Non-autoregressive System 1 decision engine. Typed choice, score and yes/no decisions over any text in a single forward pass, in 100+ languages, with a router that picks the right checkpoint per request.
Python calibrationclassificationdecision-modelhuggingfacejevmodernbertmultilingualnlppythonpytorchroutingtyped-decisionszero-shot
ci(workflows): run missing test suites in CI and guard against test omission
Swatantra Yadav6c5751328cbb
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.ymlindex 7a9cca5f..4e36a4fb 100644--- a/.github/workflows/ci.yml+++ b/.github/workflows/ci.yml@@ -132,6 +132,10 @@ jobs:python tests/test_portability.pypython tests/test_training.pypython tests/test_example_server_limits.py+ python tests/test_blank_lang_routing.py+ python tests/test_export_onnx_safety.py+ python tests/test_load_errors.py+ python tests/test_revision_pinning.py- name: Email cleaning testsrun: python tests/test_email.py@@ -141,7 +145,7 @@ jobs:- name: Chinese short-command benchmark testsrun: python -m unittest discover -s research/benchmarks/zh_short_commands/tests -v- # These four are pytest suites, and `python tests/<name>.py` does not run them: it defines+ # These are pytest suites, and `python tests/<name>.py` does not run them: it defines# the test functions and exits 0, having asserted nothing. They were counted by the suite# list above without ever executing, which is how the HTTP surface came to have no coverage# in CI at all (#374).@@ -153,6 +157,7 @@ jobs:run: |python -m pytest tests/test_serve.py \tests/test_router_batch.py \+ tests/test_predict_batch.py \tests/test_system_one_lang.py \tests/test_audit_regressions.py \tests/test_truncation_direction.py \@@ -233,13 +238,19 @@ jobs:# pass -- a change to a container script, so it belongs in its own PR, not this one.- name: Routing and language-detection testsrun: |- for t in test_router test_criteria test_download test_shortlist \- test_decision_model test_packaging test_lang_guess \- test_identifier_complexity \- test_calibration_persistence test_context_manager \+ for t in test_router test_criteria test_batch test_predict_long \+ test_structured test_structured_api test_onnx_lang_parity \+ test_download test_shortlist test_decision_model \+ test_head_checkpointing test_packaging test_doc_tables \+ test_tokenizer_cache test_tokenizer_concurrency test_lazy_import \+ test_runtime_fixes test_lang_guess test_identifier_complexity \+ test_lang_stats test_calibration_persistence test_context_manager \test_criteria_normalization test_docker_entrypoint \test_empty_questions test_router_memory test_shortlist_cosine \- test_temperature_loading; do+ test_temperature_loading test_confidence test_cli test_mcp \+ test_langchain test_portability test_training \+ test_example_server_limits test_blank_lang_routing \+ test_export_onnx_safety test_load_errors test_revision_pinning; doif [ "$t" = "test_docker_entrypoint" ]; thenecho "skipping $t: docker/entrypoint.py is Linux-container only"continue@@ -254,13 +265,14 @@ jobs:python tests/test_hooks_api.pyshell: bash- # Same five pytest suites as the Linux job. They are named explicitly rather than added to+ # Same pytest suites as the Linux job. They are named explicitly rather than added to# the loop above, because the loop runs each file as a script and that is exactly what does# not execute a pytest suite (#374).- name: Pytest suites (serve, batch, system-one, audit regressions, truncation)run: |python -m pytest tests/test_serve.py \tests/test_router_batch.py \+ tests/test_predict_batch.py \tests/test_system_one_lang.py \tests/test_audit_regressions.py \tests/test_truncation_direction.py \diff --git a/.github/workflows/release.yml b/.github/workflows/release.ymlindex 8344b4ba..27128776 100644--- a/.github/workflows/release.yml+++ b/.github/workflows/release.yml@@ -42,14 +42,23 @@ jobs:python tests/test_criteria.pypython tests/test_batch.pypython tests/test_predict_long.py+ python tests/test_hooks.py+ python tests/test_hooks_api.pypython tests/test_onnx_lang_parity.pypython tests/test_email.pypython tests/test_download.pypython tests/test_shortlist.pypython tests/test_decision_model.py+ python tests/test_head_checkpointing.pypython tests/test_packaging.pypython tests/test_doc_tables.py+ python tests/test_tokenizer_cache.py+ python tests/test_tokenizer_concurrency.py+ python tests/test_lazy_import.py+ python tests/test_runtime_fixes.pypython tests/test_lang_guess.py+ python tests/test_identifier_complexity.py+ python tests/test_lang_stats.pypython tests/test_calibration_persistence.pypython tests/test_context_manager.pypython tests/test_criteria_normalization.py@@ -58,7 +67,13 @@ jobs:python tests/test_router_memory.pypython tests/test_shortlist_cosine.pypython tests/test_temperature_loading.py+ python tests/test_confidence.pypython tests/test_cli.py+ python tests/test_training.py+ python tests/test_portability.py+ python tests/test_blank_lang_routing.py+ python tests/test_load_errors.py+ python tests/test_revision_pinning.py- name: Buildrun: python -m builddiff --git a/tests/test_packaging.py b/tests/test_packaging.pyindex 66971a35..45c5cfbf 100644--- a/tests/test_packaging.py+++ b/tests/test_packaging.py@@ -4,6 +4,7 @@ Text parsing, not tomllib: the floor is 3.10 and tomllib arrives in 3.11."""import osimport re+import shleximport sysROOT = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))@@ -93,6 +94,95 @@ missing_from_ci = []check("ci/tests every advertised Python version", missing_from_ci, [])+# Every test in tests/ must be wired into CI workflows (ci.yml or docker.yml),+# unless explicitly exempted with a documented rationale (#399).+def _clean_command_line(line):+ line = line.strip()+ if not line or line.startswith("#"):+ return ""+ try:+ return " ".join(shlex.split(line, comments=True))+ except ValueError:+ return re.sub(r"(?:\s+|^)#.*$", "", line).strip()+++def _invoked_workflow_tests(yaml_text):+ """Extract test files that are actually executed by python or pytest in run: steps."""+ run_lines = []+ in_run = False+ run_indent = 0+ for line in yaml_text.splitlines():+ indent = len(line) - len(line.lstrip())+ match = re.match(r"^(\s*)-\s+run:\s*(\|?>?)(.*)$", line) or re.match(r"^(\s*)run:\s*(\|?>?)(.*)$", line)+ if match:+ in_run = True+ run_indent = indent+ inline_cmd = match.group(3).strip()+ cleaned = _clean_command_line(inline_cmd)+ if cleaned:+ run_lines.append(cleaned)+ continue+ if in_run:+ if line.strip() and indent <= run_indent:+ in_run = False+ else:+ stripped = line.strip()+ if stripped and not stripped.startswith("#"):+ cleaned = _clean_command_line(stripped)+ if cleaned:+ if stripped.endswith("\\") and not cleaned.endswith("\\"):+ cleaned += " \\"+ run_lines.append(cleaned)++ # Merge backslash continuation lines (e.g. multi-line pytest argument lists)+ merged_lines = []+ buf = []+ for line in run_lines:+ if line.endswith("\\"):+ buf.append(line[:-1].strip())+ else:+ if buf:+ buf.append(line)+ merged_lines.append(" ".join(buf))+ buf = []+ else:+ merged_lines.append(line)+ if buf:+ merged_lines.append(" ".join(buf))++ # Match executable python or pytest invocations, ignoring mentions in echo/cat/test+ invoked = set()+ for cmd in merged_lines:+ for part in re.split(r";|&&|\|\||\|", cmd):+ part = part.strip()+ if re.search(r"\b(?:pytest|python(?:\d+(?:\.\d+)?)?\s+-m\s+pytest)\b", part):+ invoked.update(re.findall(r"\btests/(test_[a-zA-Z0-9_]+\.py)\b", part))+ elif re.search(r"\bpython(?:\d+(?:\.\d+)?)?\s+.*?tests/(test_[a-zA-Z0-9_]+\.py)\b", part):+ invoked.update(re.findall(r"\btests/(test_[a-zA-Z0-9_]+\.py)\b", part))+ return invoked+++docker_workflow = read(os.path.join(".github", "workflows", "docker.yml"))+registered_test_files = _invoked_workflow_tests(workflow) | _invoked_workflow_tests(docker_workflow)++EXEMPT_TEST_SUITES = {+ "test_local_e2e.py": "Requires local checkpoints under ~/laya_models (AGENTS.md)",+ "test_mcp_local_e2e.py": "Requires local checkpoints under ~/laya_models (AGENTS.md)",+ "test_onnx.py": "Requires onnx extra; skip-guarded on lane without it (AGENTS.md)",+ "test_fast.py": "Requires CUDA and tilelang extra (AGENTS.md)",+ "test_server_example.py": "Requires cached or downloaded weights for examples/server.py",+}++all_test_files = [+ f for f in os.listdir(os.path.join(ROOT, "tests"))+ if f.startswith("test_") and f.endswith(".py")+]+untested_suites = [+ f for f in sorted(all_test_files)+ if f not in EXEMPT_TEST_SUITES and f not in registered_test_files+]+check("ci/wires every non-exempt test suite", untested_suites, [])+# --------------------------------------------------------------- markdown links# Nothing checked these, and the README ships to PyPI and to the model card. Only