Merge branch 'feature/llamacpp' of github.com:thomasantony/text-generation-webui into thomasantony-feature/llamacpp

README.md

@@ -240,4 +240,4 @@ Before reporting a bug, make sure that you have:
 - Gradio dropdown menu refresh button, code for reloading the interface: https://github.com/AUTOMATIC1111/stable-diffusion-webui
 - Verbose preset: Anonymous 4chan user.
 - NovelAI and KoboldAI presets: https://github.com/KoboldAI/KoboldAI-Client/wiki/Settings-Presets
-- Pygmalion preset, code for early stopping in chat mode, code for some of the sliders, --chat mode colors: https://github.com/PygmalionAI/gradio-ui/
+- Code for early stopping in chat mode, code for some of the sliders: https://github.com/PygmalionAI/gradio-ui/

extensions/gallery/script.py

@@ -2,19 +2,29 @@ from pathlib import Path
 
 import gradio as gr
 
+from modules.chat import load_character
 from modules.html_generator import get_image_cache
+from modules.shared import gradio, settings
 
 
-def generate_html():
+def generate_css():
     css = """
-      .character-gallery {
+      .character-gallery > .gallery {
         margin: 1rem 0;
-        display: grid;
+        display: grid !important;
         grid-template-columns: repeat(auto-fit, minmax(150px, 1fr));
         grid-column-gap: 0.4rem;
         grid-row-gap: 1.2rem;
       }
 
+      .character-gallery > .label {
+        display: none !important;
+      }
+
+      .character-gallery button.gallery-item {
+        display: contents;
+      }
+
       .character-container {
         cursor: pointer;
         text-align: center;
@@ -45,14 +55,16 @@ def generate_html():
         overflow-wrap: anywhere;
       }
     """
+    return css
 
-    container_html = f'<style>{css}</style><div class="character-gallery">'
 
+def generate_html():
+    cards = []
     # Iterate through files in image folder
     for file in sorted(Path("characters").glob("*")):
         if file.name.endswith(".json"):
             character = file.name.replace(".json", "")
-            container_html += f'<div class="character-container" onclick=\'document.getElementById("character-menu").children[1].children[1].value = "{character}"; document.getElementById("character-menu").children[1].children[1].dispatchEvent(new Event("change"));\'>'
+            container_html = f'<div class="character-container">'
             image_html = "<div class='placeholder'></div>"
 
             for i in [
@@ -71,12 +83,24 @@ def generate_html():
 
             container_html += f'{image_html} <span class="character-name">{character}</span>'
             container_html += "</div>"
+            cards.append([container_html, character])
+
+    return cards
+
+
+def select_character(evt: gr.SelectData):
+    return (evt.value[1])
 
-    container_html += "</div>"
-    return container_html
 
 def ui():
     with gr.Accordion("Character gallery", open=False):
         update = gr.Button("Refresh")
-        gallery = gr.HTML(value=generate_html())
+        gr.HTML(value="<style>"+generate_css()+"</style>")
+        gallery = gr.Dataset(components=[gr.HTML(visible=False)],
+            label="",
+            samples=generate_html(),
+            elem_classes=["character-gallery"],
+            samples_per_page=50
+        )
     update.click(generate_html, [], gallery)
+    gallery.select(select_character, None, gradio['character_menu'])

modules/callbacks.py

@@ -12,7 +12,7 @@ import modules.shared as shared
 # Copied from https://github.com/PygmalionAI/gradio-ui/
 class _SentinelTokenStoppingCriteria(transformers.StoppingCriteria):
 
-    def __init__(self, sentinel_token_ids: list[torch.LongTensor], starting_idx: int):
+    def __init__(self, sentinel_token_ids: list, starting_idx: int):
         transformers.StoppingCriteria.__init__(self)
         self.sentinel_token_ids = sentinel_token_ids
         self.starting_idx = starting_idx

modules/shared.py

@@ -27,6 +27,7 @@ settings = {
     'max_new_tokens': 200,
     'max_new_tokens_min': 1,
     'max_new_tokens_max': 2000,
+    'seed': -1,
     'name1': 'You',
     'name2': 'Assistant',
     'context': 'This is a conversation with your Assistant. The Assistant is very helpful and is eager to chat with you and answer your questions.',
@@ -41,7 +42,7 @@ settings = {
     'chat_default_extensions': ["gallery"],
     'presets': {
         'default': 'NovelAI-Sphinx Moth',
-        '.*pygmalion': 'Pygmalion',
+        '.*pygmalion': 'NovelAI-Storywriter',
         '.*RWKV': 'Naive',
     },
     'prompts': {

presets/Pygmalion.txt

@@ -1,6 +0,0 @@
-do_sample=True
-top_p=0.9
-top_k=0
-temperature=0.5
-repetition_penalty=1.1
-typical_p=1.0

requirements.txt

@@ -1,7 +1,7 @@
 accelerate==0.18.0
 bitsandbytes==0.37.2
 flexgen==0.1.7
-gradio==3.23.0
+gradio==3.24.0
 llamacpp==0.1.10
 markdown
 numpy

server.py

@@ -166,7 +166,7 @@ def create_settings_menus(default_preset):
         with gr.Column():
             create_model_and_preset_menus()
         with gr.Column():
-            shared.gradio['seed'] = gr.Number(value=-1, label='Seed (-1 for random)')
+            shared.gradio['seed'] = gr.Number(value=shared.settings['seed'], label='Seed (-1 for random)')
 
     with gr.Row():
         with gr.Column():
@@ -408,7 +408,7 @@ def create_interface():
                 with gr.Row():
                     with gr.Column(scale=4):
                         with gr.Tab('Raw'):
-                            shared.gradio['textbox'] = gr.Textbox(value=default_text, elem_id="textbox", lines=25)
+                            shared.gradio['textbox'] = gr.Textbox(value=default_text, elem_id="textbox", lines=27)
                         with gr.Tab('Markdown'):
                             shared.gradio['markdown'] = gr.Markdown()
                         with gr.Tab('HTML'):
@@ -442,7 +442,7 @@ def create_interface():
             with gr.Tab("Text generation", elem_id="main"):
                 with gr.Row():
                     with gr.Column():
-                        shared.gradio['textbox'] = gr.Textbox(value=default_text, lines=15, label='Input')
+                        shared.gradio['textbox'] = gr.Textbox(value=default_text, lines=21, label='Input')
                         shared.gradio['max_new_tokens'] = gr.Slider(minimum=shared.settings['max_new_tokens_min'], maximum=shared.settings['max_new_tokens_max'], step=1, label='max_new_tokens', value=shared.settings['max_new_tokens'])
                         shared.gradio['Generate'] = gr.Button('Generate')
                         with gr.Row():
@@ -455,7 +455,7 @@ def create_interface():
 
                     with gr.Column():
                         with gr.Tab('Raw'):
-                            shared.gradio['output_textbox'] = gr.Textbox(lines=25, label='Output')
+                            shared.gradio['output_textbox'] = gr.Textbox(lines=27, label='Output')
                         with gr.Tab('Markdown'):
                             shared.gradio['markdown'] = gr.Markdown()
                         with gr.Tab('HTML'):

settings-template.json

@@ -2,6 +2,7 @@
     "max_new_tokens": 200,
     "max_new_tokens_min": 1,
     "max_new_tokens_max": 2000,
+    "seed": -1,
     "name1": "You",
     "name2": "Assistant",
     "context": "This is a conversation with your Assistant. The Assistant is very helpful and is eager to chat with you and answer your questions.",
@@ -18,7 +19,7 @@
     ],
     "presets": {
         "default": "NovelAI-Sphinx Moth",
-        ".*pygmalion": "Pygmalion",
+        ".*pygmalion": "NovelAI-Storywriter",
         ".*RWKV": "Naive"
     },
     "prompts": {