Instructions to use DiTy/gemma-2-9b-it-function-calling-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use DiTy/gemma-2-9b-it-function-calling-GGUF with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="DiTy/gemma-2-9b-it-function-calling-GGUF")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("DiTy/gemma-2-9b-it-function-calling-GGUF")
model = AutoModelForCausalLM.from_pretrained("DiTy/gemma-2-9b-it-function-calling-GGUF", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use DiTy/gemma-2-9b-it-function-calling-GGUF with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf DiTy/gemma-2-9b-it-function-calling-GGUF:F16
# Run inference directly in the terminal:
llama cli -hf DiTy/gemma-2-9b-it-function-calling-GGUF:F16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf DiTy/gemma-2-9b-it-function-calling-GGUF:F16
# Run inference directly in the terminal:
llama cli -hf DiTy/gemma-2-9b-it-function-calling-GGUF:F16

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf DiTy/gemma-2-9b-it-function-calling-GGUF:F16
# Run inference directly in the terminal:
./llama-cli -hf DiTy/gemma-2-9b-it-function-calling-GGUF:F16

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf DiTy/gemma-2-9b-it-function-calling-GGUF:F16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf DiTy/gemma-2-9b-it-function-calling-GGUF:F16

Use Docker

docker model run hf.co/DiTy/gemma-2-9b-it-function-calling-GGUF:F16

LM Studio
Jan

vLLM

How to use DiTy/gemma-2-9b-it-function-calling-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "DiTy/gemma-2-9b-it-function-calling-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "DiTy/gemma-2-9b-it-function-calling-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/DiTy/gemma-2-9b-it-function-calling-GGUF:F16

SGLang

How to use DiTy/gemma-2-9b-it-function-calling-GGUF with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "DiTy/gemma-2-9b-it-function-calling-GGUF" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "DiTy/gemma-2-9b-it-function-calling-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "DiTy/gemma-2-9b-it-function-calling-GGUF" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "DiTy/gemma-2-9b-it-function-calling-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use DiTy/gemma-2-9b-it-function-calling-GGUF with Ollama:
```
ollama run hf.co/DiTy/gemma-2-9b-it-function-calling-GGUF:F16
```
Unsloth Desktop

How to use DiTy/gemma-2-9b-it-function-calling-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf DiTy/gemma-2-9b-it-function-calling-GGUF:F16

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "DiTy/gemma-2-9b-it-function-calling-GGUF:F16"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use DiTy/gemma-2-9b-it-function-calling-GGUF with Docker Model Runner:
```
docker model run hf.co/DiTy/gemma-2-9b-it-function-calling-GGUF:F16
```

Lemonade

How to use DiTy/gemma-2-9b-it-function-calling-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull DiTy/gemma-2-9b-it-function-calling-GGUF:F16

Run and chat with the model

lemonade run user.gemma-2-9b-it-function-calling-GGUF-F16

List all available models

lemonade list

Hermes Agent

How to use DiTy/gemma-2-9b-it-function-calling-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf DiTy/gemma-2-9b-it-function-calling-GGUF:F16

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default DiTy/gemma-2-9b-it-function-calling-GGUF:F16

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use DiTy/gemma-2-9b-it-function-calling-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf DiTy/gemma-2-9b-it-function-calling-GGUF:F16

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "DiTy/gemma-2-9b-it-function-calling-GGUF:F16" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

DiTy commited on Sep 6, 2024

Commit

b16e16c

verified ·

1 Parent(s): 6ad5274

Fix chat template with system message

Browse files

Files changed (2) hide show

README.md +7 -7
tokenizer_config.json +1 -1

README.md CHANGED Viewed

@@ -1,17 +1,17 @@
 ---
-library_name: transformers
-tags:
-- conversational
-- gemma2
-- function-calling
-- trl
-license: apache-2.0
 base_model: google/gemma-2-9b-it
 datasets:
 - DiTy/function-calling
 language:
 - en
 pipeline_tag: text-generation
 ---
 # DiTy/gemma-2-9b-it-function-calling-GGUF

 ---
 base_model: google/gemma-2-9b-it
 datasets:
 - DiTy/function-calling
 language:
 - en
+library_name: transformers
+license: apache-2.0
 pipeline_tag: text-generation
+tags:
+- conversational
+- gemma2
+- function-calling
+- trl
 ---
 # DiTy/gemma-2-9b-it-function-calling-GGUF

tokenizer_config.json CHANGED Viewed

@@ -2000,7 +2000,7 @@
     "<end_of_turn>"
   ],
   "bos_token": "<bos>",
-  "chat_template": "\n{%- macro render_json(d, indent=4) -%}\n{%- if d is string %}\n{{ '\"' + d + '\"' }}\n{%- elif d is mapping %}\n{%- for key, value in d.items() %}\n{%- if value is string %}\n{{ \" \" * indent + '\"' + key + '\": \"' + value + '\"' }}\n{%- elif value is mapping %}\n{{ \" \" * indent + '\"' + key + '\": {' }}\n{{ render_json(value, indent + 4) }}\n{{ \" \" * indent + \"}\" }}\n{%- elif value is sequence %}\n{{ \" \" * indent + '\"' + key + '\": [\n' }}\n{%- for item in value %}\n{{- \" \" * (indent + 4) + render_json(item, indent + 4) }}\n{%- if not loop.last %},\n{% endif %}\n{%- endfor %}\n{{ '\n' + \" \" * indent + \"]\" }}\n{%- else %}\n{{ \" \" * indent + '\"' + key + '\": ' + value|string }}\n{%- endif %}\n{%- if not loop.last %},\n{% endif %}\n{%- endfor %}\n{%- elif d is sequence %}\n{%- for item in d %}\n{{ \" \" * indent + render_json(item, indent + 4) }}\n{%- if not loop.last %},\n{% endif %}\n{%- endfor %}\n{%- else %}\n{{ \" \" * indent + d|string }}\n{%- endif %}\n{%- endmacro %}\n\n\n{#- This block checks system prompt and list of tools #}\n{%- if not add_generation_prompt is defined %}\n    {%- set add_generation_prompt = false %}\n{%- endif %}\n{%- if messages[0][\"role\"] == \"system\" %}\n    {%- set system_message = messages[0][\"content\"] %}\n    {%- set loop_messages = messages[1:] %}\n{%- else %}\n    {%- set loop_messages = messages %}\n    {%- set system_message = \"\" %}\n{%- endif %}\n{%- if not tools is defined %}\n    {%- set tools = none %}\n{%- endif %}\n{%- set user_messages = loop_messages | selectattr(\"role\", \"equalto\", \"user\") | list %}\n\n\n{#- This block checks for alternating user/model messages, skipping tool calling messages #}\n{%- set ns = namespace() %}\n{%- set ns.index = 0 %}\n{%- for message in loop_messages %}\n    {%- if not (message.role == \"tool\" or message.role == \"tool_results\" or (message.tool_calls is defined and message.tool_calls is not none)) %}\n        {%- if (message[\"role\"] == \"user\" or message[\"role\"] == \"function-response\") != (ns.index % 2 == 0) %}\n            {{- raise_exception(\"After the optional system message, conversation roles must alternate user/model/user/model/... NB: 'function-call' is same as the 'model' and 'function-response' is same as the 'user'!\") }}\n        {%- endif %}\n        {%- set ns.index = ns.index + 1 %}\n    {%- endif %}\n{%- endfor %}\n\n\n{#- This block format full dialogue messages #}\n{{- bos_token }}\n{{- \"<start_of_turn>user\" + '\n' + system_message}}\n{%- if tools is not none %}\n    {%- for tool in tools %}\n        {%- set tool = tool.function %}\n        {{- '{\n' }}\n        {{- render_json(tool, 4) }}\n        {%- if not loop.last %}\n            {{- \"\n},\n\" }}\n        {%- else %}\n            {{- \"\n}\n\" }}\n        {% endif %}\n    {%- endfor %}\n{%- else %}\n    {{- '\n\n' }}\n{%- endif %}\n{{ loop_messages[0]['content'] + '<end_of_turn>\n' }}\n{%- for message in loop_messages[1:] %}\n    {%- if message[\"role\"] == \"user\" %}\n        {{- \"<start_of_turn>user\" + '\n' + message['content'] + '<end_of_turn>\n' }}\n    {%- elif message[\"role\"] == \"model\" %}\n        {{- \"<start_of_turn>model\" + '\n' + message['content'] + '<end_of_turn>\n' }}\n    {%- elif message[\"role\"] == \"function-call\" %}\n        {{- \"<start_of_turn>model\" + '\n' + \"Function call: \" + message['content'] + '<end_of_turn>\n' }}\n    {%- elif message[\"role\"] == \"function-response\" %}\n        {{- \"<start_of_turn>user\" + '\n' + \"Function response: \" + message['content'] + '<end_of_turn>\n' }}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {%- if loop_messages[-1]['role'] == \"user\" or loop_messages[-1]['role'] == \"function-response\" %}\n{{ '<start_of_turn>model\n' }}\n    {%- else %}\n        {{- raise_exception(\"For add_generation_prompt=True, the last message role must be 'user' or 'function-response'.\") }}\n    {%- endif %}\n{%- endif %}\n",
   "clean_up_tokenization_spaces": false,
   "eos_token": "<eos>",
   "model_max_length": 1000000000000000019884624838656,

     "<end_of_turn>"
   ],
   "bos_token": "<bos>",
+  "chat_template": "\n{%- macro render_json(d, indent=4) -%}\n{%- if d is string %}\n{{ '\"' + d + '\"' }}\n{%- elif d is mapping %}\n{%- for key, value in d.items() %}\n{%- if value is string %}\n{{ \" \" * indent + '\"' + key + '\": \"' + value + '\"' }}\n{%- elif value is mapping %}\n{{ \" \" * indent + '\"' + key + '\": {' }}\n{{ render_json(value, indent + 4) }}\n{{ \" \" * indent + \"}\" }}\n{%- elif value is sequence %}\n{{ \" \" * indent + '\"' + key + '\": [\n' }}\n{%- for item in value %}\n{{- \" \" * (indent + 4) + render_json(item, indent + 4) }}\n{%- if not loop.last %},\n{% endif %}\n{%- endfor %}\n{{ '\n' + \" \" * indent + \"]\" }}\n{%- else %}\n{{ \" \" * indent + '\"' + key + '\": ' + value|string }}\n{%- endif %}\n{%- if not loop.last %},\n{% endif %}\n{%- endfor %}\n{%- elif d is sequence %}\n{%- for item in d %}\n{{ \" \" * indent + render_json(item, indent + 4) }}\n{%- if not loop.last %},\n{% endif %}\n{%- endfor %}\n{%- else %}\n{{ \" \" * indent + d|string }}\n{%- endif %}\n{%- endmacro %}\n\n\n{#- This block checks system prompt and list of tools #}\n{%- if not add_generation_prompt is defined %}\n    {%- set add_generation_prompt = false %}\n{%- endif %}\n{%- if messages[0][\"role\"] == \"system\" %}\n    {%- set system_message = messages[0][\"content\"] %}\n    {%- set loop_messages = messages[1:] %}\n{%- else %}\n    {%- set loop_messages = messages %}\n    {%- set system_message = \"\" %}\n{%- endif %}\n{%- if not tools is defined %}\n    {%- set tools = none %}\n{%- endif %}\n{%- set user_messages = loop_messages | selectattr(\"role\", \"equalto\", \"user\") | list %}\n\n\n{#- This block checks for alternating user/model messages, skipping tool calling messages #}\n{%- set ns = namespace() %}\n{%- set ns.index = 0 %}\n{%- for message in loop_messages %}\n    {%- if (message[\"role\"] == \"user\" or message[\"role\"] == \"function-response\") != (ns.index % 2 == 0) %}\n        {{- raise_exception(\"After the optional system message, conversation roles must alternate user/model/user/model/... NB: 'function-call' is same as the 'model' and 'function-response' is same as the 'user'!\") }}\n    {%- endif %}\n    {%- set ns.index = ns.index + 1 %}\n{%- endfor %}\n\n\n{#- This block format full dialogue messages #}\n{{- bos_token }}\n{{- \"<start_of_turn>user\" + '\n' }}\n{%- if system_message != \"\" %}\n    {{- system_message }}\n{%- endif %}\n{%- if tools is not none %}\n    {%- for tool in tools %}\n        {%- set tool = tool.function %}\n        {{- '{\n' }}\n        {{- render_json(tool, 4) }}\n        {%- if not loop.last %}\n            {{- \"\n},\n\" }}\n        {%- else %}\n            {{- \"\n}\n\" }}\n        {% endif %}\n    {%- endfor %}\n    {%- elif system_message != \"\" %}\n        {{- '\n\n' }}\n{%- endif %}\n{{ loop_messages[0]['content'] + '<end_of_turn>\n' }}\n{%- for message in loop_messages[1:] %}\n    {%- if message[\"role\"] == \"user\" %}\n        {{- \"<start_of_turn>user\" + '\n' + message['content'] + '<end_of_turn>\n' }}\n    {%- elif message[\"role\"] == \"model\" %}\n        {{- \"<start_of_turn>model\" + '\n' + message['content'] + '<end_of_turn>\n' }}\n    {%- elif message[\"role\"] == \"function-call\" %}\n        {{- \"<start_of_turn>model\" + '\n' + \"Function call: \" + message['content'] + '<end_of_turn>\n' }}\n    {%- elif message[\"role\"] == \"function-response\" %}\n        {{- \"<start_of_turn>user\" + '\n' + \"Function response: \" + message['content'] + '<end_of_turn>\n' }}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {%- if loop_messages[-1]['role'] == \"user\" or loop_messages[-1]['role'] == \"function-response\" %}\n{{ '<start_of_turn>model\n' }}\n    {%- else %}\n        {{- raise_exception(\"For add_generation_prompt=True, the last message role must be 'user' or 'function-response'.\") }}\n    {%- endif %}\n{%- endif %}\n\n",
   "clean_up_tokenization_spaces": false,
   "eos_token": "<eos>",
   "model_max_length": 1000000000000000019884624838656,