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 / config.json

Trouter-Library

Create config.json

d82b168 verified 10 months ago

Raw

History Blame

1.55 kB

	{
	"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
	}
	}