Instructions to use KR-16/broken-model with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use KR-16/broken-model with Transformers:

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

pipe = pipeline("text-generation", model="KR-16/broken-model")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("KR-16/broken-model")
model = AutoModelForCausalLM.from_pretrained("KR-16/broken-model", 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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use KR-16/broken-model with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "KR-16/broken-model"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KR-16/broken-model",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/KR-16/broken-model

SGLang

How to use KR-16/broken-model 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 "KR-16/broken-model" \
    --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": "KR-16/broken-model",
		"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 "KR-16/broken-model" \
        --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": "KR-16/broken-model",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use KR-16/broken-model with Docker Model Runner:
```
docker model run hf.co/KR-16/broken-model
```

Fixed Model for Chat Completions API

This repository contains the fixed version of the model from yunmorning/broken-model, with corrections applied to enable /chat/completions API functionality.

Changes Made

tokenizer_config.json

Added `chat_template` field

What was changed: Added the chat_template field to tokenizer_config.json
Why: The chat completions API requires a chat template to properly format conversation messages. Without this template, the model cannot process chat messages in the expected format. The template uses Qwen's <|im_start|> and <|im_end|> tokens (which are already present in the tokenizer configuration) to delimit different message roles (system, user, assistant).
Template format: The added template follows the Qwen chat format, using Jinja2 syntax to format messages with <|im_start|>role\n and <|im_end|>\n tokens.

Model Information

Model Type: Qwen3ForCausalLM
Base Model: meta-llama/Meta-Llama-3.1-8B
Architecture: Qwen3
Vocabulary Size: 151936

Original Model

Original broken model: https://huggingface.co/yunmorning/broken-model

Downloads last month: 10

Safetensors

Model size

8B params

Tensor type

BF16

Model tree for KR-16/broken-model

Base model

meta-llama/Llama-3.1-8B

Finetuned

(1501)

this model