Skip to content

Commit

Permalink
Update Agent notebook (#2186)
Browse files Browse the repository at this point in the history 
1. replace the example tool with offline version
2. change quantization ratio from 0.8 to 0.72 for better accuracy on
GPU.
  • Loading branch information
@openvino-dev-samples
openvino-dev-samples authored Jul 11, 2024
1 parent 717f7a8 commit c20240f
Showing 1 changed file with 38 additions and 130 deletions.
168 changes: 38 additions & 130 deletions notebooks/llm-agent-functioncall/llm-agent-functioncall-qwen.ipynb
View file
Original file line number Diff line number Diff line change
Expand Up @@ -45,75 +45,17 @@
},
{
"cell_type": "code",
"execution_count": 17,
"execution_count": 16,
"id": "47d43de7-9946-482d-84cb-222294c1cda8",
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...\n",
"To disable this warning, you can either:\n",
"\t- Avoid using `tokenizers` before the fork if possible\n",
"\t- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Note: you may need to restart the kernel to use updated packages.\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...\n",
"To disable this warning, you can either:\n",
"\t- Avoid using `tokenizers` before the fork if possible\n",
"\t- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Note: you may need to restart the kernel to use updated packages.\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...\n",
"To disable this warning, you can either:\n",
"\t- Avoid using `tokenizers` before the fork if possible\n",
"\t- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Note: you may need to restart the kernel to use updated packages.\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...\n",
"To disable this warning, you can either:\n",
"\t- Avoid using `tokenizers` before the fork if possible\n",
"\t- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Note: you may need to restart the kernel to use updated packages.\n",
"Note: you may need to restart the kernel to use updated packages.\n",
"Note: you may need to restart the kernel to use updated packages.\n",
"\u001b[33mWARNING: typer 0.12.3 does not provide the extra 'all'\u001b[0m\u001b[33m\n",
"\u001b[0mNote: you may need to restart the kernel to use updated packages.\n"
]
Expand Down Expand Up @@ -141,32 +83,7 @@
"execution_count": 1,
"id": "2ec33075",
"metadata": {},
"outputs": [
{
"data": {
"application/vnd.jupyter.widget-view+json": {
"model_id": "d15bea5cc94949328b44d9a2ed0efaa0",
"version_major": 2,
"version_minor": 0
},
"text/plain": [
"openvino_log.png: 0.00B [00:00, ?B/s]"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"PosixPath('/home/ethan/intel/openvino_notebooks/notebooks/llm-agent-qwen/openvino_log.png')"
]
},
"execution_count": 1,
"metadata": {},
"output_type": "execute_result"
}
],
"outputs": [],
"source": [
"import requests\n",
"from PIL import Image\n",
Expand Down Expand Up @@ -212,28 +129,19 @@
"metadata": {},
"outputs": [],
"source": [
"import requests\n",
"\n",
"import json\n",
"\n",
"def get_current_weather(city_name):\n",
" \"\"\"Get the current weather in a given city name\"\"\"\n",
" if not isinstance(city_name, str):\n",
" raise TypeError(\"City name must be a string\")\n",
" key_selection = {\n",
" \"current_condition\": [\n",
" \"temp_C\",\n",
" \"FeelsLikeC\",\n",
" \"humidity\",\n",
" \"weatherDesc\",\n",
" \"observation_time\",\n",
" ],\n",
" }\n",
" resp = requests.get(f\"https://wttr.in/{city_name}?format=j1\")\n",
" resp.raise_for_status()\n",
" resp = resp.json()\n",
" ret = {k: {_v: resp[k][0][_v] for _v in v} for k, v in key_selection.items()}\n",
"\n",
" return str(ret)"
"def get_current_weather(location, unit=\"fahrenheit\"):\n",
" \"\"\"Get the current weather in a given location\"\"\"\n",
" if \"tokyo\" in location.lower():\n",
" return json.dumps({\"location\": \"Tokyo\", \"temperature\": \"10\", \"unit\": \"celsius\"})\n",
" elif \"san francisco\" in location.lower():\n",
" return json.dumps({\"location\": \"San Francisco\", \"temperature\": \"72\", \"unit\": \"fahrenheit\"})\n",
" elif \"paris\" in location.lower():\n",
" return json.dumps({\"location\": \"Paris\", \"temperature\": \"22\", \"unit\": \"celsius\"})\n",
" else:\n",
" return json.dumps({\"location\": location, \"temperature\": \"unknown\"})"
]
},
{
Expand All @@ -246,24 +154,25 @@
},
{
"cell_type": "code",
"execution_count": 3,
"execution_count": 9,
"id": "5ea4ce13",
"metadata": {},
"outputs": [],
"source": [
"functions = [\n",
" {\n",
" \"name\": \"get_current_weather\",\n",
" \"description\": \"Get the current weather in a given city name\",\n",
" \"description\": \"Get the current weather in a given location\",\n",
" \"parameters\": {\n",
" \"type\": \"object\",\n",
" \"properties\": {\n",
" \"city_name\": {\n",
" \"location\": {\n",
" \"type\": \"string\",\n",
" \"description\": \"The city and state, e.g. San Francisco, CA\",\n",
" },\n",
" \"unit\": {\"type\": \"string\", \"enum\": [\"celsius\", \"fahrenheit\"]},\n",
" },\n",
" \"required\": [\"city_name\"],\n",
" \"required\": [\"location\"],\n",
" },\n",
" }\n",
"]"
Expand All @@ -288,7 +197,7 @@
},
{
"cell_type": "code",
"execution_count": 4,
"execution_count": 10,
"id": "981df8fe-cfcf-455a-919e-dda36f3b5dfb",
"metadata": {
"test_replace": {
Expand All @@ -303,7 +212,7 @@
"model_path = \"Qwen2-7B-Instruct-ov\"\n",
"\n",
"if not Path(model_path).exists():\n",
" !optimum-cli export openvino --model {model_id} --task text-generation-with-past --trust-remote-code --weight-format int4 {model_path}"
" !optimum-cli export openvino --model {model_id} --task text-generation-with-past --trust-remote-code --weight-format int4 --ratio 0.72 {model_path}"
]
},
{
Expand All @@ -326,7 +235,7 @@
{
"data": {
"application/vnd.jupyter.widget-view+json": {
"model_id": "851fc266add14641961d0483cc65718c",
"model_id": "f981d582698c4400ac8997e9108c5558",
"version_major": 2,
"version_minor": 0
},
Expand Down Expand Up @@ -413,7 +322,7 @@
},
{
"cell_type": "code",
"execution_count": null,
"execution_count": 7,
"id": "ee406254",
"metadata": {},
"outputs": [],
Expand Down Expand Up @@ -464,7 +373,7 @@
},
{
"cell_type": "code",
"execution_count": 7,
"execution_count": 14,
"id": "4799540b-eee0-491f-a5b6-5bae68c22af9",
"metadata": {},
"outputs": [
Expand All @@ -475,17 +384,15 @@
"# User question:\n",
"[{'role': 'user', 'content': \"What's the weather like in San Francisco?\"}]\n",
"# Assistant Response 1:\n",
"{'role': 'assistant', 'content': '', 'function_call': {'name': 'get_current_weather', 'arguments': '{\"city_name\": \"San Francisco, CA\"}'}}\n",
"[{'role': 'assistant', 'content': '', 'function_call': {'name': 'get_current_weather', 'arguments': '{\"location\": \"San Francisco, CA\"}'}}]\n",
"# Function Response:\n",
"{'current_condition': {'temp_C': '18', 'FeelsLikeC': '18', 'humidity': '72', 'weatherDesc': [{'value': 'Clear'}], 'observation_time': '07:58 AM'}}\n",
"{\"location\": \"San Francisco\", \"temperature\": \"72\", \"unit\": \"fahrenheit\"}\n",
"# Assistant Response 2:\n",
"{'role': 'assistant', 'content': 'The current weather in San Francisco is clear with a temperature of 18°C and a humidity level of 72%. The time of this observation was 07:58 AM.'}\n"
"[{'role': 'assistant', 'content': 'The current weather in San Francisco is 72 degrees Fahrenheit.'}]\n"
]
}
],
"source": [
"import json\n",
"\n",
"print(\"# User question:\")\n",
"messages = [{\"role\": \"user\", \"content\": \"What's the weather like in San Francisco?\"}]\n",
"print(messages)\n",
Expand All @@ -494,14 +401,14 @@
"responses = []\n",
"\n",
"# Step 1: Role `user` sending the request\n",
"for responses in llm.chat(\n",
"responses = llm.chat(\n",
" messages=messages,\n",
" functions=functions,\n",
" stream=False,\n",
"):\n",
" print(responses)\n",
")\n",
"print(responses)\n",
"\n",
"messages.append(responses)\n",
"messages.extend(responses)\n",
"\n",
"# Step 2: check if the model wanted to call a function, and call the function if needed\n",
"last_response = messages[-1]\n",
Expand All @@ -513,7 +420,7 @@
" function_to_call = available_functions[function_name]\n",
" function_args = json.loads(last_response[\"function_call\"][\"arguments\"])\n",
" function_response = function_to_call(\n",
" city_name=function_args.get(\"city_name\"),\n",
" location=function_args.get(\"location\"),\n",
" )\n",
" print(\"# Function Response:\")\n",
" print(function_response)\n",
Expand All @@ -529,12 +436,12 @@
"\n",
" print(\"# Assistant Response 2:\")\n",
" # Step 4: Consolidate the observation from function into final response\n",
" for responses in llm.chat(\n",
" responses = llm.chat(\n",
" messages=messages,\n",
" functions=functions,\n",
" stream=False,\n",
" ): # get a new response from the model where it can see the function response\n",
" print(responses)"
" )\n",
" print(responses)"
]
},
{
Expand Down Expand Up @@ -580,6 +487,7 @@
"source": [
"import urllib.parse\n",
"import json5\n",
"import requests\n",
"from qwen_agent.tools.base import BaseTool, register_tool\n",
"\n",
"\n",
Expand Down

0 comments on commit c20240f

Please sign in to comment.