From ba1ba0fc8a5ee0d77f05e1a42dba0c62a83d2eb6 Mon Sep 17 00:00:00 2001 From: KoboldAI Dev Date: Fri, 7 May 2021 19:04:51 -0400 Subject: [PATCH] Reduced default max_length parameter to 512. Added warning about VRAM usage to Max Tokens tooltip. --- aiserver.py | 2 +- gensettings.py | 10 +++++----- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/aiserver.py b/aiserver.py index 2109469a..95c192f5 100644 --- a/aiserver.py +++ b/aiserver.py @@ -48,7 +48,7 @@ class vars: model = "" noai = False # Runs the script without starting up the transformers pipeline aibusy = False # Stops submissions while the AI is working - max_length = 1024 # Maximum number of tokens to submit per action + max_length = 512 # Maximum number of tokens to submit per action ikmax = 3000 # Maximum number of characters to submit to InferKit genamt = 60 # Amount of text for each action to generate ikgen = 200 # Number of characters for InferKit to generate diff --git a/gensettings.py b/gensettings.py index fdd82c2a..0150a139 100644 --- a/gensettings.py +++ b/gensettings.py @@ -29,15 +29,15 @@ gensettingstf = [{ "max": 2.0, "step": 0.05, "default": 1.0, - "tooltip": "Used to penalize words that were already generated or belong to the context" + "tooltip": "Used to penalize words that were already generated or belong to the context." }, { "uitype": "slider", "unit": "int", "label": "Amount to Generate", "id": "setoutput", - "min": 10, - "max": 500, + "min": 16, + "max": 512, "step": 2, "default": 60, "tooltip": "Number of tokens the AI should generate. Higher numbers will take longer to generate." @@ -50,8 +50,8 @@ gensettingstf = [{ "min": 512, "max": 2048, "step": 8, - "default": 1024, - "tooltip": "Number of tokens of context to submit to the AI for sampling." + "default": 512, + "tooltip": "Max number of tokens of context to submit to the AI for sampling. Make sure this is higher than Amount to Generate. Higher values increase VRAM/RAM usage." }] gensettingsik =[{