Local UI to run and train LLMs and diffusion models. Supports GGUF, MLX, Qwen3.8, DeepSeek-V4.1, MiniMax-H3, Gemma 4, FLUX and more.
Python agentaichatgptdeepseekfine-tuninggemmaimage-generationllamallmllmsopenaipythonqwenreinforcement-learningself-hostedstable-diffusiontext-to-speechttsuiunsloth
Re-measure the Studio startup budget after the Library and composer merges (#13533)
Daniel Hanf68585e2e99f
diff --git a/studio/frontend/scripts/check-bundle-budget.ts b/studio/frontend/scripts/check-bundle-budget.tsindex 25f1610ca15..5b974d9ed8b 100644--- a/studio/frontend/scripts/check-bundle-budget.ts+++ b/studio/frontend/scripts/check-bundle-budget.ts@@ -54,8 +54,14 @@ export const BUDGET = {// 6.6 KB raw / 2.0 KB transfer, mostly in the eager chat chunk (6,513.0 -> 6,519.6 KB raw,// 1,956.4 -> 1,958.4 KB transfer against d4ea5a51d, one machine and build). Lazy-loading the// folder walker recovered 0.5 KB raw and no transfer, so the budget is raised by that growth.- transferBytes: 2_007_100,- rawBytes: 6_687_800,+ // Then 19 small Studio merges took main over again with no single import doing it: 6,518.8 ->+ // 6,533.6 KB raw, 1,958.0 -> 1,963.7 KB transfer from 5f38afbc2d to 33176a75e9 (one machine and+ // build). The largest step is Regenerate title in the Library menu (#13491, +5.2 KB raw / +2.2 KB+ // transfer, code that shares the eager chat row menu module); the Library card and composer tweaks+ // and the edit_file download link add the rest, 0.2 to 1.7 KB transfer each. Raised by that growth+ // so main's margin stays where it was.+ transferBytes: 2_013_000,+ rawBytes: 6_703_000,};// The chunk count is reported but not budgeted. Splitting a page out of the entry raises it while lowering the