Instructions to use KSP-NMAI/boris-50M-superlight-cubscout with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use KSP-NMAI/boris-50M-superlight-cubscout with Transformers:

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

pipe = pipeline("text-generation", model="KSP-NMAI/boris-50M-superlight-cubscout")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("KSP-NMAI/boris-50M-superlight-cubscout")
model = AutoModelForCausalLM.from_pretrained("KSP-NMAI/boris-50M-superlight-cubscout")
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]:]))

llama-cpp-python

How to use KSP-NMAI/boris-50M-superlight-cubscout with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="KSP-NMAI/boris-50M-superlight-cubscout",
	filename="boris_v1.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use KSP-NMAI/boris-50M-superlight-cubscout 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 KSP-NMAI/boris-50M-superlight-cubscout
# Run inference directly in the terminal:
llama cli -hf KSP-NMAI/boris-50M-superlight-cubscout

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf KSP-NMAI/boris-50M-superlight-cubscout
# Run inference directly in the terminal:
llama cli -hf KSP-NMAI/boris-50M-superlight-cubscout

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 KSP-NMAI/boris-50M-superlight-cubscout
# Run inference directly in the terminal:
./llama-cli -hf KSP-NMAI/boris-50M-superlight-cubscout

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 KSP-NMAI/boris-50M-superlight-cubscout
# Run inference directly in the terminal:
./build/bin/llama-cli -hf KSP-NMAI/boris-50M-superlight-cubscout

Use Docker

docker model run hf.co/KSP-NMAI/boris-50M-superlight-cubscout

LM Studio
Jan

vLLM

How to use KSP-NMAI/boris-50M-superlight-cubscout with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "KSP-NMAI/boris-50M-superlight-cubscout"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KSP-NMAI/boris-50M-superlight-cubscout",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/KSP-NMAI/boris-50M-superlight-cubscout

SGLang

How to use KSP-NMAI/boris-50M-superlight-cubscout 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 "KSP-NMAI/boris-50M-superlight-cubscout" \
    --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": "KSP-NMAI/boris-50M-superlight-cubscout",
		"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 "KSP-NMAI/boris-50M-superlight-cubscout" \
        --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": "KSP-NMAI/boris-50M-superlight-cubscout",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use KSP-NMAI/boris-50M-superlight-cubscout with Ollama:
```
ollama run hf.co/KSP-NMAI/boris-50M-superlight-cubscout
```

Unsloth Studio

How to use KSP-NMAI/boris-50M-superlight-cubscout with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for KSP-NMAI/boris-50M-superlight-cubscout to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for KSP-NMAI/boris-50M-superlight-cubscout to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for KSP-NMAI/boris-50M-superlight-cubscout to start chatting

Atomic Chat new
Docker Model Runner
How to use KSP-NMAI/boris-50M-superlight-cubscout with Docker Model Runner:
```
docker model run hf.co/KSP-NMAI/boris-50M-superlight-cubscout
```

Lemonade

How to use KSP-NMAI/boris-50M-superlight-cubscout with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull KSP-NMAI/boris-50M-superlight-cubscout

Run and chat with the model

lemonade run user.boris-50M-superlight-cubscout-{{QUANT_TAG}}

List all available models

lemonade list

boris-50M-superlight-cubscout

boris-50M-superlight-cubscout (Boris) is a lightweight, ~50M parameter text generation model trained entirely on the roneneldan/TinyStories dataset.

It was developed entirely on one NVIDIA RTX 3060 in ~2 hours. Boris's primary use case is generating bad children's short stories.

Traning Details:

Trained on TinyStories (1000 steps)
Trained using one NVIDIA RTX 3060 (12GB VRAM)
Precision: FP16
Final Traning Loss: ~1.76

Advice:

Set max tokens to ~50-100.
This is a base model, and does not know how to stop. Add stop sequences like "the end." or ###

Evaluation Results:

Final Training Loss: ~1.76 TinyStories (Train)

Perplexity (PPL): 8.52 TinyStories (Validation)

Copyright & License:

This project and all associated files (the "Work") are licensed under the Apache License, Version 2.0 (the "License"); you may not use this project except in compliance with the License. You may obtain a copy of the License at:

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

Downloads last month: 25

Safetensors

Model size

58.1M params

Tensor type

F32

Dataset used to train KSP-NMAI/boris-50M-superlight-cubscout

Space using KSP-NMAI/boris-50M-superlight-cubscout 1

Collection including KSP-NMAI/boris-50M-superlight-cubscout

NMAI-Superlight

Collection

2 items • Updated May 14