According to a2a.computer
One question, one answer, one path — backed by the live capability graph, not branding. Each page is a single projection: readable here, structured as JSON-LD for discovery, and machine-readable for agents at /api/according/<slug>.
Model embodiment
- How do I turn a HuggingFace model into a GGUF file?Pull the model from the Hub, then convert and quantize it to GGUF with llama.cpp. a2a.computer plans this as a two-hop governed path and can realize each hop with an evidence receipt.hub_repo → gguf
- How do I package a LoRA fine-tune for local serving?Merge the LoRA adapter into its base model, then quantize the merged model to GGUF. With the base already pulled, a2a.computer plans the merge→quantize path over governed capabilities.lora_adapter → gguf
Multi-modal
- How do I turn audio into an image?Transcribe the audio to text, then generate an image from that text. a2a.computer plans this cross-modal path across the multi-modal capability family and realizes each hop under policy with evidence.audio → image
- How do I extract text from audio?Transcribe the audio with a Whisper ASR capability. a2a.computer plans this as a single governed hop and returns an evidence receipt for the transcription.audio → text