Instructions to use HelpingAI/hai3.1-checkpoint-0002 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use HelpingAI/hai3.1-checkpoint-0002 with Transformers:

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

pipe = pipeline("text-generation", model="HelpingAI/hai3.1-checkpoint-0002", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("HelpingAI/hai3.1-checkpoint-0002", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use HelpingAI/hai3.1-checkpoint-0002 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "HelpingAI/hai3.1-checkpoint-0002"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "HelpingAI/hai3.1-checkpoint-0002",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/HelpingAI/hai3.1-checkpoint-0002

SGLang

How to use HelpingAI/hai3.1-checkpoint-0002 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 "HelpingAI/hai3.1-checkpoint-0002" \
    --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": "HelpingAI/hai3.1-checkpoint-0002",
		"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 "HelpingAI/hai3.1-checkpoint-0002" \
        --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": "HelpingAI/hai3.1-checkpoint-0002",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use HelpingAI/hai3.1-checkpoint-0002 with Docker Model Runner:
```
docker model run hf.co/HelpingAI/hai3.1-checkpoint-0002
```

hai3.1-checkpoint-0002

File size: 2,366 Bytes

{
  "architectures": [
    "HelpingAIForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "auto_map": {
    "AutoConfig": "configuration_helpingai.HelpingAIConfig",
    "AutoModelForCausalLM": "modeling_helpingai.HelpingAIForCausalLM"
  },
  "bos_token_id": 151643,
  "emotion_hidden_size": 512,
  "empathy_scaling_factor": 1.2,
  "eos_token_id": 151645,
  "head_dim": 128,
  "hidden_act": "silu",
  "hidden_size": 5120,
  "initializer_range": 0.02,
  "intermediate_size": 17408,
  "layer_types": [
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention"
  ],
  "max_position_embeddings": 40960,
  "max_window_layers": 40,
  "model_type": "helpingai",
  "num_attention_heads": 40,
  "num_emotion_heads": 4,
  "num_hidden_layers": 40,
  "num_key_value_heads": 8,
  "num_thinking_stages": 3,
  "perspective_threads": 4,
  "reasoning_temperature": 0.8,
  "rms_norm_eps": 1e-06,
  "rope_scaling": null,
  "rope_theta": 1000000,
  "sliding_window": null,
  "speech_head_hidden_dim": null,
  "speech_loss_type": "l1",
  "speech_num_mels": 80,
  "speech_upsample_factor": 1,
  "structured_head_activation": "gelu",
  "structured_head_hidden_dim": 9578,
  "structured_head_type": "mlp_v1",
  "structured_output_vocab_size": 100,
  "thinking_depth": 2,
  "tie_word_embeddings": false,
  "torch_dtype": "bfloat16",
  "transformers_version": "4.55.2",
  "use_cache": true,
  "use_emotional_reasoning": false,
  "use_perspective_threading": true,
  "use_sliding_window": false,
  "use_speech_output": false,
  "vocab_size": 151669
}