Skip to content

Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦

C

jev-sdk-check: TypeSafe's SDKs against GLiNER2.5-Decide too

JustVuggcd0b86b6d509
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
index e54c0db3..975193b6 100644
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -1129,7 +1129,7 @@ jobs:
run: |
pip install -r c/tools/requirements-jev-sdk.txt
cd c/tools/jev-sdk && npm ci
- - name: Both SDKs against a language model (tiny MiMo) and a decision engine (tiny Laya)
+ - name: Both SDKs against a language model (tiny MiMo) and the decision engines (tiny Laya, tiny GLiNER2.5-Decide)
run: make -C c jev-sdk-check
qwenimage-tiny-check:
diff --git a/c/Makefile b/c/Makefile
index 30131de6..0b4a23e7 100644
--- a/c/Makefile
+++ b/c/Makefile
@@ -1310,11 +1310,13 @@ gliner-decide-tiny-check: gliner-decide-tiny-generate gliner_decide$(EXE)
GLINER_DECIDE_TINY_REQUIRED=1 $(PYTHON) -m unittest -v tests.test_gliner_decide_tiny tests.test_decision_serve
# TypeSafe's own SDKs (Python and TypeScript), unmodified, against `coli serve`
-# on a language model (tiny MiMo) and a decision engine (tiny Laya): switching
-# from Jev is a base-URL change. Needs the pinned SDKs first:
+# on a language model (tiny MiMo) and the decision engines (tiny Laya, tiny
+# GLiNER2.5-Decide): switching from Jev is a base-URL change. Needs the pinned
+# SDKs first:
# pip install -r tools/requirements-jev-sdk.txt && (cd tools/jev-sdk && npm ci)
.PHONY: jev-sdk-check
-jev-sdk-check: mimo$(EXE) laya$(EXE) mimo-tiny-generate laya-tiny-generate
+jev-sdk-check: mimo$(EXE) laya$(EXE) gliner_decide$(EXE) mimo-tiny-generate laya-tiny-generate \
+ gliner-decide-tiny-generate
JEV_SDK_REQUIRED=1 $(PYTHON) -m unittest -v tests.test_jev_sdk
# In make check: every format and ragged edges against the naive product, and
diff --git a/c/tests/test_jev_sdk.py b/c/tests/test_jev_sdk.py
index 2023449c..8611f50b 100644
--- a/c/tests/test_jev_sdk.py
+++ b/c/tests/test_jev_sdk.py
@@ -11,10 +11,10 @@ The official clients, unmodified, pointed at colibri with nothing but
@typesafe-ai/sdk TypeScript, new TypeSafeClient({apiKey, baseURL})
.systemOne(...) and .models.list(), run under node
-Each one is driven against two servers, each a real `coli serve` process:
+Each one is driven against three servers, each a real `coli serve` process:
a language model (the tiny MiMo fixture, which scores options through the
-logprob channel) and a decision engine (the tiny Laya fixture, DECIDE). The
-same requests go to both: every question type, several questions at once,
+logprob channel) and two decision engines (the tiny Laya and GLiNER2.5-Decide
+fixtures, DECIDE). The same requests go to all three: every question type, several questions at once,
text, object and array states, objects and arrays as instructions and
criteria, and a request the server refuses (422), which each SDK must raise
as its own UnprocessableEntity error.
@@ -44,6 +44,8 @@ import mimo_serve_fixture # noqa: E402
REQUIRED = os.environ.get("JEV_SDK_REQUIRED") == "1"
LAYA_FIXTURE = HERE / "laya_tiny"
LAYA_BINARY = HERE / ("laya.exe" if os.name == "nt" else "laya")
+GLINER_FIXTURE = HERE / "gliner_decide_tiny"
+GLINER_BINARY = HERE / ("gliner_decide.exe" if os.name == "nt" else "gliner_decide")
TS_SDK_DIR = Path(os.environ.get("JEV_TS_SDK_DIR", HERE / "tools" / "jev-sdk"))
TS_SCRIPT = HERE / "tests" / "jev_sdk_client.mjs"
API_KEY = "sk-colibri-test"
@@ -138,7 +140,7 @@ def long_mimo_fixture():
def servers():
- """(name, model dir, env) of the two servers, or a skip naming what is missing."""
+ """(name, model dir, env) of the three servers, or a skip naming what is missing."""
found = []
if mimo_serve_fixture.available():
found.append(("mimo-tiny (language model)", long_mimo_fixture(), {"CTX": "4096"}))
@@ -148,8 +150,14 @@ def servers():
found.append(("laya-tiny (decision engine)", LAYA_FIXTURE, {}))
elif REQUIRED:
raise AssertionError("laya or laya_tiny missing (make laya laya-tiny-generate)")
+ if GLINER_BINARY.exists() and (GLINER_FIXTURE / "model.safetensors").exists():
+ found.append(("gliner-decide-tiny (decision engine)", GLINER_FIXTURE, {}))
+ elif REQUIRED:
+ raise AssertionError("gliner_decide or gliner_decide_tiny missing "
+ "(make gliner_decide gliner-decide-tiny-generate)")
if not found:
- raise unittest.SkipTest("no tiny fixture to serve (make mimo laya laya-tiny-generate)")
+ raise unittest.SkipTest("no tiny fixture to serve (make mimo laya laya-tiny-generate "
+ "gliner_decide gliner-decide-tiny-generate)")
return found