Update app.py
Browse files
app.py
CHANGED
|
@@ -1,80 +1,382 @@
|
|
| 1 |
import gradio as gr
|
| 2 |
from huggingface_hub import InferenceClient
|
| 3 |
-
import json
|
|
|
|
| 4 |
|
| 5 |
-
|
|
|
|
| 6 |
|
| 7 |
-
def load_models():
|
| 8 |
-
if os.path.exists(MODEL_FILE):
|
| 9 |
-
return json.load(open(MODEL_FILE))
|
| 10 |
-
return {"Default (GPT‑OSS‑20B)": "openai/gpt-oss-20b"}
|
| 11 |
|
| 12 |
-
|
| 13 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 14 |
|
| 15 |
-
models = load_models()
|
| 16 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
|
| 18 |
-
def add_model(link):
|
| 19 |
-
name = link.split("/")[-1]
|
| 20 |
-
models[name] = link
|
| 21 |
-
save_models(models)
|
| 22 |
-
return gr.Dropdown.update(choices=list(models.keys()), value=name)
|
| 23 |
|
|
|
|
|
|
|
|
|
|
| 24 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 25 |
def respond(
|
| 26 |
message,
|
| 27 |
-
history
|
| 28 |
system_message,
|
| 29 |
max_tokens,
|
| 30 |
temperature,
|
| 31 |
top_p,
|
| 32 |
-
|
| 33 |
-
hf_token: gr.OAuthToken
|
| 34 |
):
|
| 35 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 36 |
client = InferenceClient(token=hf_token.token, model=model_id)
|
| 37 |
|
| 38 |
messages = [{"role": "system", "content": system_message}]
|
| 39 |
-
messages.extend(history)
|
| 40 |
messages.append({"role": "user", "content": message})
|
| 41 |
|
| 42 |
response = ""
|
| 43 |
-
|
| 44 |
-
for message in client.chat_completion(
|
| 45 |
messages,
|
| 46 |
max_tokens=max_tokens,
|
| 47 |
stream=True,
|
| 48 |
temperature=temperature,
|
| 49 |
top_p=top_p,
|
| 50 |
):
|
| 51 |
-
|
| 52 |
-
|
| 53 |
-
if len(choices) and choices[0].delta.content:
|
| 54 |
-
token = choices[0].delta.content
|
| 55 |
-
|
| 56 |
-
response += token
|
| 57 |
yield response
|
| 58 |
|
| 59 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 60 |
with gr.Blocks() as demo:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 61 |
with gr.Sidebar():
|
| 62 |
gr.LoginButton()
|
| 63 |
-
gr.Markdown("### Add a new model")
|
| 64 |
-
new_model = gr.Textbox(label="Model repo (e.g. meta-llama/Llama-3-8B)")
|
| 65 |
-
model_dropdown = gr.Dropdown(list(models.keys()), label="Choose model")
|
| 66 |
-
add_button = gr.Button("Add model")
|
| 67 |
-
add_button.click(add_model, new_model, model_dropdown)
|
| 68 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 69 |
chatbot = gr.ChatInterface(
|
| 70 |
respond,
|
|
|
|
| 71 |
type="messages",
|
| 72 |
additional_inputs=[
|
| 73 |
-
gr.Textbox(
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 78 |
],
|
| 79 |
)
|
| 80 |
|
|
|
|
| 1 |
import gradio as gr
|
| 2 |
from huggingface_hub import InferenceClient
|
| 3 |
+
import json
|
| 4 |
+
import os
|
| 5 |
|
| 6 |
+
DEFAULT_FILE = "default_models.json"
|
| 7 |
+
USER_FILE = "models.json"
|
| 8 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
|
| 10 |
+
# -----------------------------
|
| 11 |
+
# Model data loading / saving
|
| 12 |
+
# -----------------------------
|
| 13 |
+
def load_default_models():
|
| 14 |
+
with open(DEFAULT_FILE, "r", encoding="utf-8") as f:
|
| 15 |
+
return json.load(f)
|
| 16 |
|
|
|
|
| 17 |
|
| 18 |
+
def load_user_models():
|
| 19 |
+
if os.path.exists(USER_FILE):
|
| 20 |
+
with open(USER_FILE, "r", encoding="utf-8") as f:
|
| 21 |
+
try:
|
| 22 |
+
return json.load(f)
|
| 23 |
+
except json.JSONDecodeError:
|
| 24 |
+
return {}
|
| 25 |
+
return {}
|
| 26 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 27 |
|
| 28 |
+
def save_user_models(data):
|
| 29 |
+
with open(USER_FILE, "w", encoding="utf-8") as f:
|
| 30 |
+
json.dump(data, f, indent=2, ensure_ascii=False)
|
| 31 |
|
| 32 |
+
|
| 33 |
+
def merge_models():
|
| 34 |
+
"""
|
| 35 |
+
Merge default + user models into one tree:
|
| 36 |
+
Category -> Family -> Model -> meta
|
| 37 |
+
User models can introduce new categories/families.
|
| 38 |
+
"""
|
| 39 |
+
base = load_default_models()
|
| 40 |
+
user = load_user_models()
|
| 41 |
+
|
| 42 |
+
for category, families in user.items():
|
| 43 |
+
if category not in base:
|
| 44 |
+
base[category] = {}
|
| 45 |
+
for family, models in families.items():
|
| 46 |
+
if family not in base[category]:
|
| 47 |
+
base[category][family] = {}
|
| 48 |
+
for model_name, meta in models.items():
|
| 49 |
+
base[category][family][model_name] = meta
|
| 50 |
+
|
| 51 |
+
return base
|
| 52 |
+
|
| 53 |
+
|
| 54 |
+
# -----------------------------
|
| 55 |
+
# Utility: flatten and lookup
|
| 56 |
+
# -----------------------------
|
| 57 |
+
def flatten_models(model_tree):
|
| 58 |
+
"""
|
| 59 |
+
Returns a dict:
|
| 60 |
+
full_key -> (meta, category, family, model_name)
|
| 61 |
+
where full_key = "Category / Family / Model"
|
| 62 |
+
"""
|
| 63 |
+
flat = {}
|
| 64 |
+
for category, families in model_tree.items():
|
| 65 |
+
for family, models in families.items():
|
| 66 |
+
for model_name, meta in models.items():
|
| 67 |
+
full_key = f"{category} / {family} / {model_name}"
|
| 68 |
+
flat[full_key] = (meta, category, family, model_name)
|
| 69 |
+
return flat
|
| 70 |
+
|
| 71 |
+
|
| 72 |
+
# -----------------------------
|
| 73 |
+
# Add a new model
|
| 74 |
+
# -----------------------------
|
| 75 |
+
def add_model_box(
|
| 76 |
+
category,
|
| 77 |
+
family,
|
| 78 |
+
model_name,
|
| 79 |
+
model_id,
|
| 80 |
+
description,
|
| 81 |
+
link,
|
| 82 |
+
emoji
|
| 83 |
+
):
|
| 84 |
+
# Basic validation
|
| 85 |
+
if not model_id:
|
| 86 |
+
return gr.Markdown.update(
|
| 87 |
+
value="Please provide a Model ID like `user/model`."
|
| 88 |
+
)
|
| 89 |
+
|
| 90 |
+
# Fallbacks
|
| 91 |
+
if not category:
|
| 92 |
+
category = "Custom"
|
| 93 |
+
if not family:
|
| 94 |
+
family = "User-Added"
|
| 95 |
+
if not model_name:
|
| 96 |
+
model_name = model_id.split("/")[-1]
|
| 97 |
+
if not description:
|
| 98 |
+
description = "User-added model."
|
| 99 |
+
if not link:
|
| 100 |
+
link = f"https://huggingface.co/{model_id}"
|
| 101 |
+
if not emoji:
|
| 102 |
+
emoji = "✨"
|
| 103 |
+
|
| 104 |
+
user_models = load_user_models()
|
| 105 |
+
|
| 106 |
+
# Ensure nested structure exists
|
| 107 |
+
if category not in user_models:
|
| 108 |
+
user_models[category] = {}
|
| 109 |
+
if family not in user_models[category]:
|
| 110 |
+
user_models[category][family] = {}
|
| 111 |
+
|
| 112 |
+
user_models[category][family][model_name] = {
|
| 113 |
+
"id": model_id,
|
| 114 |
+
"description": description,
|
| 115 |
+
"link": link,
|
| 116 |
+
"emoji": emoji
|
| 117 |
+
}
|
| 118 |
+
|
| 119 |
+
save_user_models(user_models)
|
| 120 |
+
|
| 121 |
+
msg = (
|
| 122 |
+
f"Added model under `{category} / {family}`: "
|
| 123 |
+
f"{emoji} **{model_name}** (`{model_id}`)\n\n"
|
| 124 |
+
f"It will appear in the model tree after reloading the Space."
|
| 125 |
+
)
|
| 126 |
+
return gr.Markdown.update(value=msg)
|
| 127 |
+
|
| 128 |
+
|
| 129 |
+
# -----------------------------
|
| 130 |
+
# Chat function
|
| 131 |
+
# -----------------------------
|
| 132 |
def respond(
|
| 133 |
message,
|
| 134 |
+
history,
|
| 135 |
system_message,
|
| 136 |
max_tokens,
|
| 137 |
temperature,
|
| 138 |
top_p,
|
| 139 |
+
active_model_key,
|
| 140 |
+
hf_token: gr.OAuthToken
|
| 141 |
):
|
| 142 |
+
if active_model_key is None:
|
| 143 |
+
yield "No model selected. Please choose a model in the sidebar and click 'Use this model'."
|
| 144 |
+
return
|
| 145 |
+
|
| 146 |
+
models = merge_models()
|
| 147 |
+
flat = flatten_models(models)
|
| 148 |
+
|
| 149 |
+
meta_tuple = flat.get(active_model_key)
|
| 150 |
+
if meta_tuple is None:
|
| 151 |
+
yield "Selected model not found. Please choose a model again."
|
| 152 |
+
return
|
| 153 |
+
|
| 154 |
+
meta, _, _, _ = meta_tuple
|
| 155 |
+
model_id = meta["id"]
|
| 156 |
+
|
| 157 |
client = InferenceClient(token=hf_token.token, model=model_id)
|
| 158 |
|
| 159 |
messages = [{"role": "system", "content": system_message}]
|
| 160 |
+
messages.extend(history or [])
|
| 161 |
messages.append({"role": "user", "content": message})
|
| 162 |
|
| 163 |
response = ""
|
| 164 |
+
for msg in client.chat_completion(
|
|
|
|
| 165 |
messages,
|
| 166 |
max_tokens=max_tokens,
|
| 167 |
stream=True,
|
| 168 |
temperature=temperature,
|
| 169 |
top_p=top_p,
|
| 170 |
):
|
| 171 |
+
delta = msg.choices[0].delta.content or ""
|
| 172 |
+
response += delta
|
|
|
|
|
|
|
|
|
|
|
|
|
| 173 |
yield response
|
| 174 |
|
| 175 |
|
| 176 |
+
# -----------------------------
|
| 177 |
+
# Build the sidebar tree
|
| 178 |
+
# -----------------------------
|
| 179 |
+
def build_model_tree(
|
| 180 |
+
models,
|
| 181 |
+
active_model_state,
|
| 182 |
+
current_model_label,
|
| 183 |
+
info_markdowns,
|
| 184 |
+
use_buttons
|
| 185 |
+
):
|
| 186 |
+
"""
|
| 187 |
+
models: merged models dict (Category -> Family -> Model -> meta)
|
| 188 |
+
active_model_state: gr.State storing current active full key
|
| 189 |
+
current_model_label: gr.Markdown for 'Current model: ...'
|
| 190 |
+
info_markdowns: dict full_key -> gr.Markdown (Model Info)
|
| 191 |
+
use_buttons: dict full_key -> gr.Button (Use this model)
|
| 192 |
+
"""
|
| 193 |
+
|
| 194 |
+
flat = flatten_models(models)
|
| 195 |
+
|
| 196 |
+
# We’ll build the UI, and inside each loop create small closures
|
| 197 |
+
for category, families in models.items():
|
| 198 |
+
with gr.Accordion(category, open=False):
|
| 199 |
+
for family, model_dict in families.items():
|
| 200 |
+
with gr.Accordion(family, open=False):
|
| 201 |
+
for model_name, meta in model_dict.items():
|
| 202 |
+
emoji = meta.get("emoji", "✨")
|
| 203 |
+
full_key = f"{category} / {family} / {model_name}"
|
| 204 |
+
|
| 205 |
+
# Model button row
|
| 206 |
+
model_button = gr.Button(f"{emoji} {model_name}", size="sm")
|
| 207 |
+
|
| 208 |
+
# Model info area
|
| 209 |
+
info_md = gr.Markdown(visible=False)
|
| 210 |
+
info_markdowns[full_key] = info_md
|
| 211 |
+
|
| 212 |
+
# Use button
|
| 213 |
+
use_btn = gr.Button("Use this model", visible=False, size="sm")
|
| 214 |
+
use_buttons[full_key] = use_btn
|
| 215 |
+
|
| 216 |
+
# ---- Click model button → show this info, hide others
|
| 217 |
+
def show_info(clicked_key, fk=full_key):
|
| 218 |
+
models_local = merge_models()
|
| 219 |
+
flat_local = flatten_models(models_local)
|
| 220 |
+
|
| 221 |
+
updates = {}
|
| 222 |
+
for key, (meta_loc, _, _, _) in flat_local.items():
|
| 223 |
+
md_out = info_markdowns.get(key)
|
| 224 |
+
btn_out = use_buttons.get(key)
|
| 225 |
+
if md_out is None or btn_out is None:
|
| 226 |
+
continue
|
| 227 |
+
|
| 228 |
+
if key == fk:
|
| 229 |
+
text = (
|
| 230 |
+
f"**Model ID:** `{meta_loc['id']}` \n"
|
| 231 |
+
f"**Description:** {meta_loc['description']} \n"
|
| 232 |
+
f"[Model card]({meta_loc['link']})"
|
| 233 |
+
)
|
| 234 |
+
updates[md_out] = gr.Markdown.update(value=text, visible=True)
|
| 235 |
+
updates[btn_out] = gr.Button.update(visible=True)
|
| 236 |
+
else:
|
| 237 |
+
updates[md_out] = gr.Markdown.update(visible=False)
|
| 238 |
+
updates[btn_out] = gr.Button.update(visible=False)
|
| 239 |
+
return updates
|
| 240 |
+
|
| 241 |
+
model_button.click(
|
| 242 |
+
show_info,
|
| 243 |
+
inputs=active_model_state,
|
| 244 |
+
outputs=list(info_markdowns.values()) + list(use_buttons.values()),
|
| 245 |
+
)
|
| 246 |
+
|
| 247 |
+
# ---- Click "Use this model" → set active model + label
|
| 248 |
+
def use_model(fk=full_key):
|
| 249 |
+
models_local = merge_models()
|
| 250 |
+
flat_local = flatten_models(models_local)
|
| 251 |
+
meta_loc_tuple = flat_local.get(fk)
|
| 252 |
+
|
| 253 |
+
if not meta_loc_tuple:
|
| 254 |
+
return fk, gr.Markdown.update(
|
| 255 |
+
value="**Current model:** _none selected_"
|
| 256 |
+
)
|
| 257 |
+
|
| 258 |
+
meta_loc, _, _, mname = meta_loc_tuple
|
| 259 |
+
emoji_local = meta_loc.get("emoji", "✨")
|
| 260 |
+
label_text = f"**Current model:** {emoji_local} {mname}"
|
| 261 |
+
return fk, gr.Markdown.update(value=label_text)
|
| 262 |
+
|
| 263 |
+
use_btn.click(
|
| 264 |
+
use_model,
|
| 265 |
+
inputs=None,
|
| 266 |
+
outputs=[active_model_state, current_model_label],
|
| 267 |
+
)
|
| 268 |
+
|
| 269 |
+
|
| 270 |
+
# -----------------------------
|
| 271 |
+
# Build the UI
|
| 272 |
+
# -----------------------------
|
| 273 |
with gr.Blocks() as demo:
|
| 274 |
+
models_tree = merge_models()
|
| 275 |
+
|
| 276 |
+
# Holds full key: "Category / Family / Model"
|
| 277 |
+
active_model_key = gr.State(value=None)
|
| 278 |
+
|
| 279 |
with gr.Sidebar():
|
| 280 |
gr.LoginButton()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 281 |
|
| 282 |
+
# Collapsible "Add New Model" box
|
| 283 |
+
with gr.Accordion("Add New Model", open=False):
|
| 284 |
+
category_input = gr.Textbox(
|
| 285 |
+
label="Category (e.g. Exotic or new category)",
|
| 286 |
+
placeholder="Exotic"
|
| 287 |
+
)
|
| 288 |
+
family_input = gr.Textbox(
|
| 289 |
+
label="Family (e.g. RWKV)",
|
| 290 |
+
placeholder="RWKV"
|
| 291 |
+
)
|
| 292 |
+
model_name_input = gr.Textbox(
|
| 293 |
+
label="Model Name (e.g. RWKV-World-7B)",
|
| 294 |
+
placeholder="RWKV-World-7B"
|
| 295 |
+
)
|
| 296 |
+
model_id_input = gr.Textbox(
|
| 297 |
+
label="Model ID (e.g. BlinkDL/rwkv-7-world)",
|
| 298 |
+
placeholder="BlinkDL/rwkv-7-world"
|
| 299 |
+
)
|
| 300 |
+
description_input = gr.Textbox(
|
| 301 |
+
label="Description (optional)",
|
| 302 |
+
lines=2
|
| 303 |
+
)
|
| 304 |
+
link_input = gr.Textbox(
|
| 305 |
+
label="Link (optional, will default to https://huggingface.co/ModelID if empty)",
|
| 306 |
+
lines=1
|
| 307 |
+
)
|
| 308 |
+
emoji_input = gr.Textbox(
|
| 309 |
+
label="Emoji (optional, e.g. 🌍)",
|
| 310 |
+
lines=1
|
| 311 |
+
)
|
| 312 |
+
|
| 313 |
+
add_button = gr.Button("Add Model")
|
| 314 |
+
add_status = gr.Markdown("")
|
| 315 |
+
|
| 316 |
+
add_button.click(
|
| 317 |
+
add_model_box,
|
| 318 |
+
inputs=[
|
| 319 |
+
category_input,
|
| 320 |
+
family_input,
|
| 321 |
+
model_name_input,
|
| 322 |
+
model_id_input,
|
| 323 |
+
description_input,
|
| 324 |
+
link_input,
|
| 325 |
+
emoji_input,
|
| 326 |
+
],
|
| 327 |
+
outputs=add_status,
|
| 328 |
+
)
|
| 329 |
+
|
| 330 |
+
# Current model label under the Add box
|
| 331 |
+
current_model_label = gr.Markdown("**Current model:** _none selected_")
|
| 332 |
+
|
| 333 |
+
# Structures to hold info areas
|
| 334 |
+
info_markdowns = {}
|
| 335 |
+
use_buttons = {}
|
| 336 |
+
|
| 337 |
+
gr.Markdown("### Models")
|
| 338 |
+
|
| 339 |
+
# Build nested accordions for models
|
| 340 |
+
build_model_tree(
|
| 341 |
+
models_tree,
|
| 342 |
+
active_model_state=active_model_key,
|
| 343 |
+
current_model_label=current_model_label,
|
| 344 |
+
info_markdowns=info_markdowns,
|
| 345 |
+
use_buttons=use_buttons,
|
| 346 |
+
)
|
| 347 |
+
|
| 348 |
+
# Main chat interface
|
| 349 |
chatbot = gr.ChatInterface(
|
| 350 |
respond,
|
| 351 |
+
title="chatbot",
|
| 352 |
type="messages",
|
| 353 |
additional_inputs=[
|
| 354 |
+
gr.Textbox(
|
| 355 |
+
value="You are a friendly chatbot.",
|
| 356 |
+
label="System message"
|
| 357 |
+
),
|
| 358 |
+
gr.Slider(
|
| 359 |
+
minimum=1,
|
| 360 |
+
maximum=2048,
|
| 361 |
+
value=512,
|
| 362 |
+
step=1,
|
| 363 |
+
label="Max new tokens"
|
| 364 |
+
),
|
| 365 |
+
gr.Slider(
|
| 366 |
+
minimum=0.1,
|
| 367 |
+
maximum=4.0,
|
| 368 |
+
value=0.7,
|
| 369 |
+
step=0.1,
|
| 370 |
+
label="Temperature"
|
| 371 |
+
),
|
| 372 |
+
gr.Slider(
|
| 373 |
+
minimum=0.1,
|
| 374 |
+
maximum=1.0,
|
| 375 |
+
value=0.95,
|
| 376 |
+
step=0.05,
|
| 377 |
+
label="Top-p"
|
| 378 |
+
),
|
| 379 |
+
active_model_key, # passes current active model key into respond()
|
| 380 |
],
|
| 381 |
)
|
| 382 |
|