rixty/KoboldAI-Client
1
1
Fork 0
mirror of https://github.com/KoboldAI/KoboldAI-Client.git synced 2025-02-25 16:08:00 +01:00
Code Issues Projects Releases Wiki Activity

Threading + Memory Sizes

Browse Source 
Polish effort to suppress a warning and list more accurate VRAM as tested with the full 2048 max tokens.
This commit is contained in:
henk717 2022-01-30 13:56:25 +01:00
parent fca7f8659f
commit 5b5a479f29
1 changed files with 7 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
aiserver.py
View File

@ -11,6 +11,7 @@ eventlet.monkey_patch(all=True, thread=False)
import os import os
os.system("") os.system("")
os.environ['EVENTLET_THREADPOOL_SIZE'] = '1' os.environ['EVENTLET_THREADPOOL_SIZE'] = '1'
os.environ['TOKENIZERS_PARALLELISM'] = 'false'
from eventlet import tpool from eventlet import tpool
from os import path, getcwd from os import path, getcwd
@ -74,12 +75,12 @@ modellist = [
["Horni 2.7B (NSFW)", "KoboldAI/GPT-Neo-2.7B-Horni", "8GB"], ["Horni 2.7B (NSFW)", "KoboldAI/GPT-Neo-2.7B-Horni", "8GB"],
["Horni-LN 2.7B (Novel)", "KoboldAI/GPT-Neo-2.7B-Horni-LN", "8GB"], ["Horni-LN 2.7B (Novel)", "KoboldAI/GPT-Neo-2.7B-Horni-LN", "8GB"],
["Shinen 2.7B (NSFW)", "KoboldAI/GPT-Neo-2.7B-Shinen", "8GB"], ["Shinen 2.7B (NSFW)", "KoboldAI/GPT-Neo-2.7B-Shinen", "8GB"],
["GPT-J 6B", "EleutherAI/gpt-j-6B", "12GB"], ["GPT-J 6B", "EleutherAI/gpt-j-6B", "16GB"],
["GPT-Neo 2.7B", "EleutherAI/gpt-neo-2.7B", "8GB"], ["GPT-Neo 2.7B", "EleutherAI/gpt-neo-2.7B", "8GB"],
["GPT-Neo 1.3B", "EleutherAI/gpt-neo-1.3B", "6GB"], ["GPT-Neo 1.3B", "EleutherAI/gpt-neo-1.3B", "6GB"],
["GPT-2 XL", "gpt2-xl", "8GB"], ["GPT-2 XL", "gpt2-xl", "6GB"],
["GPT-2 Large", "gpt2-large", "6GB"], ["GPT-2 Large", "gpt2-large", "4GB"],
["GPT-2 Med", "gpt2-medium", "4GB"], ["GPT-2 Med", "gpt2-medium", "2GB"],
["GPT-2", "gpt2", "2GB"], ["GPT-2", "gpt2", "2GB"],
["OpenAI API (requires API key)", "OAI", ""], ["OpenAI API (requires API key)", "OAI", ""],
["InferKit API (requires API key)", "InferKit", ""], ["InferKit API (requires API key)", "InferKit", ""],
@ -201,7 +202,7 @@ class vars:
# Function to get model selection at startup # Function to get model selection at startup
#==================================================================# #==================================================================#
def getModelSelection(): def getModelSelection():
print(" # Model V/RAM\n =========================================") print(" # Model VRAM\n =========================================")
i = 1 i = 1
for m in modellist: for m in modellist:
print(" {0} - {1}\t\t{2}".format("{:<2}".format(i), m[0].ljust(15), m[2])) print(" {0} - {1}\t\t{2}".format("{:<2}".format(i), m[0].ljust(15), m[2]))
@ -432,7 +433,7 @@ if args.model:
vars.colaburl = args.path + "/request"; # Lets just use the same parameter to keep it simple vars.colaburl = args.path + "/request"; # Lets just use the same parameter to keep it simple
else: else:
print("{0}Welcome to the KoboldAI Server!\nSelect an AI model to continue:{1}\n".format(colors.CYAN, colors.END)) print("{0}Welcome to the KoboldAI Server!\nListed RAM is the optimal VRAM and CPU ram can be up to twice the amount.\nMost models can run at less VRAM with reduced max tokens or less layers on the GPU.\nSelect an AI model to continue:{1}\n".format(colors.CYAN, colors.END))
getModelSelection() getModelSelection()
# If transformers model was selected & GPU available, ask to use CPU or GPU # If transformers model was selected & GPU available, ask to use CPU or GPU