Instructions to use NewEden/Trinity-Mini-Futaba with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use NewEden/Trinity-Mini-Futaba with Transformers:

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

pipe = pipeline("text-generation", model="NewEden/Trinity-Mini-Futaba", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("NewEden/Trinity-Mini-Futaba", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("NewEden/Trinity-Mini-Futaba", trust_remote_code=True)
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 NewEden/Trinity-Mini-Futaba with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/NewEden/Trinity-Mini-Futaba

SGLang

How to use NewEden/Trinity-Mini-Futaba 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 "NewEden/Trinity-Mini-Futaba" \
    --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": "NewEden/Trinity-Mini-Futaba",
		"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 "NewEden/Trinity-Mini-Futaba" \
        --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": "NewEden/Trinity-Mini-Futaba",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use NewEden/Trinity-Mini-Futaba with Docker Model Runner:
```
docker model run hf.co/NewEden/Trinity-Mini-Futaba
```

You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Trinity Mini CleanBlend Step 100

Private step-100 checkpoint from the Prime RL run:

Run output: outputs/trinity-mini-cleanblend-32k-cp2-chunk1024
Checkpoint: weights/step_100
Sequence length: 32768
Training cap: 200 steps
Export format: sharded safetensors

This repository contains the full model export plus the LoRA adapter directory emitted by the trainer.

Downloads last month: 62

Safetensors

Model size

26B params

Tensor type

F32

BF16

Model tree for NewEden/Trinity-Mini-Futaba

Base model

arcee-ai/Trinity-Mini-Base-Pre-Anneal

Finetuned

arcee-ai/Trinity-Mini-Base

Finetuned

arcee-ai/Trinity-Mini

Finetuned

(4)

this model

Finetunes

1 model