SillyTavern/SillyTavern-Launcher
1
0
Fork 0
mirror of https://github.com/SillyTavern/SillyTavern-Launcher.git synced 2026-01-26 17:09:44 +00:00
Code Issues Packages Projects Releases Wiki Activity

launcher upgrades

Browse Source 
+added VRAM calculator by rolyattata
This commit is contained in:
deffcolony
2024-04-15 12:04:27 +02:00
parent b24127576c
commit 0a1f2879b0
1 changed files with 22 additions and 18 deletions
Download Patch File Download Diff File Expand all files Collapse all files

40
Launcher.bat
View File

@@ -3943,29 +3943,33 @@ echo %blue_fg_strong%/ Home / VRAM Info%reset%
echo -------------------------------------------------------------
REM Recommendations Based on VRAM Size
if %VRAM% lss 8 (
echo GPU VRAM: %cyan_fg_strong%%VRAM% GB%reset%
echo %cyan_fg_strong%%VRAM% GB%reset% - It's recommended to stick with APIs like OpenAI or OpenRouter for LLM usage, as local models might not perform well.
pause
goto :home
echo %cyan_fg_strong%GPU VRAM: %VRAM% GB%reset% - It's recommended to stick with APIs like OpenAI or OpenRouter for LLM usage, as local models might not perform well.
) else if %VRAM% lss 12 (
echo %cyan_fg_strong%%VRAM% GB%reset% - Capable of running efficient 7B models. However, APIs like OpenAI or OpenRouter will likely perform much better.
pause
goto :home
echo %cyan_fg_strong%GPU VRAM: %VRAM% GB%reset% - Capable of running efficient 7B models. However, APIs like OpenAI or OpenRouter will likely perform much better.
) else if %VRAM% lss 22 (
echo %cyan_fg_strong%%VRAM% GB%reset% - Suitable for 7B and some efficient 13B models, but APIs like OpenAI or OpenRouter are still recommended for much better performance.
pause
goto :home
echo %cyan_fg_strong%GPU VRAM: %VRAM% GB%reset% - Suitable for 7B and some efficient 13B models, but APIs like OpenAI or OpenRouter are still recommended for much better performance.
) else if %VRAM% lss 25 (
echo %cyan_fg_strong%%VRAM% GB%reset% - Good for 7B, 13B, 30B, and some efficient 70B models. Powerful local models will run well but APIs like OpenAI or Claude will still perform better than many local models.
pause
goto :home
echo %cyan_fg_strong%GPU VRAM: %VRAM% GB%reset% - Good for 7B, 13B, 30B, and some efficient 70B models. Powerful local models will run well but APIs like OpenAI or Claude will still perform better than many local models.
) else if %VRAM% gtr 25 (
echo %cyan_fg_strong%%VRAM% GB%reset% - Suitable for most models, including larger LLMs. You likely have the necessary expertise to pick your own model if you possess more than 25GB of VRAM.
pause
goto :home
echo %cyan_fg_strong%GPU VRAM: %VRAM% GB%reset% - Suitable for most models, including larger LLMs. You likely have the necessary expertise to pick your own model if you possess more than 25GB of VRAM.
) else (
echo An unexpected amount of VRAM detected or unable to detect VRAM. Check your system specifications.
pause
goto :home
)
echo.
echo Would you like to open the VRAM calculator website to check compatible models?
echo Just hit 'Load Models' after the page loads.
set /p uservram_choice=Check compatible models? [Y/N]
REM Check if user input is not empty and is neither "Y" nor "N"
if not "%uservram_choice%"=="" (
if /i not "%uservram_choice%"=="Y" if /i not "%uservram_choice%"=="N" (
echo %red_bg%[%time%]%reset% %red_fg_strong%[ERROR] Invalid input. Please enter Y for yes or N for no%reset%
pause
goto :vraminfo
)
)
if /i "%uservram_choice%"=="Y" ( start https://sillytavernai.com/llm-model-vram-calculator/?vram=%VRAM%
)
goto :home