oobabooga/text-generation-webui
1
0
Fork 0
mirror of https://github.com/oobabooga/text-generation-webui.git synced 2026-02-05 23:34:21 +01:00
Code Issues Actions Packages Projects Releases Wiki Activity

UI: Add a None option for the speculative decoding model (closes #7145)

Browse source
This commit is contained in:
oobabooga 2025-07-19 12:14:41 -07:00
parent ccf5e3e3a7
commit 0c667de7a7
1 changed files with 2 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
modules/ui_model_menu.py
View file

@ -62,8 +62,8 @@ def create_ui():
# Speculative decoding
with gr.Accordion("Speculative decoding", open=False, elem_classes='tgw-accordion') as shared.gradio['speculative_decoding_accordion']:
with gr.Row():
shared.gradio['model_draft'] = gr.Dropdown(label="model-draft", choices=utils.get_available_models(), value=lambda: shared.args.model_draft, elem_classes='slim-dropdown', info='Draft model. Speculative decoding only works with models sharing the same vocabulary (e.g., same model family).', interactive=not mu)
ui.create_refresh_button(shared.gradio['model_draft'], lambda: None, lambda: {'choices': utils.get_available_models()}, 'refresh-button', interactive=not mu)
shared.gradio['model_draft'] = gr.Dropdown(label="model-draft", choices=['None'] + utils.get_available_models(), value=lambda: shared.args.model_draft, elem_classes='slim-dropdown', info='Draft model. Speculative decoding only works with models sharing the same vocabulary (e.g., same model family).', interactive=not mu)
ui.create_refresh_button(shared.gradio['model_draft'], lambda: None, lambda: {'choices': ['None'] + utils.get_available_models()}, 'refresh-button', interactive=not mu)
shared.gradio['gpu_layers_draft'] = gr.Slider(label="gpu-layers-draft", minimum=0, maximum=256, value=shared.args.gpu_layers_draft, info='Number of layers to offload to the GPU for the draft model.')
shared.gradio['draft_max'] = gr.Number(label="draft-max", precision=0, step=1, value=shared.args.draft_max, info='Number of tokens to draft for speculative decoding. Recommended value: 4.')