v2.1.3 Flash Attention

🢃 Download Release

Discord Support Server

---

v2.1.3 Changelog

• Support Flash attention for Ampere GPUs and above. This should improve performance on supported GPUs (>=3000 series)

Note

If the install fails or doesn't show up over the auto updater, please run force_update.bat as admin in the src folder of your install.
If you are updating from <v2.0.0, please reinstall the program.

All Changelogs >=v1.0

Full Changelog: v2.1.2...v2.1.3

---

Requirements

With default settings, this program has following requirements:

• .NET 4.8.1 (Should be preinstalled on Windows 10 and up)
• Visual C++ 2015-2022 Redistributable (x64)
• SteamVR (IF ran in VR, no Oculus/Meta support as of now.)
• Inference on GPU (Recommended):
  • CUDA enabled GPU (NVIDIA ONLY), otherwise it will fall back to using CPU.
  • ~11gb of available space for installation, ~6GB of space used after successful installation and loading models.
  • ~1GB of available RAM.
  • ~320MB of available VRAM.
• Inference on CPU:
  • ~4gb of available space for installation, ~2GB of space used after successful installation and loading models.
  • ~400MB of available RAM.

Note

Depending on settings changed in the program those requirements can change exponentially:
VRAM usages per Model: (int8 Precision. English models only)
~200MB with tiny.en
~220MB with base.en
~320MB with distil-small.en
~380MB with small.en
~580MB with distil-medium.en
~900MB with medium.en
~900MB with distil-large-v2
~1.6GB with large-v2

---

Installing and Updating TextboxSTT