydantic import BaseModel
|
11 |
from loguru import logger
|
12 |
|
|
|
28 |
|
29 |
|
30 |
schema_json = json.loads(FunctionCall.schema_json())
|
31 |
+
HRMS_SYSTEM_PROMPT = """<|im_start|>system
|
|
|
32 |
You are a function calling AI agent with self-recursion.
|
33 |
You can call only one function at a time and analyse data you get from function response.
|
34 |
You are provided with function signatures within <tools></tools> XML tags.
|
|
|
35 |
|
36 |
You may use agentic frameworks for reasoning and planning to help with user query.
|
37 |
Please call a function and wait for function results to be provided to you in the next iteration.
|
|
|
66 |
{{"arguments": {{"search_query": "Spa"}}, "name": "search_points_of_interests"}}
|
67 |
</tool_call>
|
68 |
|
69 |
+
Example 3:
|
70 |
+
User: How long will it take to get to the destination?
|
71 |
+
Assistant:
|
72 |
+
<tool_call>
|
73 |
+
{{"arguments": {{"destination": ""}}, "name": "calculate_route"}}
|
74 |
|
75 |
+
When asked for the weather or points of interest, use the appropriate tool with the current location of the car. Unless the user provides a location, then use that location.
|
76 |
+
Always assume user wants to travel by car.
|
77 |
|
78 |
Use the following pydantic model json schema for each tool call you will make:
|
79 |
{schema}
|
|
|
150 |
pattern = r"<tool_call>\s*(\{.*?\})\s*</tool_call>"
|
151 |
|
152 |
if not text.startswith("<tool_call>"):
|
153 |
+
if "<tool_call>" in text:
|
154 |
+
raise ValueError("<text_and_tool_call>")
|
155 |
return [], []
|
156 |
|
157 |
matches = re.findall(pattern, text, re.DOTALL)
|
|
|
171 |
def process_response(user_query, res, history, tools, depth):
|
172 |
"""Returns True if the response contains tool calls, False otherwise."""
|
173 |
logger.debug(f"Processing response: {res}")
|
174 |
+
tool_results = f"Agent iteration {depth} to assist with user query: {user_query}\n"
|
175 |
+
tool_call_id = uuid.uuid4().hex
|
176 |
+
try:
|
177 |
+
tool_calls, errors = parse_tool_calls(res)
|
178 |
+
except ValueError as e:
|
179 |
+
if "<text_and_tool_call>" in str(e):
|
180 |
+
tool_results += f"A mix of text and tool_call was found, you must either answer the query in a short sentence or use tool_call not both. Try again, this time only using tool_call."
|
181 |
+
history.add_message(
|
182 |
+
ToolMessage(content=tool_results, tool_call_id=tool_call_id)
|
183 |
+
)
|
184 |
+
return True, [], []
|
185 |
# TODO: Handle errors
|
186 |
if not tool_calls:
|
187 |
return False, tool_calls, errors
|
188 |
# tool_results = ""
|
189 |
+
|
190 |
for tool_call in tool_calls:
|
191 |
# TODO: Extra Validation
|
192 |
# Call the function
|
|
|
202 |
|
203 |
tool_results = tool_results.strip()
|
204 |
print(f"Tool results: {tool_results}")
|
|
|
205 |
history.add_message(ToolMessage(content=tool_results, tool_call_id=tool_call_id))
|
206 |
return True, tool_calls, errors
|
207 |
|
208 |
|
209 |
+
def run_inference_step(depth, history, tools, schema_json, dry_run=False):
|
210 |
# If we decide to call a function, we need to generate the prompt for the model
|
211 |
# based on the history of the conversation so far.
|
212 |
# not break the loop
|
|
|
215 |
print(f"Prompt is:{prompt + AI_PREAMBLE}\n------------------\n")
|
216 |
|
217 |
data = {
|
218 |
+
"prompt": prompt
|
219 |
+
+ "\nThis is the first turn and you don't have <tool_results> to analyze yet"
|
220 |
+
+ AI_PREAMBLE,
|
221 |
# "streaming": False,
|
222 |
# "model": "smangrul/llama-3-8b-instruct-function-calling",
|
223 |
# "model": "elvee/hermes-2-pro-llama-3:8b-Q5_K_M",
|
224 |
# "model": "NousResearch/Hermes-2-Pro-Llama-3-8B",
|
225 |
+
# "model": "interstellarninja/hermes-2-pro-llama-3-8b",
|
226 |
+
"model": "dolphin-llama3:8b",
|
227 |
+
# "model": "dolphin-llama3:70b",
|
228 |
"raw": True,
|
229 |
"options": {
|
230 |
"temperature": 0.8,
|
231 |
# "max_tokens": 1500,
|
232 |
"num_predict": 1500,
|
233 |
+
"mirostat": 1,
|
234 |
+
# "mirostat_tau": 2,
|
235 |
+
"repeat_penalty": 1.5,
|
236 |
+
"top_k": 25,
|
237 |
+
"top_p": 0.5,
|
238 |
# "num_predict": 1500,
|
239 |
# "max_tokens": 1500,
|
240 |
},
|
|
|
243 |
if dry_run:
|
244 |
print(prompt + AI_PREAMBLE)
|
245 |
return "Didn't really run it."
|
246 |
+
|
247 |
+
client = Client(host='http://localhost:11444')
|
248 |
+
# out = ollama.generate(**data)
|
249 |
+
out = client.generate(**data)
|
250 |
logger.debug(f"Response from model: {out}")
|
251 |
res = out["response"]
|
252 |
|
|
|
254 |
|
255 |
|
256 |
def process_query(user_query: str, history: ChatMessageHistory, tools):
|
257 |
+
# Add vehicle status to the history
|
258 |
+
user_query_status = (
|
259 |
+
f"Given that:\n{vehicle_status()[0]}\nAnswer the following:\n{user_query}"
|
260 |
+
)
|
261 |
+
history.add_message(HumanMessage(content=user_query_status))
|
262 |
for depth in range(10):
|
263 |
+
out = run_inference_step(depth, history, tools, schema_json)
|
264 |
print(f"Inference step result:\n{out}\n------------------\n")
|
265 |
history.add_message(AIMessage(content=out))
|
266 |
to_continue, tool_calls, errors = process_response(
|
267 |
user_query, out, history, tools, depth
|
268 |
)
|
269 |
if errors:
|
270 |
+
history.add_message(AIMessage(content=f"Errors in tool calls: {errors}"))
|
|
|
|
|
271 |
|
272 |
if not to_continue:
|
273 |
print(f"This is the answer, no more iterations: {out}")
|
kitt/skills/__init__.py
CHANGED
@@ -6,6 +6,7 @@ from .weather import get_weather_current_location, get_weather, get_forecast
|
|
6 |
from .routing import find_route
|
7 |
from .poi import search_points_of_interests, search_along_route_w_coordinates
|
8 |
from .vehicle import vehicle_status
|
|
|
9 |
|
10 |
|
11 |
|
|
|
6 |
from .routing import find_route
|
7 |
from .poi import search_points_of_interests, search_along_route_w_coordinates
|
8 |
from .vehicle import vehicle_status
|
9 |
+
from .interpreter import code_interpreter
|
10 |
|
11 |
|
12 |
|
kitt/skills/interpreter.py
ADDED
@@ -0,0 +1,52 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import inspect
|
2 |
+
|
3 |
+
# From https://github.com/NousResearch/Hermes-Function-Calling
|
4 |
+
|
5 |
+
def code_interpreter(code_markdown: str) -> dict | str:
|
6 |
+
"""
|
7 |
+
Execute the provided Python code string on the terminal using exec.
|
8 |
+
|
9 |
+
The string should contain valid, executable and pure Python code in markdown syntax.
|
10 |
+
Code should also import any required Python packages.
|
11 |
+
|
12 |
+
Args:
|
13 |
+
code_markdown (str): The Python code with markdown syntax to be executed.
|
14 |
+
For example: ```python\n<code-string>\n```
|
15 |
+
|
16 |
+
Returns:
|
17 |
+
dict | str: A dictionary containing variables declared and values returned by function calls,
|
18 |
+
or an error message if an exception occurred.
|
19 |
+
|
20 |
+
Note:
|
21 |
+
Use this function with caution, as executing arbitrary code can pose security risks. Use it only for numerical calculations.
|
22 |
+
"""
|
23 |
+
try:
|
24 |
+
# Extracting code from Markdown code block
|
25 |
+
code_lines = code_markdown.split('\n')[1:-1]
|
26 |
+
code_without_markdown = '\n'.join(code_lines)
|
27 |
+
|
28 |
+
# Create a new namespace for code execution
|
29 |
+
exec_namespace = {}
|
30 |
+
|
31 |
+
# Execute the code in the new namespace
|
32 |
+
exec(code_without_markdown, exec_namespace)
|
33 |
+
|
34 |
+
# Collect variables and function call results
|
35 |
+
result_dict = {}
|
36 |
+
for name, value in exec_namespace.items():
|
37 |
+
if callable(value):
|
38 |
+
try:
|
39 |
+
result_dict[name] = value()
|
40 |
+
except TypeError:
|
41 |
+
# If the function requires arguments, attempt to call it with arguments from the namespace
|
42 |
+
arg_names = inspect.getfullargspec(value).args
|
43 |
+
args = {arg_name: exec_namespace.get(arg_name) for arg_name in arg_names}
|
44 |
+
result_dict[name] = value(**args)
|
45 |
+
elif not name.startswith('_'): # Exclude variables starting with '_'
|
46 |
+
result_dict[name] = value
|
47 |
+
|
48 |
+
return result_dict
|
49 |
+
|
50 |
+
except Exception as e:
|
51 |
+
error_message = f"An error occurred: {e}"
|
52 |
+
return error_message
|
kitt/skills/routing.py
CHANGED
@@ -90,10 +90,41 @@ def find_route_tomtom(
|
|
90 |
}, response
|
91 |
|
92 |
|
93 |
-
def
|
94 |
-
"""
|
95 |
-
|
96 |
-
:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
97 |
"""
|
98 |
if not destination:
|
99 |
destination = vehicle.destination
|
@@ -114,7 +145,13 @@ def find_route(destination=""):
|
|
114 |
trip_info, raw_response = find_route_tomtom(
|
115 |
lat_depart, lon_depart, lat_dest, lon_dest, departure_time
|
116 |
)
|
|
|
|
|
|
|
|
|
|
|
117 |
|
|
|
118 |
distance, duration, arrival_time = (
|
119 |
trip_info["distance_m"],
|
120 |
trip_info["duration_s"],
|
@@ -138,5 +175,4 @@ def find_route(destination=""):
|
|
138 |
arrival_hour_display = arrival_time.strftime("%H:%M")
|
139 |
|
140 |
# return the distance and time
|
141 |
-
return f"The route to {destination} is {distance_km:.2f} km which takes {time_display}. Leaving now, the arrival time is estimated at {arrival_hour_display}."
|
142 |
-
# raw_response["routes"][0]["legs"][0]["points"]
|
|
|
90 |
}, response
|
91 |
|
92 |
|
93 |
+
def find_route_a_to_b(origin="", destination=""):
|
94 |
+
"""Get a route between origin and destination.
|
95 |
+
|
96 |
+
Args:
|
97 |
+
origin (string): Optional. The origin name.
|
98 |
+
destination (string): Optional. The destination name.
|
99 |
+
"""
|
100 |
+
if not destination:
|
101 |
+
destination = vehicle.destination
|
102 |
+
lat_dest, lon_dest = find_coordinates(destination)
|
103 |
+
print(f"lat_dest: {lat_dest}, lon_dest: {lon_dest}")
|
104 |
+
|
105 |
+
if not origin:
|
106 |
+
# Extract the latitude and longitude of the vehicle
|
107 |
+
vehicle_coordinates = getattr(vehicle, "location_coordinates")
|
108 |
+
lat_depart, lon_depart = vehicle_coordinates
|
109 |
+
else:
|
110 |
+
lat_depart, lon_depart = find_coordinates(origin)
|
111 |
+
print(f"lat_depart: {lat_depart}, lon_depart: {lon_depart}")
|
112 |
+
|
113 |
+
date = getattr(vehicle, "date")
|
114 |
+
time = getattr(vehicle, "time")
|
115 |
+
departure_time = f"{date}T{time}"
|
116 |
+
|
117 |
+
trip_info, raw_response = find_route_tomtom(
|
118 |
+
lat_depart, lon_depart, lat_dest, lon_dest, departure_time
|
119 |
+
)
|
120 |
+
return _format_tomtom_trip_info(trip_info, destination)
|
121 |
+
|
122 |
+
|
123 |
+
def find_route(destination):
|
124 |
+
"""Get a route to a destination from the current location of the vehicle.
|
125 |
+
|
126 |
+
Args:
|
127 |
+
destination (string): Optional. The destination name.
|
128 |
"""
|
129 |
if not destination:
|
130 |
destination = vehicle.destination
|
|
|
145 |
trip_info, raw_response = find_route_tomtom(
|
146 |
lat_depart, lon_depart, lat_dest, lon_dest, departure_time
|
147 |
)
|
148 |
+
return _format_tomtom_trip_info(trip_info, destination)
|
149 |
+
|
150 |
+
|
151 |
+
# raw_response["routes"][0]["legs"][0]["points"]
|
152 |
+
|
153 |
|
154 |
+
def _format_tomtom_trip_info(trip_info, destination="destination"):
|
155 |
distance, duration, arrival_time = (
|
156 |
trip_info["distance_m"],
|
157 |
trip_info["duration_s"],
|
|
|
175 |
arrival_hour_display = arrival_time.strftime("%H:%M")
|
176 |
|
177 |
# return the distance and time
|
178 |
+
return f"The route to {destination} is {distance_km:.2f} km which takes {time_display}. Leaving now, the arrival time is estimated at {arrival_hour_display}."
|
|
kitt/skills/vehicle.py
CHANGED
@@ -1,13 +1,9 @@
|
|
1 |
from .common import vehicle
|
2 |
|
3 |
|
4 |
-
STATUS_TEMPLATE = """
|
5 |
-
The current
|
6 |
-
The current
|
7 |
-
The current time: {time}
|
8 |
-
The current date: {date}
|
9 |
-
The current destination is: {destination}
|
10 |
-
"""
|
11 |
|
12 |
|
13 |
def vehicle_status() -> tuple[str, dict[str, str]]:
|
|
|
1 |
from .common import vehicle
|
2 |
|
3 |
|
4 |
+
STATUS_TEMPLATE = """The current location is: {location} ({lat}, {lon})
|
5 |
+
The current date and time: {date} {time}
|
6 |
+
The current destination is: {destination}"""
|
|
|
|
|
|
|
|
|
7 |
|
8 |
|
9 |
def vehicle_status() -> tuple[str, dict[str, str]]:
|
main.py
CHANGED
@@ -26,6 +26,7 @@ from kitt.skills import (
|
|
26 |
do_anything_else,
|
27 |
date_time_info,
|
28 |
get_weather_current_location,
|
|
|
29 |
)
|
30 |
from kitt.skills import extract_func_args
|
31 |
from kitt.core import voice_options, tts_gradio
|
@@ -124,6 +125,7 @@ tools = [
|
|
124 |
StructuredTool.from_function(search_along_route),
|
125 |
StructuredTool.from_function(date_time_info),
|
126 |
StructuredTool.from_function(get_weather_current_location),
|
|
|
127 |
# StructuredTool.from_function(do_anything_else),
|
128 |
]
|
129 |
|
@@ -201,6 +203,8 @@ def run_model(query, voice_character, state):
|
|
201 |
return run_nexusraven_model(query, voice_character)
|
202 |
elif model == "llama3":
|
203 |
return run_llama3_model(query, voice_character)
|
|
|
|
|
204 |
|
205 |
|
206 |
def calculate_route_gradio(origin, destination):
|
@@ -259,12 +263,19 @@ def save_and_transcribe_audio(audio):
|
|
259 |
y = y.astype(np.float32)
|
260 |
y /= np.max(np.abs(y))
|
261 |
text = transcriber({"sampling_rate": sr, "raw": y})["text"]
|
|
|
|
|
262 |
except Exception as e:
|
263 |
print(f"Error: {e}")
|
264 |
-
return "Error transcribing audio"
|
265 |
return text
|
266 |
|
267 |
|
|
|
|
|
|
|
|
|
|
|
268 |
# to be able to use the microphone on chrome, you will have to go to chrome://flags/#unsafely-treat-insecure-origin-as-secure and enter http://10.186.115.21:7860/
|
269 |
# in "Insecure origins treated as secure", enable it and relaunch chrome
|
270 |
|
@@ -337,6 +348,18 @@ def create_demo(tts_server: bool = False, model="llama3", tts=True):
|
|
337 |
input_text = gr.Textbox(
|
338 |
value="How is the weather?", label="Input text", interactive=True
|
339 |
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
340 |
vehicle_status = gr.JSON(
|
341 |
value=vehicle.model_dump_json(), label="Vehicle status"
|
342 |
)
|
@@ -370,6 +393,11 @@ def create_demo(tts_server: bool = False, model="llama3", tts=True):
|
|
370 |
inputs=[input_text, voice_character, state],
|
371 |
outputs=[output_text, output_audio],
|
372 |
)
|
|
|
|
|
|
|
|
|
|
|
373 |
|
374 |
# Set the vehicle status based on the trip progress
|
375 |
trip_progress.release(
|
@@ -380,7 +408,10 @@ def create_demo(tts_server: bool = False, model="llama3", tts=True):
|
|
380 |
|
381 |
# Save and transcribe the audio
|
382 |
input_audio.stop_recording(
|
383 |
-
fn=
|
|
|
|
|
|
|
384 |
)
|
385 |
|
386 |
# Clear the history
|
|
|
26 |
do_anything_else,
|
27 |
date_time_info,
|
28 |
get_weather_current_location,
|
29 |
+
code_interpreter,
|
30 |
)
|
31 |
from kitt.skills import extract_func_args
|
32 |
from kitt.core import voice_options, tts_gradio
|
|
|
125 |
StructuredTool.from_function(search_along_route),
|
126 |
StructuredTool.from_function(date_time_info),
|
127 |
StructuredTool.from_function(get_weather_current_location),
|
128 |
+
StructuredTool.from_function(code_interpreter),
|
129 |
# StructuredTool.from_function(do_anything_else),
|
130 |
]
|
131 |
|
|
|
203 |
return run_nexusraven_model(query, voice_character)
|
204 |
elif model == "llama3":
|
205 |
return run_llama3_model(query, voice_character)
|
206 |
+
return "Error running model", None
|
207 |
+
|
208 |
|
209 |
|
210 |
def calculate_route_gradio(origin, destination):
|
|
|
263 |
y = y.astype(np.float32)
|
264 |
y /= np.max(np.abs(y))
|
265 |
text = transcriber({"sampling_rate": sr, "raw": y})["text"]
|
266 |
+
gr.Info(f"Transcribed text is: {text}\nProcessing the input...")
|
267 |
+
|
268 |
except Exception as e:
|
269 |
print(f"Error: {e}")
|
270 |
+
return "Error transcribing audio."
|
271 |
return text
|
272 |
|
273 |
|
274 |
+
def save_and_transcribe_run_model(audio, voice_character, state):
|
275 |
+
text = save_and_transcribe_audio(audio)
|
276 |
+
out_text, out_voice = run_model(text, voice_character, state)
|
277 |
+
return text, out_text, out_voice
|
278 |
+
|
279 |
# to be able to use the microphone on chrome, you will have to go to chrome://flags/#unsafely-treat-insecure-origin-as-secure and enter http://10.186.115.21:7860/
|
280 |
# in "Insecure origins treated as secure", enable it and relaunch chrome
|
281 |
|
|
|
348 |
input_text = gr.Textbox(
|
349 |
value="How is the weather?", label="Input text", interactive=True
|
350 |
)
|
351 |
+
with gr.Accordion("Debug"):
|
352 |
+
input_audio_debug = gr.Audio(
|
353 |
+
type="numpy",
|
354 |
+
sources=["microphone"],
|
355 |
+
label="Input audio",
|
356 |
+
elem_id="input_audio",
|
357 |
+
)
|
358 |
+
input_text_debug = gr.Textbox(
|
359 |
+
value="How is the weather?",
|
360 |
+
label="Input text",
|
361 |
+
interactive=True,
|
362 |
+
)
|
363 |
vehicle_status = gr.JSON(
|
364 |
value=vehicle.model_dump_json(), label="Vehicle status"
|
365 |
)
|
|
|
393 |
inputs=[input_text, voice_character, state],
|
394 |
outputs=[output_text, output_audio],
|
395 |
)
|
396 |
+
input_text_debug.submit(
|
397 |
+
fn=run_model,
|
398 |
+
inputs=[input_text, voice_character, state],
|
399 |
+
outputs=[output_text, output_audio],
|
400 |
+
)
|
401 |
|
402 |
# Set the vehicle status based on the trip progress
|
403 |
trip_progress.release(
|
|
|
408 |
|
409 |
# Save and transcribe the audio
|
410 |
input_audio.stop_recording(
|
411 |
+
fn=save_and_transcribe_run_model, inputs=[input_audio, voice_character, state], outputs=[input_text, output_text, output_audio]
|
412 |
+
)
|
413 |
+
input_audio_debug.stop_recording(
|
414 |
+
fn=save_and_transcribe_audio, inputs=[input_audio_debug], outputs=[input_text_debug]
|
415 |
)
|
416 |
|
417 |
# Clear the history
|