Text Generation
Transformers
Safetensors
English
Korean
llama
causal-lm
sft
dpo
grpo
long-context
keti
conversational
text-generation-inference
Instructions to use KETI-NLP/keti-llama-7b-v0.1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use KETI-NLP/keti-llama-7b-v0.1 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="KETI-NLP/keti-llama-7b-v0.1") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("KETI-NLP/keti-llama-7b-v0.1") model = AutoModelForCausalLM.from_pretrained("KETI-NLP/keti-llama-7b-v0.1", 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
- vLLM
How to use KETI-NLP/keti-llama-7b-v0.1 with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "KETI-NLP/keti-llama-7b-v0.1" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "KETI-NLP/keti-llama-7b-v0.1", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/KETI-NLP/keti-llama-7b-v0.1
- SGLang
How to use KETI-NLP/keti-llama-7b-v0.1 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 "KETI-NLP/keti-llama-7b-v0.1" \ --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": "KETI-NLP/keti-llama-7b-v0.1", "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 "KETI-NLP/keti-llama-7b-v0.1" \ --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": "KETI-NLP/keti-llama-7b-v0.1", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use KETI-NLP/keti-llama-7b-v0.1 with Docker Model Runner:
docker model run hf.co/KETI-NLP/keti-llama-7b-v0.1
| {{- bos_token if bos_token is defined else '' }} | |
| {%- if custom_tools is defined and custom_tools is not none %} | |
| {%- set tools = custom_tools %} | |
| {%- endif %} | |
| {%- if tools is not defined %} | |
| {%- set tools = none %} | |
| {%- endif %} | |
| {%- if tools is not none and tools is iterable and tools is not mapping and tools|length == 0 %} | |
| {%- set tools = none %} | |
| {%- endif %} | |
| {%- if tools_in_user_message is not defined %} | |
| {#- Llama 3.1-style tool prompting is most stable when tools are injected into the first user turn. #} | |
| {%- set tools_in_user_message = true %} | |
| {%- endif %} | |
| {%- if date_string is not defined %} | |
| {%- if strftime_now is defined %} | |
| {%- set date_string = strftime_now("%d %b %Y") %} | |
| {%- else %} | |
| {#- Avoid baking a stale Today Date into newer data. Pass date_string explicitly if needed. #} | |
| {%- set date_string = none %} | |
| {%- endif %} | |
| {%- endif %} | |
| {%- if cutoff_date_string is not defined %} | |
| {%- set cutoff_date_string = "December 2023" %} | |
| {%- endif %} | |
| {%- if image_token is not defined %} | |
| {%- set image_token = "<|image|>" %} | |
| {%- endif %} | |
| {%- if video_token is not defined %} | |
| {%- set video_token = "<|video|>" %} | |
| {%- endif %} | |
| {#- Modern content renderer inspired by Qwen-style templates. | |
| Supports: | |
| - string content | |
| - None / missing content | |
| - list content blocks: {type: text, text: ...}, {type: image/image_url/input_image, ...}, {type: video/input_video, ...} | |
| - direct {text: ...}, {image: ...}, {image_url: ...}, {video: ...} | |
| #} | |
| {%- set image_count = namespace(value=0) %} | |
| {%- set video_count = namespace(value=0) %} | |
| {%- macro render_content_item(item, do_vision_count, is_system_content=false) %} | |
| {%- if item is string %} | |
| {{- item }} | |
| {%- elif item is mapping %} | |
| {%- set item_type = item.get('type', '') %} | |
| {%- if item_type in ['image', 'image_url', 'input_image'] or 'image' in item or 'image_url' in item %} | |
| {%- if is_system_content %} | |
| {{- raise_exception('System message cannot contain images.') }} | |
| {%- endif %} | |
| {%- if do_vision_count %} | |
| {%- set image_count.value = image_count.value + 1 %} | |
| {%- endif %} | |
| {%- if add_vision_id is defined and add_vision_id %} | |
| {{- 'Picture ' ~ image_count.value ~ ': ' }} | |
| {%- endif %} | |
| {{- image_token }} | |
| {%- elif item_type in ['video', 'input_video'] or 'video' in item %} | |
| {%- if is_system_content %} | |
| {{- raise_exception('System message cannot contain videos.') }} | |
| {%- endif %} | |
| {%- if do_vision_count %} | |
| {%- set video_count.value = video_count.value + 1 %} | |
| {%- endif %} | |
| {%- if add_vision_id is defined and add_vision_id %} | |
| {{- 'Video ' ~ video_count.value ~ ': ' }} | |
| {%- endif %} | |
| {{- video_token }} | |
| {%- elif 'text' in item %} | |
| {{- item.text }} | |
| {%- elif item_type in ['text', 'input_text'] and 'content' in item %} | |
| {{- item.content }} | |
| {%- elif 'content' in item %} | |
| {{- render_content(item.content, do_vision_count, is_system_content) }} | |
| {%- else %} | |
| {{- raise_exception('Unexpected item type in content.') }} | |
| {%- endif %} | |
| {%- elif item is none or item is undefined %} | |
| {{- '' }} | |
| {%- else %} | |
| {{- raise_exception('Unexpected item type in content.') }} | |
| {%- endif %} | |
| {%- endmacro %} | |
| {%- macro render_content(content, do_vision_count, is_system_content=false) %} | |
| {%- if content is string %} | |
| {{- content }} | |
| {%- elif content is none or content is undefined %} | |
| {{- '' }} | |
| {%- elif content is mapping %} | |
| {{- render_content_item(content, do_vision_count, is_system_content) }} | |
| {%- elif content is iterable %} | |
| {%- for item in content %} | |
| {{- render_content_item(item, do_vision_count, is_system_content) }} | |
| {%- endfor %} | |
| {%- else %} | |
| {{- raise_exception('Unexpected content type.') }} | |
| {%- endif %} | |
| {%- endmacro %} | |
| {%- macro render_tool_call_json(tool_call) %} | |
| {%- if tool_call.function is defined %} | |
| {%- set fn = tool_call.function %} | |
| {%- else %} | |
| {%- set fn = tool_call %} | |
| {%- endif %} | |
| {%- if fn.name is not defined %} | |
| {{- raise_exception('Tool call is missing function name.') }} | |
| {%- endif %} | |
| {%- set args = fn.arguments if fn.arguments is defined and fn.arguments is not none else {} %} | |
| {{- '{"name": ' }}{{- fn.name | tojson }}{{- ', "parameters": ' }} | |
| {%- if args is string and args|trim %} | |
| {{- args }} | |
| {%- else %} | |
| {{- args | tojson }} | |
| {%- endif %} | |
| {{- '}' }} | |
| {%- endmacro %} | |
| {%- macro render_tool_calls_json(tool_calls) %} | |
| {%- if tool_calls|length == 1 %} | |
| {{- render_tool_call_json(tool_calls[0]) }} | |
| {%- else %} | |
| {{- '[' }} | |
| {%- for tool_call in tool_calls %} | |
| {{- render_tool_call_json(tool_call) }} | |
| {%- if not loop.last %} | |
| {{- ', ' }} | |
| {%- endif %} | |
| {%- endfor %} | |
| {{- ']' }} | |
| {%- endif %} | |
| {%- endmacro %} | |
| {%- if not messages %} | |
| {{- raise_exception('No messages provided.') }} | |
| {%- endif %} | |
| {#- Extract an initial system/developer message, so we can slot it into the Llama system header. #} | |
| {%- if messages[0]['role'] == 'system' or messages[0]['role'] == 'developer' %} | |
| {%- set system_message = render_content(messages[0]['content'], false, true)|trim %} | |
| {%- set messages = messages[1:] %} | |
| {%- else %} | |
| {%- if tools is not none %} | |
| {%- set system_message = "You are a helpful assistant with tool calling capabilities. Only reply with a tool call if the function exists in the library provided by the user. If it doesn't exist, just reply directly in natural language. When you receive a tool call response, use the output to format an answer to the original user question." %} | |
| {%- else %} | |
| {%- set system_message = "" %} | |
| {%- endif %} | |
| {%- endif %} | |
| {#- System message #} | |
| {{- "<|start_header_id|>system<|end_header_id|>\n\n" }} | |
| {%- if tools is not none %} | |
| {{- "Environment: ipython\n" }} | |
| {%- endif %} | |
| {%- if cutoff_date_string %} | |
| {{- "Cutting Knowledge Date: " + cutoff_date_string + "\n" }} | |
| {%- endif %} | |
| {%- if date_string %} | |
| {{- "Today Date: " + date_string + "\n" }} | |
| {%- endif %} | |
| {%- if tools is not none or cutoff_date_string or date_string %} | |
| {{- "\n" }} | |
| {%- endif %} | |
| {%- if tools is not none and not tools_in_user_message %} | |
| {{- "You have access to the following functions. To call a function, please respond with JSON for a function call. " }} | |
| {{- 'For a single function call, respond in the format {"name": function name, "parameters": dictionary of argument name and its value}. ' }} | |
| {{- 'If multiple function calls are required, respond with a JSON array of objects in that same format. ' }} | |
| {{- "Do not use variables.\n\n" }} | |
| {%- for t in tools %} | |
| {{- t | tojson(indent=4) }} | |
| {{- "\n\n" }} | |
| {%- endfor %} | |
| {%- endif %} | |
| {{- system_message }} | |
| {{- "<|eot_id|>" }} | |
| {#- Custom tools can be passed in the first user message with extra guidance. #} | |
| {%- if tools_in_user_message and tools is not none %} | |
| {%- if messages | length != 0 %} | |
| {%- set first_user_message = render_content(messages[0]['content'], true)|trim %} | |
| {%- set messages = messages[1:] %} | |
| {%- else %} | |
| {{- raise_exception("Cannot put tools in the first user message when there's no first user message!") }} | |
| {%- endif %} | |
| {{- '<|start_header_id|>user<|end_header_id|>\n\n' -}} | |
| {{- "Given the following functions, please respond with JSON for a function call " }} | |
| {{- "with its proper arguments that best answers the given prompt.\n\n" }} | |
| {{- 'For a single function call, respond in the format {"name": function name, "parameters": dictionary of argument name and its value}. ' }} | |
| {{- 'If multiple function calls are required, respond with a JSON array of objects in that same format. ' }} | |
| {{- "Do not use variables.\n\n" }} | |
| {%- for t in tools %} | |
| {{- t | tojson(indent=4) }} | |
| {{- "\n\n" }} | |
| {%- endfor %} | |
| {{- first_user_message + "<|eot_id|>" }} | |
| {%- endif %} | |
| {%- for message in messages %} | |
| {%- set role = message.role %} | |
| {%- if role == 'system' or role == 'developer' %} | |
| {{- '<|start_header_id|>system<|end_header_id|>\n\n' }} | |
| {{- render_content(message.content, true, true)|trim }} | |
| {{- '<|eot_id|>' }} | |
| {%- elif role == 'user' %} | |
| {{- '<|start_header_id|>user<|end_header_id|>\n\n' }} | |
| {{- render_content(message.content, true)|trim }} | |
| {{- '<|eot_id|>' }} | |
| {%- elif role == 'assistant' %} | |
| {%- set content = render_content(message.content, true)|trim %} | |
| {%- set reasoning_content = '' %} | |
| {%- if message.reasoning_content is defined and message.reasoning_content is string %} | |
| {%- set reasoning_content = message.reasoning_content|trim %} | |
| {%- elif '</think>' in content %} | |
| {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n')|trim %} | |
| {%- set content = content.split('</think>')[-1].lstrip('\n')|trim %} | |
| {%- endif %} | |
| {{- '<|start_header_id|>assistant<|end_header_id|>\n\n' }} | |
| {%- if reasoning_content %} | |
| {{- '<think>\n' + reasoning_content + '\n</think>\n\n' }} | |
| {%- endif %} | |
| {{- content }} | |
| {%- if message.tool_calls is defined and message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %} | |
| {%- if content or reasoning_content %} | |
| {{- '\n\n' }} | |
| {%- endif %} | |
| {{- render_tool_calls_json(message.tool_calls) }} | |
| {%- endif %} | |
| {{- '<|eot_id|>' }} | |
| {%- elif role == 'tool' or role == 'ipython' or role == 'function' %} | |
| {{- '<|start_header_id|>ipython<|end_header_id|>\n\n' }} | |
| {%- if message.content is mapping %} | |
| {{- {"output": message.content} | tojson }} | |
| {%- else %} | |
| {%- set tool_content = render_content(message.content, true)|trim %} | |
| {{- {"output": tool_content} | tojson }} | |
| {%- endif %} | |
| {{- '<|eot_id|>' }} | |
| {%- else %} | |
| {{- raise_exception('Unexpected message role: ' + role) }} | |
| {%- endif %} | |
| {%- endfor %} | |
| {%- if add_generation_prompt is defined and add_generation_prompt %} | |
| {{- '<|start_header_id|>assistant<|end_header_id|>\n\n' }} | |
| {%- if enable_thinking is defined and enable_thinking %} | |
| {{- '<think>\n' }} | |
| {%- endif %} | |
| {%- endif %} | |