YAML Metadata Warning:The pipeline tag "text-to-motion" is not in the official list: text-classification, token-classification, table-question-answering, question-answering, zero-shot-classification, translation, summarization, feature-extraction, text-generation, fill-mask, sentence-similarity, text-to-speech, text-to-audio, automatic-speech-recognition, audio-to-audio, audio-classification, audio-text-to-text, voice-activity-detection, depth-estimation, image-classification, object-detection, image-segmentation, text-to-image, image-to-text, image-to-image, image-to-video, unconditional-image-generation, video-classification, reinforcement-learning, robotics, tabular-classification, tabular-regression, tabular-to-text, table-to-text, multiple-choice, text-ranking, text-retrieval, time-series-forecasting, text-to-video, image-text-to-text, image-text-to-image, image-text-to-video, visual-question-answering, document-question-answering, zero-shot-image-classification, graph-ml, mask-generation, zero-shot-object-detection, text-to-3d, image-to-3d, image-feature-extraction, video-text-to-text, keypoint-detection, visual-document-retrieval, any-to-any, video-to-video, other
FlowMDM ONNX (WebGPU Optimized)
This repository contains the standalone, single-file ONNX export of FlowMDM (Human Motion Diffusion Model) compiled specifically for browser execution via ONNX Runtime Web and WebGPU.
It is designed for real-time, client-side text-to-motion generation in AIRI.
Model Details
- Model Type: FlowMDM (Motion Diffusion Model)
- Dataset: HumanML3D (263 motion features)
- Conditioning: CLIP Text Encoder (
Xenova/clip-vit-base-patch32, 512-dim embedding) - Sampling: 50-step DDIM deterministic solver ($t \in [0, 999]$, outputting $x_0$)
- Format: Single-file ONNX with inline weight tensor serialization (no external
.datadependencies, CORS / WASM sandbox safe).
Repository Contents
flow_mdm.onnx: The WebGPU-optimized motion denoiser graph (86.8 MB).diffusion_stats.json: Gaussian diffusion variance schedules andalphas_cumprodcoefficients.Mean.npy: HumanML3D feature normalization means.Std.npy: HumanML3D feature normalization standard deviations.
Usage in AIRI
import * as ort from 'onnxruntime-web/webgpu';
// Load the model on WebGPU
const session = await ort.InferenceSession.create('https://huggingface.co/dasilva333/flowmdm-onnx/resolve/main/flow_mdm.onnx', {
executionProviders: ['webgpu']
});