Instructions to use Minami-su/IA_14B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Minami-su/IA_14B with Transformers:

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

pipe = pipeline("text-generation", model="Minami-su/IA_14B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Minami-su/IA_14B")
model = AutoModelForCausalLM.from_pretrained("Minami-su/IA_14B")
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

vLLM

How to use Minami-su/IA_14B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Minami-su/IA_14B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Minami-su/IA_14B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Minami-su/IA_14B

SGLang

How to use Minami-su/IA_14B 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 "Minami-su/IA_14B" \
    --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": "Minami-su/IA_14B",
		"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 "Minami-su/IA_14B" \
        --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": "Minami-su/IA_14B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Minami-su/IA_14B with Docker Model Runner:
```
docker model run hf.co/Minami-su/IA_14B
```

Exploring Collaboration in LLM Systems & AI Roleplay

by hrquartday - opened 22 days ago

Discussion

hrquartday

22 days ago

Hi,

I recently came across your work on GitHub, and it’s clear that you have a strong ability to bridge model capabilities with real-world applications.

I’m currently with Crushon.AI, where we focus on building an AI roleplay platform. What we’re exploring is how to move large language models beyond “general conversation” toward experiences that are more controllable, stable, and immersive. For example:

Maintaining character consistency over long interactions
Controlling output structure (format, tone, style)
Long-context memory / RAG systems
Multi-model adaptation and alignment

At the moment, we’re looking for engineers or researchers with hands-on experience in the LLM space, especially those who have:

Worked on model alignment or fine-tuning (e.g., LoRA, instruction tuning)
Experience with RAG or long-term memory systems
Experience in multi-model adaptation or evaluation
Built practical, production-level LLM systems

That said, I also understand you may already be focused on your own projects or research. So this outreach isn’t just about a “role”—we’d also love to connect in a more open-ended way, such as:

Project collaboration or joint exploration
Technical discussions and idea exchange
Or potential long-term collaboration in the future

If this sounds even slightly interesting, I’d be happy to share more about what we’re building, our tech stack, and some of the challenges we’re tackling. I’d also love to hear more about what you’re currently focusing on.

Best regards,
Luna
HRBP
Crushon.AI

Minami-su

Owner 21 days ago

Hi,

I recently came across your work on GitHub, and it’s clear that you have a strong ability to bridge model capabilities with real-world applications.

I’m currently with Crushon.AI, where we focus on building an AI roleplay platform. What we’re exploring is how to move large language models beyond “general conversation” toward experiences that are more controllable, stable, and immersive. For example:

Maintaining character consistency over long interactions
Controlling output structure (format, tone, style)
Long-context memory / RAG systems
Multi-model adaptation and alignment

At the moment, we’re looking for engineers or researchers with hands-on experience in the LLM space, especially those who have:

Worked on model alignment or fine-tuning (e.g., LoRA, instruction tuning)
Experience with RAG or long-term memory systems
Experience in multi-model adaptation or evaluation
Built practical, production-level LLM systems

That said, I also understand you may already be focused on your own projects or research. So this outreach isn’t just about a “role”—we’d also love to connect in a more open-ended way, such as:

Project collaboration or joint exploration
Technical discussions and idea exchange
Or potential long-term collaboration in the future

If this sounds even slightly interesting, I’d be happy to share more about what we’re building, our tech stack, and some of the challenges we’re tackling. I’d also love to hear more about what you’re currently focusing on.

Best regards,
Luna
HRBP
Crushon.AI

Hi Luna,
Thanks for reaching out! I really appreciate your interest in my work.
What you are building at Crushon.AI sounds incredibly interesting and highly relevant to my current focus. I’ve been deeply involved in LLM alignment, empathy reward modeling, and evaluating AI character adherence for companionship applications. The challenges you mentioned—especially around character consistency, controllable output, and multi-model adaptation—are right up my alley, and I'd love to hear more about your tech stack and the specific hurdles you're tackling.
I am definitely open to connecting for technical discussions, idea exchange, or exploring potential collaborations.
I'm happy to exchange contact information to make chatting easier. You can reach me via email, WeChat, or QQ—whichever is most convenient for you:
Email: 2012456373@qq.com
WeChat: a2012456373
QQ: 2012456373
Looking forward to connecting and sharing ideas!
Best regards,
Minami-su

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment