| Name | Modified | Size | Downloads / Week |
|---|---|---|---|
| Parent folder | |||
| rocm-python-wheels-Linux.zip | 2026-10-25 | 389.3 MB | |
| rocm-python-wheels-Windows.zip | 2026-10-25 | 158.7 MB | |
| README.md | 2026-10-25 | 1.6 kB | |
| v4.8.4 source code.tar.gz | 2026-10-25 | 3.6 MB | |
| v4.8.4 source code.zip | 2026-10-25 | 3.9 MB | |
| Totals: 5 Items | 555.5 MB | 0 | |
New features
-
Build Python 3.15 wheels (#2129) by @shaanmajid
-
Add an experimental implementation of Drama 3 offline text-to-speech, including model conversion and waveform decoding on CPU with optional CUDA and ROCm acceleration by @jordimas
- Add CPU and GPU implementations of CausalDepthwiseConv1D, ConvTranspose1D, DequantizeFP8, GatedDeltaNet, MoeTopK, and Snake operations for audio models (#2121) by @jordimas
See: https://opennmt.net/CTranslate2/guides/drama3.html
Fixes and improvements
- Guard integer overflow in tensor size computation, reshape, and growth (#2097) by @Kyozzz and @jordimas
- Fix CPU Conv1D dilation and INT8 quantization scale overflow for tiny weight rows (#2117) by @jordimas
- Use the caching allocator by default on HIP (#2118) by @jordimas
- Fix Wav2Vec2 lookups of packed linear weights (#2119) by @arcusbuilds
- Fix Wav2Vec2-BERT float16 inference (#2126) by @jordimas
- Rebuild rotary caches when the device or data type changes by @jordimas
- Restore the Transformer include in the public attention header by @jordimas
- Improve CI reliability with updated GitHub Actions and retries for ROCm package index updates and macOS OneAPI downloads (#2125) by @jordimas
