Aitrepreneur
patreon
1-Click INSTALL SillyTavern - BEST ROLEPLAY UI EVER!
🕑 Added 2023-12-22 20:45:41 +0000 UTCHey everyone! I've created a 1-click auto installer for the AMAZING SillyTavern UI, the best Ui for LLM roleplay. Featuring character animation, text-to-speech, voice recognition, group chat, Dynamic Audio and so much more! You Absolutely need to try this out!
Watch the video on SillyTavern: https://youtu.be/enWO16x6tRM
The installer of course automates the entire install process!
Here's how it works:
1. Download the installer, drag and drop it inside a folder (without any spaces) then run file (without admin rights)
2. Select your GPU ("1" for Nvidia, "2" for AMD, "3" for CPU) then press enter
3.???
4. Profit π
The Sillytavern UI will be automatically installed as well as the text to speech Xtts api server and the SillyTavern extras folders will be created inside the Sillytavern folder.
Don't forget to use the launcher bat files for xtts and extras that you can find right here:
https://www.patreon.com/posts/95126591
And as always, supporting me on Patreon allows me to keep creating helpful resources like this for the community. Thank you for your support - now go have some...huh... fun *wink*π *wink* π!
Comments
Grey Walker
Try installing the v4 file instead of the v3.
Nathan Trent
Install seems fine but when I launch the xxts Bat inside the silly tavern folder, prompt just opens and closes. I'm running my llm on ollama, not sure if that would prevent it from executing properly?
Thomas Foster
What is your GPU? NVIDIA AMD CPU Enter 1, 2 or 3 and press enter: 1 [INFO] Winget is already installed. [INFO] winget already exists in PATH. Git already installed. [WARN] Miniconda3 is not installed on this system. [INFO] Installing Miniconda3 using Winget... Found an existing package already installed. Trying to upgrade the installed package... No available upgrade found. No newer package versions are available from the configured sources. [INFO] Miniconda3 installed successfully. Please restart the Installer. Press any key to continue . . . keeps saying restart the installer n im on my new computer and no files in folder the link to we share aslo down
Sophie Sinclair
I installed, but how do I run it?
Robbieddanger
Right you are. I'm an idiot...
Aitrepreneur
It's on this page
Robbieddanger
Where is the installer? I don't see any download link
Kankatzu
Thanks, works pretty OK on my 3080. (just Got a 3090 for a decent price on ebay. The extra 14gb of vram should help)
Aitrepreneur
then no way you ran without admin though, it probably installed it in system32 folder
Aitrepreneur
no this is only sillytavern, text gen is in another post: https://www.patreon.com/posts/1-click-install-92531137
MarcAndreMueller
Hey! I put the installer v4 inside a folder I named "SillyTavern" and then I ran the installer without admin rights. The installation went through, but there are no files in the folder I ran the installer from. Any help please?
Kankatzu
does this come with text generation web ui?
Aitrepreneur
this error means you don't have nodejs installed correctly: https://nodejs.org/en
Tim McCabe
whenever I try and open the start.bat file I keep getting this: 'npm' is not recognized as an internal or external command, operable program or batch file. 'node' is not recognized as an internal or external command, operable program or batch file. Press any key to continue . . . And then anykey just causes it to close T_T
Aitrepreneur
Try updating node js to the latest version (20 something, right now you have 16.20) and retry the installer again
Prince
Firas Mghames
so i tried installing the V4, and for some reason its not creating a folder called silly tavern with the files you showed in the vid inside, it just makes a folder called xtts thats 6gb in size
Anthony Sparks
when i run start i get this error: 'npm' is not recognized as an internal or external command, operable program or batch file. 'node' is not recognized as an internal or external command, operable program or batch file. Press any key to continue . . .
Netblinder
Thanks boss
Aitrepreneur
I've added the V4 file to the patreon post
Netblinder
What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 1 [INFO] Winget is already installed. [INFO] winget already exists in PATH. Git already installed. [WARN] Miniconda3 is not installed on this system. [INFO] Installing Miniconda3 using Winget... Found an existing package already installed. Trying to upgrade the installed package... No available upgrade found. No newer package versions are available from the configured sources. [INFO] Miniconda3 installed successfully. Please restart the Installer. Press any key to continue . . . Looks like I need the v4 but none of the links are live.
Steven De Page
Since there was no confirmation from the others yet, in case someone else comes here looking: Was getting the same error: What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 1 [INFO] Winget is already installed. [INFO] winget already exists in PATH. Git already installed. [INFO] Miniconda3 is already installed. [INFO] Execution Alias for python.exe was already removed. [INFO] Execution Alias for python3.exe was already removed. [WARN] Node.js is not installed on this system. [INFO] Installing Node.js using Winget... Failed when opening source(s); try the 'source reset' command if the problem persists. [INFO] Node.js installed successfully. Please restart the Installer. Press any key to continue . . . Tried the v4 debug recommended to others by Aitrepreneur (https://we.tl/t-a4e0jkO2to), and now it is installing. SillyTavern Installation complete, NOW GO HAVE FUN! Press any key to continue . . . :)
Aitrepreneur
try this one then: https://we.tl/t-a4e0jkO2to
Aitrepreneur
then try this one then: https://we.tl/t-a4e0jkO2to
Anthony Sparks
Hello getting the error like others but the links are dead: What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 1 [INFO] Winget is already installed. [INFO] winget already exists in PATH. Git already installed. [WARN] Miniconda3 is not installed on this system. [INFO] Installing Miniconda3 using Winget... Found an existing package already installed. Trying to upgrade the installed package... No available upgrade found. No newer package versions are available from the configured sources. [INFO] Miniconda3 installed successfully. Please restart the Installer. Press any key to continue . . . after restart it just says same thing over and over again
Firas Mghames
yes i've restarted it countless times
Aitrepreneur
have you tried simply restarting the installer? Or the message is still there?
Firas Mghames
Hello, im getting this issue What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 1 [INFO] Winget is already installed. [INFO] winget already exists in PATH. Git already installed. [INFO] Miniconda3 is already installed. [INFO] Execution Alias for python.exe was already removed. [INFO] Execution Alias for python3.exe was already removed. [WARN] Node.js is not installed on this system. [INFO] Installing Node.js using Winget... Failed when opening source(s); try the 'source reset' command if the problem persists. [INFO] Node.js installed successfully. Please restart the Installer. Press any key to continue . . . i tried the links you sent on the previous messages but they still didn't work
Aitrepreneur
try this one: https://we.tl/t-Bt5Tmc18mk
Peter Greene
im having the same issue with the miniconda here
Peter Greene
What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 1 [INFO] Winget is already installed. [INFO] winget already exists in PATH. Git already installed. [WARN] Miniconda3 is not installed on this system. [INFO] Installing Miniconda3 using Winget... Found Miniconda3 [Anaconda.Miniconda3] Version py312_24.5.0-0 This application is licensed to you by its owner. Microsoft is not responsible for, nor does it grant any licenses to, third-party packages. Downloading https://repo.anaconda.com/miniconda/Miniconda3-py312_24.5.0-0-Windows-x86_64.exe ββββββββββββββββββββββββββββββ 83.1 MB / 83.1 MB Successfully verified installer hash Starting package install... Successfully installed [INFO] Miniconda3 installed successfully. Please restart the Installer. Press any key to continue . . .
Aitrepreneur
https://we.tl/t-Bt5Tmc18mk
Da_AwesomeKai
Could you please send the link again?
Aitrepreneur
try this one: https://we.tl/t-X5cCnVehN4
William Troy Green
I'm stuck on the "restart installer" as well. Using the v3 one linked in post.
Aitrepreneur
https://we.tl/t-X5cCnVehN4
iioooooii
please upload a new one , this file no longer exist either
Aitrepreneur
here: https://we.tl/t-RhmZI2Pd9l
Jack Kipps
i am having the same issue i tried this link but the file no longer exist.
Metalhead RetroGamer
Thank you sir, you are a hero
Aatrox
could you tell me how to clean every instance of it so i can do a clean reinstall?
Domenico
tks you
Martin Ramirez
Was this ever resolved? I am having the same problem
Suneko
i have the same problem
Aitrepreneur
dm me your pc specs and python version
Joseph Christensen
also ive tried other one click installers from you and nothing has ever shown up.
Joseph Christensen
i am unable to use the one click installer. it runs and then says its done installing, but then i have no files for anything except the installer. ive tried restarting my pc, and running it multiple times ,and doing a full pc file search and there is nothing. need help
Aitrepreneur
you can try this one: https://we.tl/t-ZagQjppqEr
koozy
I keep getting this What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 1 [INFO] Winget is already installed. [INFO] winget already exists in PATH. Git already installed. [WARN] Miniconda3 is not installed on this system. [INFO] Installing Miniconda3 using Winget... Found an existing package already installed. Trying to upgrade the installed package... No available upgrade found. No newer package versions are available from the configured sources. [INFO] Miniconda3 installed successfully. Please restart the Installer. Press any key to continue . . .
Albedo
i have the same issue as the guy
Albedo
could i trouble you for another link
Aitrepreneur
apparently Fimbulvetr V2 11B is great
Zeis
Which model would you recommend I run with this? I've currently got Nous-Hermes-13B.Q4_K_M and guanaco-13b-uncensored.Q5_K_M downloaded, is there a better one I should use?
Aitrepreneur
The installer has been updated no need for another one, what issue are you having?
Ali
Can you drop the link again?
Aitrepreneur
The installer has been updated just dl it from the page
Nicholas Lalonde
Could I get the link for the new installer? Every time I try the link for Wetransfer it says that the link has expired. Thank you
Starseed.1111
For anyone else that's reading this later and has a similar problem. I had to uninstall Miniconda, making sure to reinstall the version that matches with Python 3.10 which I already had on my machine. Then make sure you add it to user PATH not system PATH and place it above python in the list. Also you need to open the Miniconda Prompt as an Administrator. and run conda init followed by conda activate. That cleared up the following error for me: [INFO] Activating Conda environment extras... CondaError: Run 'conda init' before 'conda activate' [ERROR] Failed to activate Conda environment extras.
SomeRandomGuy
Worked for me too. Thank you.
Starseed.1111
Can the following be ignored or am I better off going for a manual install instead of using the 1-click installer? (Assuming it will have the same issue if I run it again.) pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts. tts 0.21.3 requires numpy==1.22.0; python_version Also I didn't expect you to respond so quickly I accidentally pressed enter when I was adding a space to write something and before I could edit it, you'd already replied. Sorry lol.
Aitrepreneur
looks like miniconda was not installed properly, either install miniconda manuall or redo the install again in a new folder, note any issues during install just in case
Starseed.1111
This is so frustrating.. I installed SillyTavern with the 1-click-installer, but for some reason it didn't install xtts properly but I wasn't bothered as I wasn't ready to start using it yet. However I just tried to run it again and during the installation process I got these error messages: 'conda' is not recognized as an internal or external command, operable program or batch file. [INFO] Creating Conda environment extras... The system cannot find the path specified. 'conda' is not recognized as an internal or external command, operable program or batch file. [INFO] Activating Conda environment extras... 'conda' is not recognized as an internal or external command, operable program or batch file. [ERROR] Failed to activate Conda environment extras. [INFO] Installing Python and Git in the Conda environment... 'conda' is not recognized as an internal or external command, operable program or batch file. and ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts. tts 0.21.3 requires numpy==1.22.0; python_version <= "3.10", but you have numpy 1.26.4 which is incompatible. How can conda not be recognised if it's literally installed as part of the .bat file? And I'm not sure why there's a dependency conflict. Anyways I tried to test the LaunchXTTS.bat file anyway just to see if it'd run this time and it did, but I ended up getting another error message. UserWarning: torch.utils._pytree._register_pytree_node is deprecated. Please use torch.utils._pytree.register_pytree_node instead. I'm don't seem to be having much luck, but at least it opened this time. Do you have any idea what I'm could be doing wrong?
MattMatt
Like Kevin wrote above: AFTER you get into the miniconda loop, try editing line 70 to this (line 69 shown so you can see you are in the right place) call conda --version > nul 2>&1 if %errorlevel% equ 0 (
Marta M
The same problem for me
Aitrepreneur
then try this one: https://we.tl/t-SkELYHCP6n
Dorian
What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 1 [INFO] Winget is already installed. [INFO] winget already exists in PATH. Git already installed. [WARN] Miniconda3 is not installed on this system. [INFO] Installing Miniconda3 using Winget... An existing package has already been installed. Attempting to upgrade the installed package... Available upgrade not found. No newer package versions are available from the configured sources. [INFO] Miniconda3 installed successfully. Please restart the Installer.
Dorian
thanks for your time, but i got a error with the new BAT,stuck at miniconda3, it restart endlessly
Aitrepreneur
yup just checked, they apparently updated some stuff a few days ago, I just updated the installer, should work now, dl the V3 version and reinstall in a new folder
Dorian
it makes an error when installing "spacy-3.0.6.tar.gz" it looks like
Dorian
I have an error during installation with the BAT file, can you help me? Using cached spacy-3.0.6.tar.gz (7.1 MB) Installing build dependencies ... done Getting requirements to build wheel ... error error: subprocess-exited-with-error Γ Getting requirements to build wheel did not run successfully. β exit code: 1 β°β> [164 lines of output] Error compiling Cython file: ------------------------------------------------------------ ... int length cdef class Vocab: cdef Pool mem cpdef readonly StringStore strings ^ ------------------------------------------------------------ spacy\vocab.pxd:28:10: Variables cannot be declared with 'cpdef'. Use 'cdef' instead.
trevor gus
this helped me continue thanks a lot
Kevin
AFTER you get into the miniconda loop, try editing line 70 to this (line 69 shown so you can see you are in the right place) call conda --version > nul 2>&1 if %errorlevel% equ 0 ( basically changing the neq to equ
Raz 23
Same Problem here: What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 1 [INFO] Winget is already installed. [INFO] winget already exists in PATH. Git already installed. [WARN] Miniconda3 is not installed on this system. [INFO] Installing Miniconda3 using Winget... Found an existing package already installed. Trying to upgrade the installed package... No available upgrade found. No newer package versions are available from the configured sources. [INFO] Miniconda3 installed successfully. Please restart the Installer. Press any key to continue . . . Already tried: - Restart in new Folder (not as Admin) - Deinstall Miniconda3, same result after reinstalling - "Restart installer" several times ... - Same process with all Protection turned off Additional Info: Prior to this I installed stable-cascade via "OneClick" installer. There I had problems with Python and Git, fixed those by: -> Python: Reinstall and activate "Set PYTHON-PATH" as SystemVariable -> Git: Simply manually reinstall -> Restart installer after each step and was able to finish the installation of stable-cascade ------ Edit: I fixed it by not using the "OneClick" Installer but git clone instead.
Aitrepreneur
restart the installer
AwesomeTM
I ran the v2 .bat, and I got the same error: What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 2 [INFO] Winget is already installed. [INFO] winget already exists in PATH. Git already installed. [WARN] Miniconda3 is not installed on this system. [INFO] Installing Miniconda3 using Winget... Found an existing package already installed. Trying to upgrade the installed package... The package cannot be upgraded using winget. Please use the method provided by the publisher for upgrading this package. [INFO] Miniconda3 installed successfully. Please restart the Installer. Press any key to continue . . .
Aitrepreneur
I meant from the actual patreon page
Sense Departed
the file is not available on that site
Aitrepreneur
I just updated the installer, download the v2 from the page and try again
KOR HyperTiger
I have same problem and your link said expired.
Aitrepreneur
Yeah this is not normal, something is stuck during install, try disabling any antivirus and firewall and relaunch the installer in a new folder, run the installer normally not as admin
Tom Caughlin
I had the same problem with the first Bat file. Below is the result of the second Bat file. 4 hours later and I'm still waiting on a prompt. The slash is still spinning. What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 1 [INFO] Winget is already installed. [INFO] winget already exists in PATH. [INFO] Removed Execution Alias for python.exe [INFO] Removed Execution Alias for python3.exe [WARN] Node.js is not installed on this system. [INFO] Installing Node.js using Winget... Found Node.js LTS [OpenJS.NodeJS.LTS] Version 20.12.2 This application is licensed to you by its owner. Microsoft is not responsible for, nor does it grant any licenses to, third-party packages. Downloading https://nodejs.org/dist/v20.12.2/node-v20.12.2-x64.msi ββββββββββββββββββββββββββββββ 25.3 MB / 25.3 MB Successfully verified installer hash Starting package install... The installer will request to run as administrator, expect a prompt.
Aitrepreneur
try this installer: https://we.tl/t-Vn1ia2GzFW
Aitrepreneur
here: https://we.tl/t-Vn1ia2GzFW
grindbane
What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 1 [INFO] Winget is already installed. [INFO] winget already exists in PATH. [INFO] Git is already installed. [WARN] Miniconda3 is not installed on this system. [INFO] Installing Miniconda3 using Winget... Found an existing package already installed. Trying to upgrade the installed package... No available upgrade found. No newer package versions are available from the configured sources. [INFO] Miniconda3 installed successfully. Please restart the Installer. Press any key to continue . . . Can I get a fix?
Enrique Mora
Hi I'm having the same problem :c Can you please share with me a new link?
Aitrepreneur
here: https://we.tl/t-Vn1ia2GzFW
Sixlow
Have the same problem as the others. I can't download the alternate installer because it's expired, any help please?
Aitrepreneur
here: https://we.tl/t-RWK9HWbmOh
Aitrepreneur
have you watched my sillytavern video? have you installed the text gen webui and a model corresponding to your gpu?
aziboy
hi, im having the same issue, but the link you sent is expired, can you please reupload the file again?
Kapoks
thank you, there still is a problem though. everything got installed but when i start it, it just launches the normal SillyTavern. there is no chat ai or anything.
Aitrepreneur
here: https://we.tl/t-RWK9HWbmOh
Kapoks
i have the same problem, could i also have an alternate installer link
Aitrepreneur
use this one: https://we.tl/t-mVZoJGnUg2
Jos
have the same problem as the others. I saw you provided an alternate installer but it's expired, any help please?
Lusaca
The Bat file doesnt work in the correct way, because winget or miniconda do not add the system variable! After the installation of miniconda you have to run "set PATH=%PATH%;C:\Users\%USERNAME%\miniconda3\Scripts\" to do it manually!
Aitrepreneur
try this: go the xtts folder, click on the folder path type cmd press enter, this will bring the command prompt window and there type venv\Scripts\activate then press enter and then type pip install deepspeed then press enter if that still doesn't work for some reason, then edit the launchxtts.bat file and remove the deepspeed argument
Akshay
I get this error while running Launch XTTS CoquiEngine: Error initializing main coqui engine model: No module named 'deepspeed' Traceback (most recent call last): File "E:\TORR\Tavern\SILLYTAVERN\SillyTavern\xtts\venv\lib\site-packages\xtts_api_server\RealtimeTTS\engines\coqui_engine.py", line 289, in _synthesize_worker tts.load_checkpoint( File "E:\TORR\Tavern\SILLYTAVERN\SillyTavern\xtts\venv\lib\site-packages\TTS\tts\models\xtts.py", line 772, in load_checkpoint self.gpt.init_gpt_for_inference(kv_cache=self.args.kv_cache, use_deepspeed=use_deepspeed) File "E:\TORR\Tavern\SILLYTAVERN\SillyTavern\xtts\venv\lib\site-packages\TTS\tts\layers\xtts\gpt.py", line 222, in init_gpt_for_inference import deepspeed ModuleNotFoundError: No module named 'deepspeed' Process Process-1: Traceback (most recent call last): File "C:\Users\aksha\miniconda3\lib\multiprocessing\process.py", line 315, in _bootstrap self.run() File "C:\Users\aksha\miniconda3\lib\multiprocessing\process.py", line 108, in run self._target(*self._args, **self._kwargs) File "E:\TORR\Tavern\SILLYTAVERN\SillyTavern\xtts\venv\lib\site-packages\xtts_api_server\RealtimeTTS\engines\coqui_engine.py", line 289, in _synthesize_worker tts.load_checkpoint( File "E:\TORR\Tavern\SILLYTAVERN\SillyTavern\xtts\venv\lib\site-packages\TTS\tts\models\xtts.py", line 772, in load_checkpoint self.gpt.init_gpt_for_inference(kv_cache=self.args.kv_cache, use_deepspeed=use_deepspeed) File "E:\TORR\Tavern\SILLYTAVERN\SillyTavern\xtts\venv\lib\site-packages\TTS\tts\layers\xtts\gpt.py", line 222, in init_gpt_for_inference import deepspeed ModuleNotFoundError: No module named 'deepspeed'
Aitrepreneur
if everything was installer correctly you can try with this installer instead: https://we.tl/t-mVZoJGnUg2
Alpha Ghost 47
I keep getting this then it closes and when I rerun it does same thing
Alpha Ghost 47
What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 1 [INFO] Winget is already installed. [INFO] winget already exists in PATH. [INFO] Git is already installed. [WARN] Miniconda3 is not installed on this system. [INFO] Installing Miniconda3 using Winget... Found an existing package already installed. Trying to upgrade the installed package... No available upgrade found. No newer package versions are available from the configured sources. [INFO] Miniconda3 installed successfully. Please restart the Installer. Press any key to continue . . .
Aitrepreneur
here: https://we.tl/t-c6ATNTrqZt
mysteryintime
Hey, could you please re-update the file? It expired and I am having the same issue.
Aitrepreneur
here: https://we.tl/t-c6ATNTrqZt
Skylark8756
I'm also getting this issue, but the link you've provided appears to have expired.
Aitrepreneur
might be an antivirus or firewall blocking the file and connection, disable them both and try also with this installer: https://we.tl/t-gXA9XVnBi9
Emu
[INFO] Installing vs_BuildTools... % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0 curl: (35) schannel: next InitializeSecurityContext failed: Unknown error (0x80092012) - The revocation function was unable to check revocation for the certificate. [ERROR] Download failed. Please restart the installer Press any key to continue . . .
Aitrepreneur
try this: go the xtts folder, click on the folder path type cmd press enter, this will bring the command prompt window and there type venv\Scripts\activate then press enter and then type pip install deepspeed then press enter if that still doesn't work for some reason, then edit the launchxtts.bat file and remove the deepspeed argument
Joe
I get this error when running the .bat (Using paython 3.9 because other versions do not run the .bat at all) D:\SillyTavern\SillyTavern\xtts\venv\lib\site-packages\pydub\utils.py:170: RuntimeWarning: Couldn't find ffmpeg or avconv - defaulting to ffmpeg, but may not work warn("Couldn't find ffmpeg or avconv - defaulting to ffmpeg, but may not work", RuntimeWarning) 2024-03-28 14:42:30.074 | INFO | xtts_api_server.modeldownloader:install_deepspeed_based_on_python_version:47 - Python version: 3.9 2024-03-28 14:42:30.074 | ERROR | xtts_api_server.modeldownloader:install_deepspeed_based_on_python_version:64 - Unsupported Python version on Windows. 2024-03-28 14:42:30.077 | WARNING | xtts_api_server.server::63 - 'Streaming Mode' has certain limitations, you can read about them here https://github.com/daswer123/xtts-api-server#about-streaming-mode 2024-03-28 14:42:30.077 | INFO | xtts_api_server.server::66 - You launched an improved version of streaming, this version features an improved tokenizer and more context when processing sentences, which can be good for complex languages like Chinese 2024-03-28 14:42:30.078 | INFO | xtts_api_server.RealtimeTTS.engines.coqui_engine:__init__:103 - Loading official model 'v2.0.2' for streaming v2.0.2 D:\SillyTavern\SillyTavern\xtts\venv\lib\site-packages\pydub\utils.py:170: RuntimeWarning: Couldn't find ffmpeg or avconv - defaulting to ffmpeg, but may not work warn("Couldn't find ffmpeg or avconv - defaulting to ffmpeg, but may not work", RuntimeWarning) > Using model: xtts CoquiEngine: Error initializing main coqui engine model: No module named 'deepspeed' Traceback (most recent call last): File "D:\SillyTavern\SillyTavern\xtts\venv\lib\site-packages\xtts_api_server\RealtimeTTS\engines\coqui_engine.py", line 289, in _synthesize_worker tts.load_checkpoint( File "D:\SillyTavern\SillyTavern\xtts\venv\lib\site-packages\TTS\tts\models\xtts.py", line 772, in load_checkpoint self.gpt.init_gpt_for_inference(kv_cache=self.args.kv_cache, use_deepspeed=use_deepspeed) File "D:\SillyTavern\SillyTavern\xtts\venv\lib\site-packages\TTS\tts\layers\xtts\gpt.py", line 222, in init_gpt_for_inference import deepspeed ModuleNotFoundError: No module named 'deepspeed' Process Process-1: Traceback (most recent call last): File "C:\Program Files\WindowsApps\PythonSoftwareFoundation.Python.3.9_3.9.3568.0_x64__qbz5n2kfra8p0\lib\multiprocessing\process.py", line 315, in _bootstrap self.run() File "C:\Program Files\WindowsApps\PythonSoftwareFoundation.Python.3.9_3.9.3568.0_x64__qbz5n2kfra8p0\lib\multiprocessing\process.py", line 108, in run self._target(*self._args, **self._kwargs) File "D:\SillyTavern\SillyTavern\xtts\venv\lib\site-packages\xtts_api_server\RealtimeTTS\engines\coqui_engine.py", line 289, in _synthesize_worker tts.load_checkpoint( File "D:\SillyTavern\SillyTavern\xtts\venv\lib\site-packages\TTS\tts\models\xtts.py", line 772, in load_checkpoint self.gpt.init_gpt_for_inference(kv_cache=self.args.kv_cache, use_deepspeed=use_deepspeed) File "D:\SillyTavern\SillyTavern\xtts\venv\lib\site-packages\TTS\tts\layers\xtts\gpt.py", line 222, in init_gpt_for_inference import deepspeed ModuleNotFoundError: No module named 'deepspeed'
Aitrepreneur
You need to run the xtts server before choosing the xtts option
Joe
Hi. When choosing XTTS in the SillyTavern options I get this error: "TTS provider failed to return voice ids" Ca you help with that?
Aitrepreneur
try this installer then: https://we.tl/t-yjgra7qU7N
Notracks
Failed opening source Source reset command? Gpu = 3 Wingit already exists in path Warn git Info installing git Failed when opening source Info git installed successfully Press any key ..... Same thing
Aitrepreneur
you just need to relaunch the installer again, but since everything seems to have been installed correctly you can also try this installer instead : https://we.tl/t-lBzybSwm7Q
Aitrepreneur
you just need to relaunch the installer again, but since everything seems to have been installed correctly you can also try this installer instead : https://we.tl/t-lBzybSwm7Q
Volkfang
It seems like I'm having a similar issue as people above: What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 1 [INFO] Winget is already installed. [INFO] winget already exists in PATH. [INFO] Git is already installed. [WARN] Miniconda3 is not installed on this system. [INFO] Installing Miniconda3 using Winget... Failed in attempting to update the source: winget Failed when searching source: msstore An unexpected error occurred while executing the command: 0x8a15005e : The server certificate did not match any of the expected values. The following packages were found among the working sources. Please specify one of them using the --source option to proceed. Name Id Source ------------------------------------- Miniconda3 Anaconda.Miniconda3 winget [INFO] Miniconda3 installed successfully. Please restart the Installer. Press any key to continue . . .
Chris Smith
I installed it but the folder i installed on has nothing in it. when I launch the installer again it says this. What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 1 [INFO] Winget is already installed. [INFO] winget already exists in PATH. [INFO] Git is already installed. [WARN] Miniconda3 is not installed on this system. [INFO] Installing Miniconda3 using Winget... Found an existing package already installed. Trying to upgrade the installed package... No available upgrade found. No newer package versions are available from the configured sources. [INFO] Miniconda3 installed successfully. Please restart the Installer. Press any key to continue . . . I do not have any idea what I am doing wrong.
Aitrepreneur
you can try CaPlatTessDolXaBoros-Yi-34B-200K-DARE-Ties or Noromaid or MxLewd
Owlcatcher
Do you have any recommendations for 30B - 33B models that fit well for this purpose?
Aitrepreneur
not sure, I know it works on windows and linux but maybe not for Mac
Steven Barbon
does it work for mac? i have an macbook air m2
Aitrepreneur
dm me some screenshots of the session tab in the text gen webui and the api window inside sillytavern
The Pysio
it did work thank you now im trying to connect the text generation to the silly tavern. for some reason its says Trying to connect to API: { api_server: 'http://the site/', api_type: 'ooba', legacy_api: false } Models endpoint is offline. while the model is working on the text generator site
Aitrepreneur
If everything was installed correctly, you can try this installer instead: https://we.tl/t-5n2NCOf264
xxx
hello. I totally setup my windows11 pc new for this. When I start the file I get this message: What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 1 [INFO] Winget is already installed. [INFO] winget already exists in PATH. [INFO] Git is already installed. [WARN] Miniconda3 is not installed on this system. [INFO] Installing Miniconda3 using Winget... \ [INFO] Miniconda3 installed successfully. Please restart the Installer. DrΓΌcken Sie eine beliebige Taste . . . and this start over and over
Aitrepreneur
Thanks unfortunately I can't include model downloads since it all depends on everyone's gpu and on what they want from it, so it's up to you to choose which model you want to use. Also having separate installers for 2 different programs is easier to solve people's errors. The webui installer is here btw: https://www.patreon.com/file?h=92531137&i=16491564
Khairi Johari
my good sir, can i ask if possible, respectfully, would there be a way for you to create a "one click installer" that would include the sillytavern, webui and the models at once? so that we can start using the program that has the models loaded automatically? that way, we/i wont have to worry about errors? THank you!! I LOVE YOUR VIDEOS!! HUGE FAN HERE!!!
euphobie
same error
Aitrepreneur
this error is ok, this shouldn't be a problem to launch it though
Khairi Johari
hi there, i need some assistance please! when i did the one click install, im getting the following: "Error: pip's dependency resolver does not currently take into account all the packages that are installed. This behavior is the source of the following dependency conflicts. TTS 0.21.3 requires pandas <2.0,>=1.4, but you have pandas 2.2.1 which is incompatible." what does this mean? im running windows 10 pro, with RTX 4090. im a noob at this, please help. also i dont really know how to navigate in silly tavern as i believe i dont have the other stuffs installed, i think its the web ui? and other extensions in the video. im soo lost T_T
Aitrepreneur
Hmm not sure elevenlabs is linked to gpu but more like internet speed since you are using the api key, if you want tts you can try xtts instead
Poole
Thank you. I've found it within SIllytavern. My response times are over 2 minutes. Besides getting more VRAM, is there a way to speed up the reply?
Aitrepreneur
If you have everything correctly installed and it still doesnt' work then try this installer instead: https://we.tl/t-MF3QJTaAnj
The Pysio
I got some problems with the miniconda instalation it says that i got it when i dont and when i do it doesnt work.
Aitrepreneur
In Sillytavern or text-gen webui? It was removed in text-gen but is still here in Sillytavern
Poole
I cannot see the Elevenlabs extension checkbox in the ui. Did it get removed?
trevor gus
you are a King Amongst Kings!!! btw love your tutorials, any chance of you maybe making one on an AI that can make video clips?
Aitrepreneur
here: https://we.tl/t-XQZormytLr
trevor gus
i tried link but file is no long shared im having the same winget problem Fatty was after a fresh install of windows 11
Aitrepreneur
launch the start.bat file
Mather Man
how do i start it after 1 click install?
MemeL0rd420
For some reason, it installed into the system32 folder... which worried me lol.
Aitrepreneur
just search for the name of the folder "sillytavern" or just create a new folder somewhere you know, put the installer inside and launch it
MemeL0rd420
thats very strange then. I thought it would be but its not there
Aitrepreneur
it's installed in the same place where you launched the installer...
MemeL0rd420
I have installed it but for the life of me I have no idea where its installed to. Please help
Aitrepreneur
send me a dm with more info, like GPU, model used and screenshots of the session tab inside the text gen webui, sillytavern and the command prompt windows for both
Aitrepreneur
try this: go the xtts folder, click on the folder path type cmd press enter, this will bring the command prompt window and there type venv\Scripts\activate then press enter and then type pip install deepspeed then press enter if that still doesn't work for some reason, then edit the launchxtts.bat file and remove the deepspeed argument
JD
I am unable to get any reply after connecting to WEBUI and my WEBUI Chat works fine. Settings are: API=Text Completion, API Type=Default, Server URL=http://127.0.0.1:5000/, and connected to my AWQ (green indication). Any ideas?
Carlos ιΎζ΅©ε
Hey, when I load the LounchXTTS.bat file in the SillyTavern folder it says ModuleNotFound: No Module Named deepspeed Any tips?
Aitrepreneur
here: https://we.tl/t-6x3Mq9xaD1
Anthony Adams
hi yes I am having the same problem and the link above does not work, please help <3 and thank you for all your hard work!
Aitrepreneur
you don't select horde, rewatch the video again carefully
Ramses CΓ’mara
error (No Horde model selected or the selected models are no longer available. Please choose another model) what to do?
NateMac00
nevermind, i jumped to far into the tutorial, not ALL 1-click
NateMac00
when i try to connect, i get an error, not sure where to go from here. used the installer. https://imgur.com/a/AeQUoCh
Aitrepreneur
You can try this installer: https://we.tl/t-0o3WbtAfVf
wanyu jerry
It tell me Please restart the Installer. Can you help me?
Aitrepreneur
try this: go the xtts folder, click on the folder path type cmd press enter, this will bring the command prompt window and there type venv\Scripts\activate then press enter and then type pip install deepspeed then press enter if that still doesn't work for some reason, then edit the launchxtts.bat file and remove the deepspeed argument
Richard Rambert
When I choose XTTSV2 as the TTS provider, I get the error "TTS provider failed to return voice ids" meanwhile in the window for the "launchXTTS.bat" the following is shown: > Using model: xtts CoquiEngine: Error initializing main coqui engine model: No module named 'deepspeed' Traceback (most recent call last): File "D:\TAVERN\SillyTavern\xtts\venv\lib\site-packages\xtts_api_server\RealtimeTTS\engines\coqui_engine.py", line 289, in _synthesize_worker tts.load_checkpoint( File "D:\TAVERN\SillyTavern\xtts\venv\lib\site-packages\TTS\tts\models\xtts.py", line 772, in load_checkpoint self.gpt.init_gpt_for_inference(kv_cache=self.args.kv_cache, use_deepspeed=use_deepspeed) File "D:\TAVERN\SillyTavern\xtts\venv\lib\site-packages\TTS\tts\layers\xtts\gpt.py", line 222, in init_gpt_for_inference import deepspeed ModuleNotFoundError: No module named 'deepspeed' Process Process-1: Traceback (most recent call last): File "C:\Users\drtho\miniconda3\lib\multiprocessing\process.py", line 315, in _bootstrap self.run() File "C:\Users\drtho\miniconda3\lib\multiprocessing\process.py", line 108, in run self._target(*self._args, **self._kwargs) File "D:\TAVERN\SillyTavern\xtts\venv\lib\site-packages\xtts_api_server\RealtimeTTS\engines\coqui_engine.py", line 289, in _synthesize_worker tts.load_checkpoint( File "D:\TAVERN\SillyTavern\xtts\venv\lib\site-packages\TTS\tts\models\xtts.py", line 772, in load_checkpoint self.gpt.init_gpt_for_inference(kv_cache=self.args.kv_cache, use_deepspeed=use_deepspeed) File "D:\TAVERN\SillyTavern\xtts\venv\lib\site-packages\TTS\tts\layers\xtts\gpt.py", line 222, in init_gpt_for_inference import deepspeed ModuleNotFoundError: No module named 'deepspeed' Any ideas?
Aitrepreneur
dm me more info, gpu and what error you get
Aitrepreneur
that mobile gpu only has 6GB of vram and you are trying to run models that are way too big for that GPU. Try this model instead: TheBloke/Xwin-MLewd-7B-V0.2-GPTQ:gptq-4bit-32g-actorder_True past that line in the model field, dl it and then use the exllama loader to load it
#The
I need help running silly tavern. It's not connecting when I paste the local url. Please help
Mishal Khan
Hi, almost everything is fine but when I type something it says API error(Internal server error) and after checking for a while it says "ValueError: No model is loaded! Select one in the Model tab." in the cmd which think might be the issue. How do I fix this? (Btw I tried to use TheBloke_LLaMA2-13B-Tiefighter-AWQ, TheBloke_LLaMA2-13B-Tiefighter-GPTQ, and wojtab_llava-13b-v0-4bit-128g models and I have a rtx 3060 notebook.
Aitrepreneur
the start.bat is there, next to the start.sh
todd
i am given a start.sh file but i am unable to run it. why isn't it a start.bat file like mentioned in the video?
Aitrepreneur
Oh sorry didn't see the reply, him send me in dm some screenshots, webui session tab, syllytavern xtts, and command prompt windows with the errors
Skuttlebot
Any thoughts or just abandoning ship? lol
Aitrepreneur
here, try this: https://we.tl/t-PJIivn9QZ5
Redd
I keep getting the miniconda installing done, "restart installer", over and over error. I saw that you offered another bat file for setup but the wetransfer link for that expired.
Skuttlebot
As far as I know, being that both the OpenAi and Sillytavern are working the installs were successful. This happens when I launch the XTTS.Bat, after both OpenAi and Sillytavern are running. The first issue seems to be the Coqui not initializing but you are right, there is a lot to get through and I cant tell how many of those are chained to eachother. Further I have reinstalled both OpenAi and Sillytavern and the issue remains. (I am somewhat savvy, but this has me stumped -but may just be missing something obvious). This looks to me like 2 issues that seem to have a similar common cause: CoquiEngine: Error initializing main coqui engine model: PytorchStreamReader failed reading zip archive: failed finding central directory RuntimeError: PytorchStreamReader failed reading zip archive: failed finding central directory Process Process-1:
Aitrepreneur
oh jeez that's a lot of errors...when exactly does that happen? Was the initial install successful?
Skuttlebot
Issue with getting Coqui (xtts) to work with Tavern/OpenAI 1) used the 1 click to install OpenAi 2) used the 1 click to install Sillytavern. 3) Verified coms work between the two. 4) Attempt to ingegrate xtts: Get this: E:\AI\SillyTavern\xtts\venv\lib\site-packages\pydub\utils.py:170: RuntimeWarning: Couldn't find ffmpeg or avconv - defaulting to ffmpeg, but may not work warn("Couldn't find ffmpeg or avconv - defaulting to ffmpeg, but may not work", RuntimeWarning) > Using model: xtts CoquiEngine: Error initializing main coqui engine model: PytorchStreamReader failed reading zip archive: failed finding central directory Traceback (most recent call last): File "E:\AI\SillyTavern\xtts\venv\lib\site-packages\xtts_api_server\RealtimeTTS\engines\coqui_engine.py", line 289, in _synthesize_worker tts.load_checkpoint( File "E:\AI\SillyTavern\xtts\venv\lib\site-packages\TTS\tts\models\xtts.py", line 760, in load_checkpoint checkpoint = self.get_compatible_checkpoint_state_dict(model_path) File "E:\AI\SillyTavern\xtts\venv\lib\site-packages\TTS\tts\models\xtts.py", line 710, in get_compatible_checkpoint_state_dict checkpoint = load_fsspec(model_path, map_location=torch.device("cpu"))["model"] File "E:\AI\SillyTavern\xtts\venv\lib\site-packages\TTS\utils\io.py", line 54, in load_fsspec return torch.load(f, map_location=map_location, **kwargs) File "E:\AI\SillyTavern\xtts\venv\lib\site-packages\torch\serialization.py", line 993, in load with _open_zipfile_reader(opened_file) as opened_zipfile: File "E:\AI\SillyTavern\xtts\venv\lib\site-packages\torch\serialization.py", line 447, in __init__ super().__init__(torch._C.PyTorchFileReader(name_or_buffer)) RuntimeError: PytorchStreamReader failed reading zip archive: failed finding central directory Process Process-1: Traceback (most recent call last): File "C:\Program Files\Python310\lib\multiprocessing\process.py", line 314, in _bootstrap self.run() File "C:\Program Files\Python310\lib\multiprocessing\process.py", line 108, in run self._target(*self._args, **self._kwargs) File "E:\AI\SillyTavern\xtts\venv\lib\site-packages\xtts_api_server\RealtimeTTS\engines\coqui_engine.py", line 289, in _synthesize_worker tts.load_checkpoint( File "E:\AI\SillyTavern\xtts\venv\lib\site-packages\TTS\tts\models\xtts.py", line 760, in load_checkpoint checkpoint = self.get_compatible_checkpoint_state_dict(model_path) File "E:\AI\SillyTavern\xtts\venv\lib\site-packages\TTS\tts\models\xtts.py", line 710, in get_compatible_checkpoint_state_dict checkpoint = load_fsspec(model_path, map_location=torch.device("cpu"))["model"] File "E:\AI\SillyTavern\xtts\venv\lib\site-packages\TTS\utils\io.py", line 54, in load_fsspec return torch.load(f, map_location=map_location, **kwargs) File "E:\AI\SillyTavern\xtts\venv\lib\site-packages\torch\serialization.py", line 993, in load with _open_zipfile_reader(opened_file) as opened_zipfile: File "E:\AI\SillyTavern\xtts\venv\lib\site-packages\torch\serialization.py", line 447, in __init__ super().__init__(torch._C.PyTorchFileReader(name_or_buffer)) RuntimeError: PytorchStreamReader failed reading zip archive: failed finding central directory 5) Attempts to fix the issues using the 2-click installer into the OpenAI folder. Breaks OpenAi. 6) No proffit. =( 7) Grovels for assistance.
Aitrepreneur
You can use models that have more context size limits than others like the mistral based models but keep in mind that the bigger the context length the more vram it uses
Ryan Clancy
That's fair enough! Any tips on getting the longest possible context length limit?
Aitrepreneur
Simple answer: you can't. It's a context length limit issue. All llm have a context length limit (imagine a virtual memory) that gets filled up over time and when it reaches the model's limit, it will start erasing older information to make space for the new information, so it cannot remember "everything" you say. It's a current limitation of any Ai models, that a lot of researchers are trying to solve, you need to wait for the tech to get better basically, that's it.
Ryan Clancy
A general question to anyone who may see it. Over time it seems these break down and conversations kinda collapse, even with excellent settings and source material. Is there a way to take an existing conversation threat, and kinda.... Start fresh from that point? Not even just for role play, but for example, if I were writing a novel and conversing with an AI through the process, how could I be sure that the conversation persisted, long term, even over months?
Aitrepreneur
Dm me some screenshots on discord and your pc specs
Thomas Heshe
I tried to install it again from the documents folder. I basically get the same errors. I did have to restart the installer 3 times, but it didn't seem like an error. Otherwise I didn't notice anything strange during install. It said Node.js is already installed. Could there be a problem with that?
Ryan Clancy
That loaded. Used a 13b one instead cause i like pushing my luck.π Many thanks, Overlord!
Ryan Clancy
I'll give that a try. They awq were working fine for me for hours... And suddenly.... Nothin. π€·π»ββοΈ Thanks. I'll report back on results.
Aitrepreneur
nope, but don't use awq format, now I advise against it, too many issues for most people, use the gptq model format with exllama compatible loader, the gptq-4bit-32g-actorder_True so for example, a model I recommend: https://huggingface.co/TheBloke/Xwin-MLewd-7B-V0.2-GPTQ put this in the model field: TheBloke/Xwin-MLewd-7B-V0.2-GPTQ:gptq-4bit-32g-actorder_True dl it and use the exllama loader
Ryan Clancy
Rtx3080, and I've tried yarn mistral 7b awq solar 10.7b instruct awq Llama2 13b awq It was just running all three of them fine a few minutes before. Perhaps, your question implies that I'm lacking in Vram? , which could be cause I've been running Stable Diffusion at the same time. π Maybe I'll just try a reboot?
Aitrepreneur
what gpu you got and what model did you try to load?
Aitrepreneur
are you sure the install was done correctly? Seems like a few things went wrong. Have you tried installing it in another folder? This time don't install it at the root of C drive but somewhere else (desktop, or documents, etc)
Ryan Clancy
Anyone got any idea why I suddenly can't load any models into Oobabooga? I just get a huge text demp when I try. the most relevant part seems to be " File "G:\OOBABOOGA\text-generation-webui\installer_files\env\Lib\site-packages\torch\nn\modules\module.py", line 1158, in convert return t.to(device, dtype if t.is_floating_point() or t.is_complex() else None, non_blocking) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ NotImplementedError: Cannot copy out of meta tensor; no data! " What did I break, and how did I break it?
Thomas Heshe
I used the 1-click install, and it seemed to install fine. But when I run the start.bat, I get these error messages: C:\SillyTavern>pushd C:\SillyTavern\ C:\SillyTavern>call npm install --no-audit 'npm' is not recognized as an internal or external command, operable program or batch file. C:\SillyTavern>node server.js C:\SillyTavern\server.js:164 app.use('/proxy/:url(*)', async (req, res) => { ^ SyntaxError: Unexpected token ( at createScript (vm.js:56:10) at Object.runInThisContext (vm.js:97:10) at Module._compile (module.js:549:28) at Object.Module._extensions..js (module.js:586:10) at Module.load (module.js:494:32) at tryModuleLoad (module.js:453:12) at Function.Module._load (module.js:445:3) at Module.runMain (module.js:611:10) at run (bootstrap_node.js:387:7) at startup (bootstrap_node.js:153:9) C:\SillyTavern>pause Press any key to continue . . .
Vinicius Appel
Get error while install xtts
Aitrepreneur
Sure try this one: https://we.tl/t-NzUfKJARyJ
Aitrepreneur
Try this installer: https://we.tl/t-NzUfKJARyJ
Chu
I keep getting the "restart installer" error. I saw that you offered another bat file for setup but the wetransfer link for that expired
Link Hearth
Hello, first thanks for your awsome work. I have the issue with the miniconda installing done, reset again and again, can you help me please
Aitrepreneur
GPU is always better
Solar Zaffiro
I have a Ryzen 9 5900x 12 core and a RTX 3080 GPU... which one would run Sillytavern better?
Aky
done, I sent it to you, I hope you received it and you can help me
Aitrepreneur
Just dm me here on patreon and use imgur for hosting screenshots
Aky
how can i contact you?? I'm new on patreon and I don't really know.
Aitrepreneur
dm me more info, full pc specs and screenshots of every webui application running and textgen webui also
Aky
I tried to use SillyTavern, but when I try to talk in the chat, a notification "API Error Internal Server Error" I did the same as in the video, I don't know what else to do. Can you help me?
Aitrepreneur
you can try this model: https://huggingface.co/TheBloke/Xwin-MLewd-7B-V0.2-GPTQ
Aitrepreneur
you can try this one: https://we.tl/t-7A4MqmxScq
Sam2es
Please make a perma link, this is expired again :(
Aky
I have an RTX 3060 laptop can you recommend something that works well for my graphics card? 6gb vram
Sam2es
I keep getting "restart installer" only to be told to do it again over and over
i tried to download manually than i tried to download through one click and moved it to its own folder, i keep getting this and having it close after i press any button What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 1 [INFO] Winget is already installed. [INFO] winget already exists in PATH. [INFO] Git is already installed. [INFO] Miniconda3 is already installed. [INFO] Execution Alias for python.exe was already removed. [INFO] Execution Alias for python3.exe was already removed. [WARN] Node.js is not installed on this system. [INFO] Installing Node.js using Winget... Found an existing package already installed. Trying to upgrade the installed package... No available upgrade found. No newer package versions are available from the configured sources. [INFO] Node.js installed successfully. Please restart the Installer. Press any key to continue . . .
Aitrepreneur
I like this one: https://huggingface.co/TheBloke/LLaMA2-13B-Psyfighter2-GPTQ
Aitrepreneur
try running the latest nvidia drivers and disabling system memory fallback: https://nvidia.custhelp.com/app/answers/detail/a_id/5490/~/system-memory-fallback-for-stable-diffusion
Memphisto ImOne
Whats one/some of the best Models for NSFW Roleplay on a 4080? So 16 Gb of VRAM?
Asbu
I've been struggling with really absurd response times, sometimes up to 700 seconds. I've fiddled with the settings, and it's brought it down to like 200 on average, but so far it just seems like I've just missed something. Win 11 Pro Processor AMD Ryzen 9 5900HS with Radeon Graphics 3.30 GHz Nvidia Geforce 3080 laptop gpu Installed RAM 32,0 GB
Elijah Matthew Moses
Not sure what this means but when I go to launch coqui_tts after using the one click installer from the patreon page, it gives me this error: LLVM ERROR: Symbol not found: __svml_cosf8_ha Please, how does one fix this? Windows 10 Pro Processor Intel(R) Core(TM) i9-9900K CPU @ 3.60GHz 3.60 GHz Nvidia GeForce RTX 3090 Installed RAM 128 GB
Nick K
Something to run on the Mac would be great for me as well
Aitrepreneur
hi dm me exactly when does that happen? What model are you using + pc specs?
Tonic69
Lazza
that didnt work....but i did get it to work miraculously. i watched your video again and did this: python -m venv venv venv\Scripts\activate pip install xtts-api-server pip install torch==2.1.1+cu118 torchaudio==2.1.1+cu118 --index-url https://download.pytorch.org/whl/cu118 staight after it finished installing i didnt close the command windows and it gave me a prompt to install an update in the command window. i just copied a pasted what was there and it updated and works!
does this installer work for mac, and if not is it possible to get one for mac?
pixelgeniee
yes this worked, thank you :)
Aitrepreneur
You can try this one: https://we.tl/t-slzg12dLKy
pixelgeniee
it says minoconda installed successfully and then asks to restart installer. Upon restarting installer again says the same thing..pls help
Randy
I'm getting the Error: Cannot find module 'stream/promises' when I try to run the start.bat file Edit: This was due to me running a previous version of node through NVM. If anyone else has this issue just use NVM to switch the version of node.js you are using.
Aitrepreneur
you can try this one, can't promise it will work since I don't have any apple products but try it out and let me know: https://we.tl/t-QtvMv1KLUW
Thomas Levenhagen
I wouldn't mind one for Mac if it's possible. In the mean time I'll be trying to fuddle my way through it.
Bane
thanks so much. that helped
Aitrepreneur
try completely uninstalling python from your computer then try relaunching/reinstalling Sillytavern in a new folder or reinstall python manually, just don't forget to check "ADD TO PATH" checkbox: https://www.python.org/ftp/python/3.10.6/python-3.10.6-amd64.exe
Aitrepreneur
I could make one if people ask for it yes, most people usually use windows to run Ai tools though
Aitrepreneur
I'll make more sillytavern content soon ;)
Aitrepreneur
here: https://we.tl/t-tIZ5KdKGiZ
Aitrepreneur
I mean that depends on the character's personality and how YOU respond, if you respond with short sentences, after a while the character will start doing the same, so make your answer longer too
Lazza
I've completely broken it trying. now silly tavern wont install and saying "Found an existing package already installed. Trying to upgrade the installed package..."
Mingrui
I have the same issue. The python version seems to be old. But after i installed the newest version with the python-3.12.1-amd64.exe it still is the old version in that error. I guess it didn't land in the correct folder. Which ever that folder might be.^^
Lazza
Hi, I have and error saying '2024-01-08 21:18:50.058 | ERROR | xtts_api_server.modeldownloader:install_deepspeed_based_on_python_version:64 - Unsupported Python version on Windows.' when trying to use xtts
SurfaceTension
Is there a non-windows version of the one-click SillyTavern installer (e.g. for MacOS/Linux)?
todd beebe
Hi can you do a tutorial on Video chat with AI characters - SillyTavern this sometime I tried to follow the video that MustacheAI did with the sillytavern and extras i got from the one click installer from this patron and its not working. i tried it from scratch too also cant get it work if a tutorial would be possible sometime thanks
Feri
I have the same reinstall issue as many people before, I tried to download the solution from this link: https://we.tl/t-rtY2FmgmLY But it's writing: Sorry, this transfer has expired and is not available any more.Can you reupload again?
hwa
Thanks! This helped, now my response times are about 1-2 seconds. The responses themselves are fairly short though, only about a sentence. Is there a way to make them longer? I tried changing Response (Tokens) but it doesn't seem to actually do anything >.<
Aitrepreneur
Dm me the full error log and pc specs
Aitrepreneur
indeed that's a bit slow, you shouldn't actually need to do something, it's probably the model you are using, try this one: TheBloke/LLaMA2-13B-Tiefighter-GPTQ:gptq-4bit-32g-actorder_True copy and paste that line in the model field, dl it and then use the exllama loader
TreFeKs Gaming
Yeah already did that afterwards still doesnt work :(
hwa
What's the average response time? I set this all up without issue, but the response times around between 18-30seconds for me (Running a 3090 RTX/AMD 5950X and KoboldAI Psyfighter). I'm new to this and I don't know how to layer my GPU or if I even should? I looked around for an in-UI option to change layers but I have no idea what I'm doing. Any advice to speed up response times?
Aitrepreneur
you should have a sillytavern folder, that's where the start file is located
Aitrepreneur
there might be an issue with your folder path with the # character, try installing it in a folder with no spaces or special character, something like: C:\Users\USERNAME\Desktop\AI
Neker Play
I have used the 1click installer, it has gone all good but i have a folder called sillyTavern-extras with inside another folder called sillyTavern-extras, bue there are no start.bat files, only start.sh, i am on windows
TreFeKs Gaming
When i try to start the silly tavern start.bat it doesnt work npm WARN ws@8.16.0 requires a peer of bufferutil@^4.0.1 but none was installed. npm WARN ws@8.16.0 requires a peer of utf-8-validate@>=5.0.2 but none was installed. npm WARN node-fetch@2.7.0 requires a peer of encoding@^0.1.0 but none was installed. npm WARN pkg@5.8.1 requires a peer of node-notifier@>=9.0.1 but none was installed. npm WARN sillytavern@1.11.1 No description npm ERR! Windows_NT 10.0.19045 npm ERR! argv "C:\\Program Files (x86)\\nodejs\\node.exe" "C:\\Program Files (x86)\\nodejs\\node_modules\\npm\\bin\\npm-cli.js" "install" "--no-audit" npm ERR! node v6.10.3 npm ERR! npm v3.10.10 npm ERR! path D:\#Anwendungen\AI\SillyTavern\node_modules\.staging\@types\http-cache-semantics-2294ad44\package.json npm ERR! code ENOTDIR npm ERR! errno -4058 npm ERR! syscall open npm ERR! ENOENT: no such file or directory, open 'D:\#Anwendungen\AI\SillyTavern\node_modules\.staging\@types\http-cache-semantics-2294ad44\package.json'
Aitrepreneur
send me the full error log and pc specs
Admiral Squirrel
What am I doing/done wrong?
Admiral Squirrel
67 error code ELIFECYCLE 68 error errno 1 69 error sillytavern@1.11.1 postinstall: `node post-install.js` 69 error Exit status 1 70 error Failed at the sillytavern@1.11.1 postinstall script. 70 error This is probably not a problem with npm. There is likely additional logging output above. 71 verbose exit [ 1, true ]
Admiral Squirrel
Hi, I'm using the 1 click install and I'm getting this error code
Aitrepreneur
dm me the full error log
Alex Nick
I keep getting an error "Windows cannot find "Microsoft.DesktopAppInstaller"
Aitrepreneur
there are new things getting added all the time so there is never a lack of subjects that's for sure.... π
cool1
What might make a good video is if it went into detail on the Vector Storage option in SillyTavern and the Smart Context extension. Is one better than the other or does it depend? Is it best not to use them both at the same time? The Vector Storage option also has an "enabled for files" option too so I assume you could ask it for info about what's in different document files. But a video explaining it all would be helpful I think, including the pros and cons of using them with different settings.
Aitrepreneur
Weird, never saw that happens, I don't use firefox (and neither should you lol) but hey as long as it works for you
Aitrepreneur
Not sure I get the opposite effect tbh, maybe try another model/character and check the generation parameters if they are similar to the ones you use in text gen but it's the first time I hear someone having worse perf compared to the normal text gen
cool1
I tried installing sillytavern and clicking on "start.bat" or "UpdateAndStart.bat" says: Running from source SillyTavern 1.11.1 'release' .. Launching... SillyTavern is listening on: http://127.0.0.1:8000/ but the it shows a black webpage (even after pressing F5 to reload) and isn't showing any sillytavern UI options. Is there anything I can fix it with? Do I need to change anything in the config file? edit: I just tried it in Chrome icognito and it displayed the "welcome to sillytavern" text in the web page, after loading that URL, but it's not working in Firefox (just shows the blank/black page) and it's Firefox that I want to run it on. In Noscript I've enabled scripts. edit2: It's now worked (showing stuff on the local web page URL) when I enabled cookies for the session for that local URL in Firefox (normally I only enable cookies for sites I specify). So maybe I need to permanently enable cookies for that local URL in Firefox. I'm not sure why they're needed for a local URL but I'll try that if needed.
Jerry
The conversations in Silly Tavern doesn't seem as engaging as the text gen webui. A character imported seems bland in comparison in Silly Tavern. What am I doing wrong.
DuFFy
Fault repaired steps: 1. run latest Node.js installer, 2. select repair, this added missing files, 3. PROFIT!!!!
Aitrepreneur
dm me the full error log
DuFFy
HELP~! Performed the install manually ST was running fine. Changed a language model in ooga booga to a cpu based model from a gpu model to free up gpu for Xtts (2080 ti 11g). Wroked then wanted to go back to old model. Nope it then died. Decided to do a fresh install using "one click installer" Would not run server.js? So I git cloned... same result whenever i try to start... I don't know what to do please help!
Dropped Dice Gaming
I tried the basic ones that come with silly tavern and downloaded a maid one from the Chub site you recommended. I tried running them with the tie fighter 13b and the little 7b AIs you mention in the video.
cuefy
in the video you mentioned the step regarding this problem wasn't necessary because im using the 1 step installer. When i launch the webui i dont have the same menu as you in the video, it takes me to sillytavern straight away.
omegazone
Thanks! It works flawless when it comes to textwriting and generating time. Made my day but it cant even do simple math :P not that it matters in this case Thank you for responding so quickly ^^
RΓ©mi Snd
Perfect, everything is fine now.
Ray Kliviz
I did find the solution for this, when you have similar error like me make sure you selected "openai" in "session" settings, when using text-gen. After this you should be able to connect again, I thought it would just automaticly select whenever I clicked on in previously, but it didnt, just make sure whenever you are trying to run it again that "openai" is selected. Plus make sure you unload any models you use, othervise you will use multiple models and it will use all your CPU or GPU memory and will show you error that your memory is full.
Aitrepreneur
then try this one: https://we.tl/t-1wmmU9qgHb
RΓ©mi Snd
When i run the file i get : \ [INFO] Miniconda3 installed successfully. Please restart the Installer. Over and over again.
Aitrepreneur
it's usually a model issue, try this one for your GPU: TheBloke/Pygmalion-2-7B-GPTQ:gptq-4bit-32g-actorder_True copy and paste this line in the model field, dl it and then use the exllama loader to load the model
omegazone
Why does the ai only spill out random letters to my question? Is it because i "Only" have a 3070 (yes compared to 24GBVRAM my 8GB arent alot) if that would be the case what is the next best model to use instead LLaMA2-13B-Tiefighter? Besides that was the video super easy to follow and your one click installer is a time saver
Aitrepreneur
this basically means that it cannot find the api server so make sure that the text gen is running correctly with a model loaded and that you don't have any other extension enabled except openai. Also you can try doing the install all over again in a new folder if everything else is working fine
Aitrepreneur
have you loaded a model in oobabooga text gen?
Aitrepreneur
I would need the whole log to be sure but nothing bad if you just install visual C++14 and then relaunch the full installer again in a new folder
Ray Kliviz
I have this error, I know it was shown before, but I still havent seen way to fix it, I checked around the internet no luck. I maked sure ports is not blocked, as well disabled my antivirus and windows firewall, still no luck, I tryed used different models and still the same error. At this point I tried everything that I can think of, if there is any ideas how to fix that let me know, plus I made sure that port is not in use as well, and nope its not. "Trying to connect to API: { api_server: 'http://127.0.0.1:5000', api_type: 'ooba', legacy_api: false } FetchError: request to http://127.0.0.1:5000/v1/models failed, reason: connect ECONNREFUSED 127.0.0.1:5000 at ClientRequest. (E:\AIMODEL\AITAVERN\AITAVERN\SillyTavern\node_modules\node-fetch\lib\index.js:1505:11) at ClientRequest.emit (node:events:514:28) at Socket.socketErrorListener (node:_http_client:495:9) at Socket.emit (node:events:514:28) at emitErrorNT (node:internal/streams/destroy:151:8) at emitErrorCloseNT (node:internal/streams/destroy:116:3) at process.processTicksAndRejections (node:internal/process/task_queues:82:21) { type: 'system', errno: 'ECONNREFUSED', code: 'ECONNREFUSED'"
Marc De Vries
When i try to connect with the server URL it keeps saying no connection, in the sillytavern program it also says that Models endpoint is offline.
John Harris
I'm a good portion of the way through the 1-click install, but I got a big chunk of red error text in the middle. It says "ERROR: Failed building wheel for tts" followed shortly after by "ERROR: Could not build wheels for tts which use PEP 517 and cannot be installed directly." This appears next to another error saying that "Microsoft Visual C++ 14.0 or greater is required" along with a link to download it. My questions for you: (1) Is that the likely cause of the error? (2) If so, do I need to delete everything and restart the 1-click install after installing/upgrading to VS 14? Or is just re-running the installer afterward fine?
Aitrepreneur
make sure that you have loaded a model in the text gen, but at least if it's green it means it's correctly connected to text gen
RMR
Under the server url and connect it shows a green dot with none. What am I doing wrong to connect to the Bloke model? I get internal server error when i text.
Aitrepreneur
it's a false positive, buildtools is a microsoft tool
Aitrepreneur
I think it should work but it needs internet to download the xtts model first
Aitrepreneur
Watch my video, I show everything there
Aitrepreneur
might be a bit too low but you can run it on your cpu, with text gen also using gguf models
Wag Mods
I got a malware detection by Bitdefender for vs_buildtools.exe? Any reason why that would happen?
Dougomite
Thanks so much that worked perfectly. I went from averaging 1-2 tokens/sec to ~12tokens/sec. I had to also lower the max_seq_len when loading the model. Seems like my PC was dipping into the System memory just a bit but that was killing the speed.
Ljtrix
Does it not allow you to run XTTS offline? I tried using Xtts without an internet connection and it didn't work.
cuefy
I used this to install silly tavern, but im not sure how to get the API url now
Boogiepop Phantom
GTX 970. im trying to run tavern but looks like too much for my pc lol
TomiTom1234
OK finally it worked.. I had to delete the old SillyTavern, then re-installed everything again.. The bat file installed deepspeed this time after launching xtts.bat file with no problems. You mentioned to "accept the terms and conditions to be able to download the xttsv2 model", Where exactly and in which step should I do that? Although everything is working fine and the voice play.. Thank you for your help!
TomiTom1234
I did all what you told me to do, uninstalled python then re-installed from your link, re-installed SillyTavern (I didn't delete the old folder though), and now the xtts bat file won't start, it opens then closes in milli-second, tried starting it as administration, no luck. I am screwed now more than ever lol. P.S: the EXTRAS.bat starts, but not XTTS bat file, I redownloaded it and replaced the old one with it, same, doesn't start
Aitrepreneur
No need to reinstall deepspeed, this error shouldn't even appear to begin with. I helped someone else fixe this recently on a laptop, what we did is completely uninstall python using the installer https://www.python.org/ftp/python/3.10.6/python-3.10.6-amd64.exe (just select uninstall) then relaunch it again and this time don't forget to check the ADD TO PATH checkbox and then install it. We then redid the full sillytavern reinstall with the 1 click installer (sometimes having to relaunch it multiple times) and then relaunching the launchxtts bat file did the trick, don't forget here to accept the terms and conditions to be able to download the xttsv2 model. NO need to do anything else
TomiTom1234
Sadly the same errors when re-installing deepspeed, and when running xtts.bat with --deepspeed in it. The error when trying to install deepspeed inside the xtts folder is as the SS before which is: (venv) E:\AI\SillyTavern\xtts>pip install deepspeed Collecting deepspeed Using cached deepspeed-0.12.6.tar.gz (1.2 MB) Installing build dependencies ... done Getting requirements to build wheel ... error error: subprocess-exited-with-error Γ Getting requirements to build wheel did not run successfully. β exit code: 1 β°β> [20 lines of output] [WARNING] Unable to import torch, pre-compiling ops will be disabled. Please visit https://pytorch.org/ to see how to properly install torch on your system. [WARNING] unable to import torch, please install it if you want to pre-compile any deepspeed ops. DS_BUILD_OPS=1 Traceback (most recent call last): File "e:\ai\sillytavern\xtts\venv\lib\site-packages\pip\_vendor\pyproject_hooks\_in_process\_in_process.py", line 353, in main() File "e:\ai\sillytavern\xtts\venv\lib\site-packages\pip\_vendor\pyproject_hooks\_in_process\_in_process.py", line 335, in main json_out['return_val'] = hook(**hook_input['kwargs']) File "e:\ai\sillytavern\xtts\venv\lib\site-packages\pip\_vendor\pyproject_hooks\_in_process\_in_process.py", line 118, in get_requires_for_build_wheel return hook(config_settings) File "C:\Users\abu-m\AppData\Local\Temp\pip-build-env-q1iokgn7\overlay\Lib\site-packages\setuptools\build_meta.py", line 325, in get_requires_for_build_wheel return self._get_build_requires(config_settings, requirements=['wheel']) File "C:\Users\abu-m\AppData\Local\Temp\pip-build-env-q1iokgn7\overlay\Lib\site-packages\setuptools\build_meta.py", line 295, in _get_build_requires self.run_setup() File "C:\Users\abu-m\AppData\Local\Temp\pip-build-env-q1iokgn7\overlay\Lib\site-packages\setuptools\build_meta.py", line 480, in run_setup super(_BuildMetaLegacyBackend, self).run_setup(setup_script=setup_script) File "C:\Users\abu-m\AppData\Local\Temp\pip-build-env-q1iokgn7\overlay\Lib\site-packages\setuptools\build_meta.py", line 311, in run_setup exec(code, locals()) File "", line 147, in AssertionError: Unable to pre-compile ops without torch installed. Please install torch before attempting to pre-compile ops. [end of output] note: This error originates from a subprocess, and is likely not a problem with pip. error: subprocess-exited-with-error Γ Getting requirements to build wheel did not run successfully. β exit code: 1 β°β> See above for output. note: This error originates from a subprocess, and is likely not a problem with pip. And also same error whjen starting it. Any idea? Maybe something about Torch? Which I have no idea what it is lol.
Aitrepreneur
try completely uninstalling python from your computer, just download this: https://www.python.org/ftp/python/3.10.6/python-3.10.6-amd64.exe and choose uninstall. Then rerun the python installer manually again and don't forget to check the ADD TO PATH checkbox in the beginning then try again relaunching the launchxtts bat file
TomiTom1234
I decided to install "Deepspeed" so that the sound runs faster (I had to remove it from the bat file to let it run), I followed steps you mentioned to other subscriber here which are: "type cmd press enter, this will bring the command prompt window and there type venv\Scripts\activate then press enter and then type pip install deepspeed then press enter" I followed the steps, it installed Deepspeed but some errors appeared also in the command prompt. Can you please help me finish the installing of DeepSpeed without errors? Here is the SS of the Command Prompt window: https://ibb.co/wCCVH9Q And here is the SS of the XTTS.bat file with the "Deepspeed" paramater added in the bat file, it seems somehow "unsupported pythong version" or something like that? I have no idea: https://ibb.co/Q97zqTf Much appreciate your time and effort!
Aitrepreneur
how low end are we talking about?
Aitrepreneur
you probably have another text gen instance open
Aitrepreneur
it's usually the model's problem though, what model have you used?
Aitrepreneur
try this: go the xtts folder, click on the folder path type cmd press enter, this will bring the command prompt window and there type venv\Scripts\activate then press enter and then type pip install deepspeed then press enter if that still doesn't work for some reason, then edit the launchxtts.bat file and remove the deepspeed argument
Aitrepreneur
Dm me what error you get
Aitrepreneur
No worries have fun ;)
Aitrepreneur
For your model make sure that you are running a 7B model since you have only 8GB of Vram and for the speed make sure that you have the latest nvidia drivers and you can also try disabling system memory fallback and see if that works better: https://nvidia.custhelp.com/app/answers/detail/a_id/5490/~/system-memory-fallback-for-stable-diffusion
Boogiepop Phantom
Any advices for low end pc?
Bane
Yeah no chance. Sadly I can't use it.
John Smith
I found an error message. "[Errno 10048] error while attempting to bind on address ('127.0.0.1', 5000): only one usage of each socket address (protocol/network address/port) is normally permitted." How do I fix this?
Dropped Dice Gaming
I got everything set up but the AI's just answer with random gibberish that makes no sense. I tried a few different models but same thing. Any ideas how to fix them?
caidicus
I think this is currently not working due to an issue with deepspeed or something?
John Smith
I couldn't get the Silly Tavern to Connect to the Webui. I copied the "5000" url and deleted the "/api" as the video said, but it would never connect.
Fatty McFatass
thanks for your help :)
Bane
I only have Windows Defender and the Port should be open.
TomiTom1234
Oh my, I feel bad. I bothered you a lot of times. But There is no voice coming from Sandra, Only text with no voice, although I enabled the TTS and everything, and started the XTTS bat file. Here is the SS: https://ibb.co/SKGKm95 And here is the SS of the command prompt window of the XTTS Launcher: https://ibb.co/Mnz9BTg Thank you very much for all your help! **EDIT**: I figured out that I have to tick the βAuto Generationβ under the "TTS" tab.
TomiTom1234
Thank you this worked now.
Remgrandt
putting the cache folder in place seemed to fix it (I didn't have a cache folder at all prior).
Remgrandt
The only AV I have is windows defender. I notice in the window for LaunchEXTRAS.bat, it says it launched: Initializing text classification pipeline with model nateraw/bert-base-uncased-emotion which does not match the ST log: Initializing transformers.js pipeline for task text-classification with model Cohee/distilbert-base-uncased-go-emotions-onnx
Dougomite
Great video! I had no idea it was so doable to set up an AI locally. Anywhere I can find out the generation times with each model or GPUs? I'm using a 2080 Super and the speed isn't seeming great, are there smaller faster models to try or how much better of a card would get decent generation times.
Aitrepreneur
you might have an antivirus blocking the connection, it has trouble downloading the model from the web. Try disabling it and try again or you can try downloading this archive: https://we.tl/t-aJkXiiy0o7 extract it and place the cache folder inside the sillytavern folder
Remgrandt
I am getting this message when I enable Character Expressions, can you help? Initializing transformers.js pipeline for task text-classification with model Cohee/distilbert-base-uncased-go-emotions-onnx TypeError: response.body.getReader is not a function at readResponse (file:///C:/Users/remgrandt/Documents/SillyTavern/SillyTavern/node_modules/sillytavern-transformers/src/utils/hub.js:572:34) at getModelFile (file:///C:/Users/remgrandt/Documents/SillyTavern/SillyTavern/node_modules/sillytavern-transformers/src/utils/hub.js:493:26) at processTicksAndRejections (node:internal/process/task_queues:96:5) at async getModelJSON (file:///C:/Users/remgrandt/Documents/SillyTavern/SillyTavern/node_modules/sillytavern-transformers/src/utils/hub.js:542:18) at async Promise.all (index 1) at async loadTokenizer (file:///C:/Users/remgrandt/Documents/SillyTavern/SillyTavern/node_modules/sillytavern-transformers/src/tokenizers.js:56:16) at async Function.from_pretrained (file:///C:/Users/remgrandt/Documents/SillyTavern/SillyTavern/node_modules/sillytavern-transformers/src/tokenizers.js:3777:48) at async Promise.all (index 0) at async loadItems (file:///C:/Users/remgrandt/Documents/SillyTavern/SillyTavern/node_modules/sillytavern-transformers/src/pipelines.js:2110:5) at async pipeline (file:///C:/Users/remgrandt/Documents/SillyTavern/SillyTavern/node_modules/sillytavern-transformers/src/pipelines.js:2056:19)
Rina
Ah, I chose the AMD auto install and ran it
Aitrepreneur
Rewatch the video again, you have chose the horde api instead of text gen
Rina
"No Horde model selected or the selected models are no longer available. Please choose another model"
Aitrepreneur
try this one then: https://we.tl/t-rtY2FmgmLY
Aitrepreneur
No worries
Aitrepreneur
as I said sillytavern does not run any local models, it's just a skin. You need to have watched the text gen install video where I install the local model
Aitrepreneur
never saw that error before but it could be an antivirus blocking the connection for some reason, disable it and try again
Aitrepreneur
yeah you have a bunch of extensions activated in oobabooga, disable them all except open ai then apply flags/extension and restart ui, and reconnect to sillytavern.
Fatty McFatass
I m pretty new to this and am a little confused, when i use the script it says - What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 1 [WARN] Winget is not installed on this system. [INFO] Installing Winget... % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0 100 233M 100 233M 0 0 47.0M 0 0:00:04 0:00:04 --:--:-- 55.9M [INFO] Winget installed successfully. The operation completed successfully. [INFO] winget added to PATH. [INFO] Git is already installed. [WARN] Miniconda3 is not installed on this system. [INFO] Installing Miniconda3 using Winget... 'winget' is not recognized as an internal or external command, operable program or batch file. [INFO] Miniconda3 installed successfully. Please restart the Installer. Press any key to continue . . ., if i open it again its rinse and repeat ?
oxymoron02
Ahh I think this is the answer I was looking for. I assumed Aitrepreneur's BAT had installed a model locally as part of the process.
MichaΕ Ε»yto
Ok... Solved it myself. Your installer is golden ;) (the problem was that I had a second instance of ST and had to remove it and clean my system from old libs)
Karsten G
Works perfectly fine offline. You need to use your own mdels locally. It does not wokr offline when connected to the coud
Michael woods
figured it out had to run it a third time for some reason my apologies. On my end.
Michael woods
This is the second "1click" that refuses to work for me. it installs then askes me to restart the installer.
oxymoron02
When I disconnect my network, this stops working. Try it for yourself! (I meant SillyTavern, not your installer).
Aitrepreneur
What are you talking about?
oxymoron02
So why does this only work when your network is enabled? I'm deleting it now because that's sus as fuck.
Bane
same error
TomiTom1234
Here are screenshots: https://ibb.co/6Bn88pS https://ibb.co/HzgLVjr https://ibb.co/85YHbxq https://ibb.co/S64ffyP https://ibb.co/Gt31WdQ https://ibb.co/WBwQ3Ws Many thanks!
MichaΕ Ε»yto
I get the same error when running the script. (I tried to install it manually but i get the same error)
Andy Merz
trying to connect and getting main_api: 'textgenerationwebui', api_server: 'http://127.0.0.1:5000/', api_type: 'ooba', legacy_api: false } FetchError: request to http://127.0.0.1:5000/v1/models failed, reason: connect ECONNREFUSED 127.0.0.1:5000 at ClientRequest. (C:\Users\andy_\Downloads\SillyTavern\node_modules\node-fetch\lib\index.js:1505:11) at ClientRequest.emit (node:events:514:28) at Socket.socketErrorListener (node:_http_client:495:9) at Socket.emit (node:events:514:28) at emitErrorNT (node:internal/streams/destroy:151:8) at emitErrorCloseNT (node:internal/streams/destroy:116:3) at process.processTicksAndRejections (node:internal/process/task_queues:82:21) { type: 'system', errno: 'ECONNREFUSED', code: 'ECONNREFUSED' }
Akshay
Also, I'm not able to connect with text gen web UI from Collab to my local Tavern UI. Sent you the SS through doscord in DM's.
Akshay
shared.model, shared.tokenizer = load_model(selected_model, loader) output = load_func_map[loader](model_name) model, tokenizer = ExllamaModel.from_pretrained(model_name) model = ExLlama(config) with safe_open(path, framework = "pt", device = "cpu") as f: safetensors_rust.SafetensorError: Error while deserializing header: HeaderTooLarge Getting this error while loading the KoboldAI model in Text Gen Web UI
Stormy Adams
Teamwork my friend, thank you so much for the content and lesson design. I really appreciate it. Happy holidays to you as well.
Aitrepreneur
Damn yeah you're right, looks like I erased that 1s part during editing and didn't notice... thanks for the heads up
Aitrepreneur
you can use this colab: https://colab.research.google.com/github/oobabooga/text-generation-webui/blob/main/Colab-TextGen-GPU.ipynb to run the text gen webui, and connect with the api to a local install of sillytavern (just install sillytavern on your computer, it's just a UI so you don't need a big GPU to run it)
Stormy Adams
I noticed one other slight thing in the video. Within the video you mention placing the aqua character in the live2d folder within public. However the directory needs to be public/assets/live2d. ^^ Apologies if this is just me misunderstanding. But I wanted to post in the case this is helpful for others.
Akshay
Bro, just one help, if you can guide me on kaggle/ Collab repository for the same. how to do those. It would help a lot.
Stormy Adams
Wow, thank you so much for the quick reply. I hope you have a great rest of your day/night!
Aitrepreneur
hi, it's chub.ai
Aitrepreneur
try reinstalling it, it definitely didn't install correctly if you have errors when launching it
Stormy Adams
Hey there, this was an amazing video, and I'm very thankful for the one click install and the in depth guide! -- Ask: What is the website where you can download the Tavern AI character from? I didn't find it in the description of the latest video.
Jonathan Kinsman
NPM is not installed
Jonathan Kinsman
C:\Users\jkins\Desktop\NPCai\TavernAI\SillyTavern>node server.js 'node' is not recognized as an internal or external command, operable program or batch file.
Jonathan Kinsman
I think it installed right. but then when I click start i get:
Aitrepreneur
You don't have to run the command, the installer does everything already. xtts was already installed automatically during the sillytavern installation
MichaΕ Ε»yto
@Aitrepreneur: when I try running this command to install torch (as instructed) @Wojciech: Windows 10 Pro AMD Ryzen 5 3600 6-Core
Wojciech Lenartowicz
What's your OS, exactly? And the CPU architecture?
Aitrepreneur
when does that happen?
Demitrius Stone
this helped solve my problem, thank you!
Aitrepreneur
hi thanks, I've already updated the script for everyoneπ
MichaΕ Ε»yto
any suggestions? (venv) C:\AI\SillyTavern-MainBranch\xtts>pip install torch==2.1.1+cu118 torchaudio==2.1.1+cu118 --index-url https://download.pytorch.org/whl/cu118 Looking in indexes: https://download.pytorch.org/whl/cu118 ERROR: Could not find a version that satisfies the requirement torch==2.1.1+cu118 (from versions: none) ERROR: No matching distribution found for torch==2.1.1+cu118
Wojciech Lenartowicz
Glad it helped :) Yeah, I could leave an easier instruction, but for some reason I insisted on fixing the script so that you will have a working one for future usage ^^' EDIT: I've improved the comment. I hope it will be more helpful now if anybody still needs it :)
Aitrepreneur
send me screens of the extensions enabled in text gen and sillytavern
TomiTom1234
Oh, I really followed everything you said. Ready to respond to the character, enabled the XTTS voice, after sending the response, I get errors in the text-generation-webui command prompt window as follows: Output generated in 17.18 seconds (0.99 tokens/s, 17 tokens, context 1085, seed 381708695) Exception in ASGI application Traceback (most recent call last): File "E:\AI\text-generation-webui\installer_files\env\Lib\site-packages\uvicorn\protocols\http\h11_impl.py", line 408, in run_asgi result = await app( # type: ignore[func-returns-value] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\AI\text-generation-webui\installer_files\env\Lib\site-packages\uvicorn\middleware\proxy_headers.py", line 84, in __call__ return await self.app(scope, receive, send) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\AI\text-generation-webui\installer_files\env\Lib\site-packages\fastapi\applications.py", line 1106, in __call__ await super().__call__(scope, receive, send) File "E:\AI\text-generation-webui\installer_files\env\Lib\site-packages\starlette\applications.py", line 122, in __call__ await self.middleware_stack(scope, receive, send) File "E:\AI\text-generation-webui\installer_files\env\Lib\site-packages\starlette\middleware\errors.py", line 184, in __call__ raise exc File "E:\AI\text-generation-webui\installer_files\env\Lib\site-packages\starlette\middleware\errors.py", line 162, in __call__ await self.app(scope, receive, _send) File "E:\AI\text-generation-webui\installer_files\env\Lib\site-packages\starlette\middleware\cors.py", line 83, in __call__ await self.app(scope, receive, send) File "E:\AI\text-generation-webui\installer_files\env\Lib\site-packages\starlette\middleware\exceptions.py", line 79, in __call__ raise exc File "E:\AI\text-generation-webui\installer_files\env\Lib\site-packages\starlette\middleware\exceptions.py", line 68, in __call__ await self.app(scope, receive, sender) File "E:\AI\text-generation-webui\installer_files\env\Lib\site-packages\fastapi\middleware\asyncexitstack.py", line 20, in __call__ raise e File "E:\AI\text-generation-webui\installer_files\env\Lib\site-packages\fastapi\middleware\asyncexitstack.py", line 17, in __call__ await self.app(scope, receive, send) File "E:\AI\text-generation-webui\installer_files\env\Lib\site-packages\starlette\routing.py", line 718, in __call__ await route.handle(scope, receive, send) File "E:\AI\text-generation-webui\installer_files\env\Lib\site-packages\starlette\routing.py", line 276, in handle await self.app(scope, receive, send) File "E:\AI\text-generation-webui\installer_files\env\Lib\site-packages\starlette\routing.py", line 66, in app response = await func(request) ^^^^^^^^^^^^^^^^^^^ File "E:\AI\text-generation-webui\installer_files\env\Lib\site-packages\fastapi\routing.py", line 274, in app raw_response = await run_endpoint_function( ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\AI\text-generation-webui\installer_files\env\Lib\site-packages\fastapi\routing.py", line 191, in run_endpoint_function return await dependant.call(**values) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\AI\text-generation-webui\extensions\openai\script.py", line 114, in openai_completions response = OAIcompletions.completions(to_dict(request_data), is_legacy=is_legacy) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\AI\text-generation-webui\extensions\openai\completions.py", line 503, in completions return deque(generator, maxlen=1).pop() ^^^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\AI\text-generation-webui\extensions\openai\completions.py", line 388, in completions_common for a in generator: File "E:\AI\text-generation-webui\modules\text_generation.py", line 31, in generate_reply for result in _generate_reply(*args, **kwargs): File "E:\AI\text-generation-webui\modules\text_generation.py", line 109, in _generate_reply reply = apply_extensions('output', reply, state) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\AI\text-generation-webui\modules\extensions.py", line 229, in apply_extensions return EXTENSION_MAP[typ](*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\AI\text-generation-webui\modules\extensions.py", line 87, in _apply_string_extensions text = func(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^ File "E:\AI\text-generation-webui\extensions\coqui_tts\script.py", line 143, in output_modifier output_file = Path(f'extensions/coqui_tts/outputs/{state["character_menu"]}_{int(time.time())}.wav') ~~~~~^^^^^^^^^^^^^^^^^^ KeyError: 'character_menu' Any idea please?
Aitrepreneur
I changed it in the new one
TomiTom1234
Yes thank you! This one worked for me. The line: "winget install --id OpenJS.NodeJS.LTS -e --silent -y" had to be changed into : "winget install --id OpenJS.NodeJS.LTS -e --silent --accept-source-agreements --accept-package-agreements".
TomiTom1234
Yes thank you, this one wroked.
Aitrepreneur
try this one: https://we.tl/t-rtY2FmgmLY
TomiTom1234
I got you. It seems the installer didn't install it for me, I get error messages, here is the message I get: [INFO] Winget is already installed. [INFO] winget already exists in PATH. [INFO] Git is already installed. [INFO] Miniconda3 is already installed. [INFO] Execution Alias for python.exe was already removed. [INFO] Execution Alias for python3.exe was already removed. [WARN] Node.js is not installed on this system. [INFO] Installing Node.js using Winget... Windows Package Manager v1.6.3421 Copyright (c) Microsoft Corporation. All rights reserved. Argument alias was not recognized for the current command: '-y' Installs the selected package, either found by searching a configured source or directly from a manifest. By default, the query must case-insensitively match the id, name, or moniker of the package. Other fields can be used by passing their appropriate option. By default, install command will check package installed status and try to perform an upgrade if applicable. Override with --force to perform a direct install. usage: winget install [[-q] ...] [] The following command aliases are available: add The following arguments are available: -q,--query The query used to search for a package The following options are available: -m,--manifest The path to the manifest of the package --id Filter results by id --name Filter results by name --moniker Filter results by moniker -v,--version Use the specified version; default is the latest version -s,--source Find package using the specified source --scope Select install scope (user or machine) -a,--architecture Select the architecture --installer-type Select the installer type -e,--exact Find package using exact match -i,--interactive Request interactive installation; user input may be needed -h,--silent Request silent installation --locale Locale to use (BCP47 format) -o,--log Log location (if supported) --custom Arguments to be passed on to the installer in addition to the defaults --override Override arguments to be passed on to the installer -l,--location Location to install to (if supported) --ignore-security-hash Ignore the installer hash check failure --skip-dependencies Skips processing package dependencies and Windows features --ignore-local-archive-malware-scan Ignore the malware scan performed as part of installing an archive type package from local manifest --dependency-source Find package dependencies using the specified source --accept-package-agreements Accept all license agreements for packages --no-upgrade Skips upgrade if an installed version already exists --header Optional Windows-Package-Manager REST source HTTP header --accept-source-agreements Accept all source agreements during source operations -r,--rename The value to rename the executable file (portable) --uninstall-previous Uninstall the previous version of the package during upgrade --force Direct run the command and continue with non security related issues -?,--help Shows help about the selected command --wait Prompts the user to press any key before exiting --logs,--open-logs Open the default logs location --verbose,--verbose-logs Enables verbose logging for winget --disable-interactivity Disable interactive prompts More help can be found at: https://aka.ms/winget-command-install [INFO] Node.js installed successfully. Please restart the Installer. Press any key to continue . . . Any idea what is the problem here please?
Aitrepreneur
try this one then: https://we.tl/t-rtY2FmgmLY
littlerunning frog
it worked, thanks!
Renato Estefano
I managed it with the original installer, I applied the video tutorial until 2 minutes and 30 seconds, and from then on I ran the installer and it worked 100%. Thank you very much for all the help. I have not tested the DEBUG installer.
Aitrepreneur
ok no that's just not possible The installer installs sillytavern not tavern ai, what you are showing me is tavern ai Just delete the whole tavern Ai folder and redownload the installer again, launch it and it will create in the same folder a sillytavern folder not a tavern ai folder
Paul
it seems to just spit out [INFO] Winget is already installed. [INFO] winget already exists in PATH. [INFO] Git is already installed. [WARN] Miniconda3 is not installed on this system. [INFO] Installing Miniconda3 using Winget... Failed in attempting to update the source: winget Failed when searching source: winget An unexpected error occurred while executing the command: 0x8a15000f : Data required by the source is missing No packages were found among the working sources. [INFO] Miniconda3 installed successfully. Please restart the Installer. Press any key to continue . . .
TomiTom1234
Or maybe it is my mistake? I guess I installed TavernAI before but forgot about it. When I dropped your SillyTavern installer in a folder called "AI" then installed it, I got no SillyTavern folder, maybe I thougt TavernAI is the result of the installation. So where exactly is the SillyTavern folder installed?
TomiTom1234
Sure.. Here is the first screenshot of the main folder (the 2 bat files I dropped them here), and the second screenshot of what inside "public" folder: https://ibb.co/qgQV1MD https://ibb.co/Jj6mKB1
Aitrepreneur
Ok..then try this one then: https://we.tl/t-rtY2FmgmLY
Renato Estefano
It repeats, the result in the end is the same when I open the installer again.
Aitrepreneur
what the.... that's not possible...since this is not sillytavern but the old tavernai.... could you give me more screens of your folder??
TomiTom1234
Here is a screenshot of the interface I get when I start SillyTavern: https://ibb.co/KX9Wgxw
littlerunning frog
the same will come back again -----What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 1 [INFO] Winget is already installed. [INFO] winget already exists in PATH. [INFO] Git is already installed. [WARN] Miniconda3 is not installed on this system. [INFO] Installing Miniconda3 using Winget... | [INFO] Miniconda3 installed successfully. Please restart the Installer. Press any key to continue . . .
Aitrepreneur
what do you mean? Send a screeshot
TomiTom1234
I used the installer, but the interface I get when launching it is different from what you show in the video. I can't see the icons at the top of the screen. Any idea please?
Aitrepreneur
it install sillytavern in the same folder where you launched the installer
Aitrepreneur
I might add one for Mac if people want it, most people are on windows that's why
Aitrepreneur
Then what happens when you relaunch the installer?
Renato Estefano
same error here
Eddie Torres
I had the same issue and just installed Node.js
Steven Michael
So this is only for windows, and not macos?
Alex C
I had this too, thanks! For those wondering how, just open cmd and paste what the above says :)
Silas
I have the same error
Wojciech Lenartowicz
I had an error while installing Node.js. It's easy to fix. Below I will show you how to quickly fix the issue first, and later how to fix the entire script for future usage. I recommend the latter, but it's up to you :) If you simply want to fix the problem, then simply press Win+R, type "cmd" (no quotes) and press enter, copy and paste "winget install --id OpenJS.NodeJS.LTS -e --silent --accept-source-agreements --accept-package-agreements" (again, without quotes) into the terminal that you've just opened and hit enter. That's it! Now run the script again. However, if you want to fix the script, then right-click the original .bat file and click "edit" (look for it almost at the top of the list) to open it up for edition. You can also open it in Notepad, Notepad++ or any IDE (no Word nor WordPad though! They will likely "ruin" a raw text file, and you won't be able to run it afterward!). Next, replace the entire line 102's content with "winget install --id OpenJS.NodeJS.LTS -e --silent --accept-source-agreements --accept-package-agreements" (without quotes) and save the file. So, basically, I'm removing the faulty, nonexistent in winget "-y" argument and I'm replacing it with "--accept-source-agreements --accept-package-agreements" arguments that will more or less do what the author of the script had originally in mind, probably. Now the script should work if you run it again. You're welcome :)
littlerunning frog
What is your GPU? 1. NVIDIA 2. AMD 3. CPU Enter 1, 2 or 3 and press enter: 1 [INFO] Winget is already installed. [INFO] winget already exists in PATH. [INFO] Git is already installed. [WARN] Miniconda3 is not installed on this system. [INFO] Installing Miniconda3 using Winget... | [INFO] Miniconda3 installed successfully. Please restart the Installer. Press any key to continue . . .
littlerunning frog
I can't install it as I keep getting the following:
watari
So sorry if it's a dumb question but I am brand new here. anyways after following first step of installing and choosing the GPU everything finishes and nothing pop's up. It says it should be installed but I can't find the folder to open up. so what should be my next step? I'm completely lost.
Chroma
<3
Dark Dude Maurycy
Ooooh yeaaah
Alex C
I literally had the github page open to install this and was preparing myself to trawl through it (i know it's not that hard but still), then i saw this notification. Brilliant stuff :D
