A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 2, 14:45 UTC
Fix `OlmoHybridForCausalLM` not initialising (#43846)
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com> Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
Harry Mellor committed
May 28, 2026 at 13:33 UTC
19af4e6dd49bdbe5c7c521092c33a94279c9ccea
1 file changed
+2
-2
vllm/transformers_utils/config.py
+2
-2
@@ -430,9 +430,9 @@ def patch_legacy_rope_type(rope_parameters: dict[str, Any] | None) -> None:
430
if "rope_type" not in rope_parameters and "type" in rope_parameters:
431
rope_parameters["rope_type"] = rope_parameters["type"]
432
logger.info("Replacing legacy 'type' key with 'rope_type'")
433
- # Case 3: No rope_type field at all - cannot determine RoPE type, raise error
433
+ # Case 3: No rope_type field present - nothing to patch
434
if "rope_type" not in rope_parameters:
435
- raise ValueError("rope_parameters should have a 'rope_type' key")
435
+ return
436
# Patch legacy rope_type values with warning
437
if rope_parameters["rope_type"] == "su":
438
rope_parameters["rope_type"] = "longrope"