turbollm 1.10.3 → 1.10.4
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/dist/cli.js +4 -1
- package/package.json +1 -1
package/dist/cli.js
CHANGED
|
@@ -8188,7 +8188,7 @@ function profileToArgs(p, m, caps, cores = 0, sys, binPath) {
|
|
|
8188
8188
|
}
|
|
8189
8189
|
if (has("--parallel")) a.push("--parallel", String(p.parallel));
|
|
8190
8190
|
if (p.parallel > 1 && p.kvUnified && has("--kv-unified")) a.push("--kv-unified");
|
|
8191
|
-
if (m.moe && !p.nCpuMoeFit && p.
|
|
8191
|
+
if (m.moe && !p.nCpuMoeFit && p.ngl > 0 && has("--n-cpu-moe")) a.push("--n-cpu-moe", String(p.nCpuMoe));
|
|
8192
8192
|
const kvOk = (t) => caps.kvTypes.includes(t);
|
|
8193
8193
|
if (p.kvTypeK !== "f16" && has("--cache-type-k") && kvOk(p.kvTypeK)) a.push("--cache-type-k", p.kvTypeK);
|
|
8194
8194
|
if (p.kvTypeV !== "f16" && has("--cache-type-v") && kvOk(p.kvTypeV)) a.push("--cache-type-v", p.kvTypeV);
|
|
@@ -22019,6 +22019,9 @@ function registerApi(app2, d) {
|
|
|
22019
22019
|
return err4(c, 400, "invalid_profile_value", "gpu.tensorParallelSize must be an integer \u2265 1.");
|
|
22020
22020
|
}
|
|
22021
22021
|
}
|
|
22022
|
+
if (p.nCpuMoe !== void 0 && (typeof p.nCpuMoe !== "number" || !Number.isFinite(p.nCpuMoe) || p.nCpuMoe < 0)) {
|
|
22023
|
+
return err4(c, 400, "invalid_profile_value", "nCpuMoe must be a non-negative number.");
|
|
22024
|
+
}
|
|
22022
22025
|
const engineId = c.req.query("engine") || d.registry.active()?.id || "*";
|
|
22023
22026
|
d.store.update((cfg2) => {
|
|
22024
22027
|
setModelProfile(cfg2, key, engineId, p);
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "turbollm",
|
|
3
|
-
"version": "1.10.
|
|
3
|
+
"version": "1.10.4",
|
|
4
4
|
"description": "TurboLLM — local LLM platform: run any inference engine auto-tuned to your GPU, with a web UI and OpenAI/Anthropic-compatible API. Point Claude Code at your own machine in one command.",
|
|
5
5
|
"license": "FSL-1.1-ALv2",
|
|
6
6
|
"author": "Mohit Soni",
|