jeffemmett
/
rspace-online
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

fix(blender): use correct Ollama model (qwen2.5-coder:7b) 

qwen2.5:14b doesn't exist on the server, causing silent 404 from
Ollama and 502 to the client. Also added error logging for non-ok
Ollama responses.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
Jeff Emmett 2026-03-25 16:10:11 -07:00
parent 1a422f06ac
commit 55b973ebc2
1 changed files with 4 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
server/index.ts
View File

@ -1594,7 +1594,7 @@ app.post("/api/blender-gen", async (c) => {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({
model: process.env.OLLAMA_MODEL || "qwen2.5:14b",
model: process.env.OLLAMA_MODEL || "qwen2.5-coder:7b",
prompt: `Generate a Blender Python script that creates: ${prompt}\n\nThe script should:\n- Import bpy\n- Clear the default scene\n- Create the described objects with materials\n- Set up basic lighting and camera\n- Render to /tmp/render.png at 1024x1024\n\nOnly output the Python code, no explanations.`,
stream: false,
}),
@ -1606,6 +1606,9 @@ app.post("/api/blender-gen", async (c) => {
// Extract code block if wrapped in markdown
const codeMatch = script.match(/```(?:python)?\n([\s\S]*?)```/);
if (codeMatch) script = codeMatch[1].trim();
} else {
const errText = await llmRes.text().catch(() => "");
console.error(`[blender-gen] Ollama ${llmRes.status}: ${errText}`);
}
} catch (e) {
console.error("[blender-gen] LLM error:", e);