← Back to blog
Posts tagged docker
-
Cloning Your Own Voice on an iGPU That ROCm Doesn't Support
Qwen3-TTS with voice cloning on an integrated Radeon 890M — entirely in Docker, with an OpenAI-compatible API and voice profiles that are a folder with two files. Including the operational quirks that appear in no guide.
Aug 13, 2026 · 13 min read
-
WhisperX on an AMD Radeon 890M: GPU Transcription with ROCm in Docker
Word-level timestamps and speaker diarization as a searchable edit list for video editing — with WhisperX on an integrated AMD Radeon 890M inside Docker, over 3× faster than CPU. Getting there meant finding three silent pitfalls, all of which left the GPU present but never used.
Aug 13, 2026 · 16 min read