Instructions to use vovaRL/NanoColibri-Instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use vovaRL/NanoColibri-Instruct with Transformers:

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

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

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("vovaRL/NanoColibri-Instruct")
model = AutoModelForCausalLM.from_pretrained("vovaRL/NanoColibri-Instruct", 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 vovaRL/NanoColibri-Instruct with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/vovaRL/NanoColibri-Instruct

SGLang

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

Docker Model Runner
How to use vovaRL/NanoColibri-Instruct with Docker Model Runner:
```
docker model run hf.co/vovaRL/NanoColibri-Instruct
```

NanoColibri-Instruct / training_state.json

vovaRL

Upload folder using huggingface_hub

b3ac35d verified 3 days ago

Raw

History Blame Contribute Delete

1.89 kB

	{
	"step": 21500,
	"global_tokens": 5439488000,
	"contributor": "vovaRL",
	"cloud": "home",
	"base_model": "scratch:colibri-nano",
	"wall_clock_s": 11548.808067560196,
	"history_tail": [
	{
	"step": 21460,
	"t": 1785260063.755,
	"ntp": 1.8235763311386108,
	"zloss": 0.3778407573699951,
	"lr": 3.0002803191548873e-05,
	"grad_norm": 0.322265625,
	"ppl": 6.193970576015481,
	"balance_max_imbalance": 0.0843505859375,
	"balance_dead_frac": 0.015625,
	"tokens": 5434245120
	},
	{
	"step": 21470,
	"t": 1785260140.412,
	"ntp": 3.5947928428649902,
	"zloss": 0.40486156940460205,
	"lr": 3.000157679763363e-05,
	"grad_norm": 0.37109375,
	"ppl": 36.40815699711934,
	"balance_max_imbalance": 0.229736328125,
	"balance_dead_frac": 0.109375,
	"tokens": 5435555840
	},
	{
	"step": 21480,
	"t": 1785260217.057,
	"ntp": 1.5007089376449585,
	"zloss": 0.40441009402275085,
	"lr": 3.0000700799706186e-05,
	"grad_norm": 0.33984375,
	"ppl": 4.484867434931063,
	"balance_max_imbalance": 0.0440673828125,
	"balance_dead_frac": 0.0,
	"tokens": 5436866560
	},
	{
	"step": 21490,
	"t": 1785260293.595,
	"ntp": 1.7874301671981812,
	"zloss": 0.3486020863056183,
	"lr": 3.000017520004023e-05,
	"grad_norm": 0.328125,
	"ppl": 5.974080335333217,
	"balance_max_imbalance": 0.2159423828125,
	"balance_dead_frac": 0.015625,
	"tokens": 5438177280
	},
	{
	"step": 21500,
	"t": 1785260370.219,
	"ntp": 1.4909950494766235,
	"zloss": 0.3707721531391144,
	"lr": 2.9999999999999997e-05,
	"grad_norm": 0.333984375,
	"ppl": 4.441512845942532,
	"balance_max_imbalance": 0.076904296875,
	"balance_dead_frac": 0.015625,
	"tokens": 5439488000
	}
	]
	}