FLAS-Llama3.1
Collection
2 items • Updated
Zehao Jin PRO
Lunamos
·
AI & ML interests
None yet
Recent Activity
new activity 42 minutes ago
Lunamos/flas-demo:Stream baseline too; loading notice on first-use download; fill_width=True; relabel StepsOrganizations
Sliders 2-per-row + min_width (no more wrapped labels); native loading spinner (drop text notice)
#13 opened 33 minutes ago
by Lunamos
Stream baseline too; loading notice on first-use download; fill_width=True; relabel Steps
#12 opened 42 minutes ago
by Lunamos
Pin flas to commit 33b0ac0 (forces fresh install with generate_stream; fixes AttributeError) + widen to 1600px
#11 opened about 1 hour ago
by Lunamos
Disable SSR (ssr_mode=False) so ZeroGPU generator streaming reaches the browser (gradio#10914)
#10 opened about 1 hour ago
by Lunamos
Stream steered output token-by-token + Stop button (interruptible, via generate_stream)
#9 opened about 1 hour ago
by Lunamos
Stream steered output token-by-token + Stop button (interruptible); uses generate_stream
#8 opened about 1 hour ago
by Lunamos
Default to Llama-3.1-8B-Instruct; remove Qwen3 thinking toggle (always non-thinking); drop 'steering strength' label
#7 opened about 1 hour ago
by Lunamos
updated 4
collections about 1 hour ago
Pin flas to stable checkpoint20250604 branch (qwen3-thinking-fix was deleted post-merge)
#6 opened about 2 hours ago
by Lunamos
Add experimental Qwen3 thinking-mode toggle (default off, shown only for Qwen3 models)
#4 opened about 4 hours ago
by Lunamos
updated a
model about 4 hours ago