Start here if you are new
You do not need to understand every GPU setting first. Pick the path that matches your goal, prepare one dataset, then let the training console guide the launch.
Rule of thumb: use LoRA when you want a small reusable style, character, product, or motion adapter. Use pretraining pods when you need full control over code, checkpoints, and runtime.
- 01Decision
Choose the result you want
Pick LoRA post-training if you want downloadable .safetensors weights. Pick a pretraining pod if you need to run your own training code on a GPU machine.
Choose LoRA or pretraining - 02/infra/storage
Prepare one dataset
For LoRA, upload a zip of images or short clips in Storage and mark it as kind=dataset. For pretraining, you can attach a dataset now or bring data in after the pod starts.
Upload in Storage - 03Dashboard
Open the training console
Go to Infrastructure -> Training. The top form launches a pretraining pod; the LoRA form starts a managed fine-tune and shows your credit hold before launch.
Open /infra/training - 04One click
Launch and watch progress
Click Deploy pretraining pod or Start training. Pretraining redirects to the pod page. LoRA runs appear in the Runs list with progress, cancel, refund, and output state.
Start the run - 05Finish
Finish the run safely
Download LoRA weights when the run completes. For pretraining pods, save checkpoints and stop or terminate the pod when you are done so hourly billing stops.
Download or stop
What is live today
Pretraining pod
Deploy a dedicated GPU pod for LLM, multimodal, or continued pretraining with your runtime image and dataset.
- Own Docker image, GPU type, GPU count, disk, and volume size
- Optional owned dataset signed into HYPEREAL_DATASET_URL for 24 hours
- SSH, TensorBoard, API, and notebook ports opened automatically
LoRA post-training
Train curated image and video LoRA models against private datasets, then download the resulting weights.
- Flux Dev, Qwen Image, Wan 2.2 Image, and Wan 2.2 I2V trainers
- Trigger word, steps, learning rate, and LoRA rank controls
- Progress, cancel, refund, webhook reconciliation, and R2 output storage
Multi-node pretraining cluster
Plan and request distributed training capacity with Mercury topology and runtime hints.
- Quote GPU count, network, orchestrator, region, and storage
- Inspect topology, scheduler, NCCL, and runtime hints
- Not marketed as one-click physical multi-node launch yet
Single-node pretraining and LoRA post-training are one-click. Multi-node pretraining remains a capacity workflow until physical cluster provisioning is wired end to end.
Curated LoRA trainers
Flux Dev (LoRA)
Photoreal characters, products, styles
The general-purpose pick. Train a face, a product, or a visual style and use the weights anywhere Flux runs.
Qwen Image (LoRA)
High-fidelity illustration and character LoRAs
Strong on stylized illustration and East-Asian character likeness. Use the same zipped dataset flow as Flux.
Wan 2.2 Image (LoRA)
Stylized high-resolution image LoRAs
High-resolution image LoRA on the Wan 2.2 base. Returns dual high and low-noise weights for advanced workflows.
Wan 2.2 I2V (LoRA)
Image-to-video motion control
Train on short video clips to lock in motion, camera, or character animation. The flagship for custom video.
Credits are held when you start and refunded on submit failure, upstream failure, or cancellation before delivery.
The complete training workflow
Pretraining and continued pretraining
Pick GPU runtime
Choose GPU type, count, Docker image, disk, volume size, framework, precision, and sequence length.
Attach data
Use an owned storage dataset or attach it later. Dataset URLs are signed for the pod at launch.
Deploy pod
The first hour is held up front, the pod is recorded locally, and normal pod lifecycle controls take over.
Train and checkpoint
Use SSH, TensorBoard, API, or notebook access, then write checkpoints under HYPEREAL_OUTPUT_DIR.
LoRA post-training
Upload dataset
Upload a zip of images or short clips to Storage and mark it as kind=dataset.
Choose trainer
Pick Flux, Qwen Image, Wan 2.2 Image, or Wan 2.2 I2V and set optional LoRA hyperparameters.
Start and monitor
Hypereal holds credits, submits the run, tracks progress, and lets you cancel before delivery.
Download weights
Completed runs copy outputs into private R2 storage and expose downloadable .safetensors weights.
What teams deploy
Continued LLM pretraining
Spin up a reproducible PyTorch, DeepSpeed, or Nanotron pod for domain data, tokenizer experiments, or checkpoint continuation.
Character and product LoRAs
Train consistent people, mascots, products, packaging, and house styles without rebuilding prompts from scratch.
Custom video motion
Train short-clip LoRAs for repeatable camera moves, character animation, or branded motion assets.