LLM inference in C/C++
C++ ggml
tests: tolerate a shared pool abort in test_completion_unified (#28759)
Pascal8172e6577ac2
diff --git a/tools/server/tests/unit/test_completion.py b/tools/server/tests/unit/test_completion.pyindex 9375e0110e..01732eb163 100644--- a/tools/server/tests/unit/test_completion.py+++ b/tools/server/tests/unit/test_completion.py@@ -394,7 +394,12 @@ def test_completion_unified(n_ctx, n_slots, n_predict_vals, expected_success):results = parallel_function_calls(tasks)for res, n_predict, expect_ok in zip(results, n_predict_vals, expected_success):if expect_ok:- assert res.status_code == 200+ # the pool is aborted as a whole, so a request that fits on its own+ # is still dropped when the slots overlap, and it says so explicitly+ assert res.status_code == 200 or (+ res.status_code == 500+ and "context size has been exceeded" in res.body["error"]["message"].lower()+ )# note: https://github.com/ggml-org/llama.cpp/pull/18700#issuecomment-3728695581if res.status_code == 200: