Instructions to use namelessai/TrillionHelp-7.5b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use namelessai/TrillionHelp-7.5b with Transformers:

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

pipe = pipeline("text-generation", model="namelessai/TrillionHelp-7.5b")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("namelessai/TrillionHelp-7.5b")
model = AutoModelForMultimodalLM.from_pretrained("namelessai/TrillionHelp-7.5b")
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 namelessai/TrillionHelp-7.5b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "namelessai/TrillionHelp-7.5b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "namelessai/TrillionHelp-7.5b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/namelessai/TrillionHelp-7.5b

SGLang

How to use namelessai/TrillionHelp-7.5b 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 "namelessai/TrillionHelp-7.5b" \
    --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": "namelessai/TrillionHelp-7.5b",
		"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 "namelessai/TrillionHelp-7.5b" \
        --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": "namelessai/TrillionHelp-7.5b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use namelessai/TrillionHelp-7.5b with Docker Model Runner:
```
docker model run hf.co/namelessai/TrillionHelp-7.5b
```

Model Card for TrillionHelp

TrillionHelp uses trillionlabs/Trillion-7B-preview as the backbone.

Model Details

This model is fine-tuned on the namelessai/helply dataset designed to enhance mental health reasoning capabilities.

Model Description

This model was fine-tuned for assisting pyschologists in assiting patients.

Developed by: Alex Scott
Model type: Language Model, Adapter Model (available in a folder in the model repo)
Finetuned from model: trillionlabs/Trillion-7B-preview

Usage (Adapter Only, full model snippet coming soon)

Use the code snippet below to load the base model and apply the adapter for inference:

from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel

# Load the base model
base_model_name = "trillionlabs/Trillion-7B-preview"
adapter_path = "/path/to/adapter"  # Replace with actual adapter path
tokenizer = AutoTokenizer.from_pretrained(base_model_name)
base_model = AutoModelForCausalLM.from_pretrained(base_model_name)

# Apply the adapter
model = PeftModel.from_pretrained(base_model, adapter_path)
model = model.merge_and_unload()

# Run inference
input_text = "Your input text here"
inputs = tokenizer(input_text, return_tensors="pt")
outputs = model.generate(**inputs)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))