Instructions to use OsaurusAI/Step-3.7-Flash-JANG_2L with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use OsaurusAI/Step-3.7-Flash-JANG_2L with MLX:

# Make sure mlx-vlm is installed
# pip install --upgrade mlx-vlm

from mlx_vlm import load, generate
from mlx_vlm.prompt_utils import apply_chat_template
from mlx_vlm.utils import load_config

# Load the model
model, processor = load("OsaurusAI/Step-3.7-Flash-JANG_2L")
config = load_config("OsaurusAI/Step-3.7-Flash-JANG_2L")

# Prepare input
image = ["http://images.cocodataset.org/val2017/000000039769.jpg"]
prompt = "Describe this image."

# Apply chat template
formatted_prompt = apply_chat_template(
    processor, config, prompt, num_images=1
)

# Generate output
output = generate(model, processor, formatted_prompt, image)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps Settings
LM Studio

How to use OsaurusAI/Step-3.7-Flash-JANG_2L with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "OsaurusAI/Step-3.7-Flash-JANG_2L"

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "OsaurusAI/Step-3.7-Flash-JANG_2L"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent

How to use OsaurusAI/Step-3.7-Flash-JANG_2L with Hermes Agent:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "OsaurusAI/Step-3.7-Flash-JANG_2L"

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default OsaurusAI/Step-3.7-Flash-JANG_2L

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use OsaurusAI/Step-3.7-Flash-JANG_2L with OpenClaw:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "OsaurusAI/Step-3.7-Flash-JANG_2L"

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "OsaurusAI/Step-3.7-Flash-JANG_2L" \
  --custom-provider-id mlx-lm \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Step-3.7-Flash-JANG_2L / jang_config.json

Osaurus-AI

Add files using upload-large-folder tool

660e38c verified 4 months ago

Raw

History Blame Contribute Delete

1.54 kB

	{
	"architecture": {
	"has_audio": false,
	"has_mtp_tensors": false,
	"has_vision": true,
	"text_model_type": "step3p5",
	"type": "step3p7"
	},
	"capabilities": {
	"cache_type": "kv",
	"family": "step3p7",
	"modality": "vision",
	"reasoning_parser": "qwen3",
	"supports_thinking": true,
	"supports_tools": true,
	"think_in_template": true,
	"tool_parser": "step3p5"
	},
	"format": "jang",
	"format_version": "2.0",
	"mxtq_bits": {
	"attention": 8,
	"embedding": 6,
	"routed_expert": {
	"down_proj": 3,
	"gate_proj": 4,
	"up_proj": 2
	}
	},
	"quantization": {
	"actual_bits": 3.398,
	"bit_widths_used": [
	2,
	3,
	4,
	6,
	8
	],
	"block_size": 128,
	"method": "jang-importance",
	"passthrough_bit_widths_used": [
	16,
	32
	],
	"profile": "JANG_2L",
	"quantization_backend": "mx.quantize",
	"source_weight_decode": "modelopt-nvfp4-for-routed-moe",
	"target_bits": 2.0,
	"total_quantized_bits": 669270114304,
	"total_source_bits": 3151291744256
	},
	"runtime": {
	"requires": [
	"step3p7-vlm-wrapper",
	"step3p5-text-runtime",
	"full-and-sliding-kv-cache",
	"head-wise-attention-gate",
	"qk-rmsnorm",
	"kv-scale-sidecars",
	"image-patch-processing"
	],
	"shard_count": 67,
	"total_shard_bytes": 87621944984
	},
	"source_model": {
	"dtype": "nvfp4+bf16",
	"hub_id": "stepfun-ai/Step-3.7-Flash-NVFP4",
	"name": "Step-3.7-Flash-NVFP4"
	}
	}