Instructions to use DeepXR/Helion-V1.5-XL with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use DeepXR/Helion-V1.5-XL with Transformers:

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

pipe = pipeline("text-generation", model="DeepXR/Helion-V1.5-XL", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("DeepXR/Helion-V1.5-XL", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use DeepXR/Helion-V1.5-XL with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "DeepXR/Helion-V1.5-XL"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "DeepXR/Helion-V1.5-XL",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/DeepXR/Helion-V1.5-XL

SGLang

How to use DeepXR/Helion-V1.5-XL 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 "DeepXR/Helion-V1.5-XL" \
    --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": "DeepXR/Helion-V1.5-XL",
		"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 "DeepXR/Helion-V1.5-XL" \
        --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": "DeepXR/Helion-V1.5-XL",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use DeepXR/Helion-V1.5-XL with Docker Model Runner:
```
docker model run hf.co/DeepXR/Helion-V1.5-XL
```

Helion-V1.5-XL

File size: 1,552 Bytes

d82b168

{
  "architectures": [
    "HelionForCausalLM"
  ],
  "model_type": "helion",
  "torch_dtype": "bfloat16",
  "transformers_version": "4.35.0",
  
  "_name_or_path": "DeepXR/Helion-V1.5-XL",
  "auto_map": {
    "AutoConfig": "configuration_helion.HelionConfig",
    "AutoModelForCausalLM": "modeling_helion.HelionForCausalLM"
  },
  
  "vocab_size": 100000,
  "hidden_size": 6144,
  "intermediate_size": 24576,
  "num_hidden_layers": 48,
  "num_attention_heads": 32,
  "num_key_value_heads": 8,
  
  "max_position_embeddings": 16384,
  "rope_theta": 10000.0,
  "rope_scaling": {
    "type": "linear",
    "factor": 2.0
  },
  
  "hidden_act": "silu",
  "initializer_range": 0.02,
  "rms_norm_eps": 1e-06,
  
  "use_cache": true,
  "pad_token_id": 0,
  "bos_token_id": 1,
  "eos_token_id": 2,
  
  "tie_word_embeddings": false,
  "attention_bias": false,
  "attention_dropout": 0.0,
  "pretraining_tp": 1,
  
  "use_flash_attention_2": true,
  "use_sliding_window": false,
  "sliding_window": null,
  
  "quantization_config": {
    "quant_method": "bitsandbytes",
    "load_in_8bit": false,
    "load_in_4bit": false,
    "llm_int8_threshold": 6.0,
    "llm_int8_has_fp16_weight": false,
    "bnb_4bit_compute_dtype": "bfloat16",
    "bnb_4bit_use_double_quant": true,
    "bnb_4bit_quant_type": "nf4"
  },
  
  "generation_config": {
    "max_length": 16384,
    "max_new_tokens": 512,
    "temperature": 0.7,
    "top_p": 0.9,
    "top_k": 50,
    "repetition_penalty": 1.1,
    "do_sample": true,
    "num_beams": 1,
    "early_stopping": false
  }
}