Name | Modified | Size | Downloads / Week |
---|---|---|---|
Parent folder | |||
textgen-portable-3.9.1-windows-cuda12.4.zip | 2025-08-07 | 833.8 MB | |
textgen-portable-3.9.1-windows-cuda11.7.zip | 2025-08-07 | 722.3 MB | |
textgen-portable-3.9.1-windows-cpu.zip | 2025-08-07 | 193.7 MB | |
textgen-portable-3.9.1-windows-vulkan.zip | 2025-08-07 | 202.7 MB | |
textgen-portable-3.9.1-linux-cuda12.4.zip | 2025-08-07 | 838.7 MB | |
textgen-portable-3.9.1-linux-cuda11.7.zip | 2025-08-07 | 766.9 MB | |
textgen-portable-3.9.1-linux-vulkan.zip | 2025-08-07 | 240.8 MB | |
textgen-portable-3.9.1-linux-cpu.zip | 2025-08-07 | 231.7 MB | |
textgen-portable-3.9.1-macos-x86_64.zip | 2025-08-07 | 164.8 MB | |
textgen-portable-3.9.1-macos-arm64.zip | 2025-08-07 | 176.8 MB | |
README.md | 2025-08-07 | 251 Bytes | |
v3.9.1 source code.tar.gz | 2025-08-07 | 24.9 MB | |
v3.9.1 source code.zip | 2025-08-07 | 25.0 MB | |
Totals: 13 Items | 4.4 GB | 20 |
Changes
- Several improvements to the GPT-OSS template handling. Special actions like "Continue" and "Impersonate" now work correctly.
- Update llama.cpp to https://github.com/ggml-org/llama.cpp/tree/5fd160bbd9d70b94b5b11b0001fd7f477005e4a0