[Bugfix] Fix LFM2 ShortConv prefix breaking quant ignore list (#51495)
Signed-off-by: ylsun <1654366371@qq.com> Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
ylsun committed
Aug 8, 2026 at 19:27 UTC
0fe9a916e88e884d6ee7fa20c4b15f5a9610218f
2 files changed
+2
-2
vllm/model_executor/models/lfm2.py
+1
-1
@@ -258,7 +258,7 @@ class Lfm2ShortConvDecoderLayer(nn.Module):
258
model_config=model_config,
259
cache_config=cache_config,
260
quant_config=quant_config,
261
- prefix=f"{prefix}.conv",
261
+ prefix=f"{prefix}.short_conv",
262
)
263
264
self.feed_forward = Lfm2MLP(
vllm/model_executor/models/lfm2_moe.py
+1
-1
@@ -354,7 +354,7 @@ class Lfm2MoeShortConvDecoderLayer(nn.Module):
354
model_config=model_config,
355
cache_config=cache_config,
356
quant_config=quant_config,
357
- prefix=f"{prefix}.conv",
357
+ prefix=f"{prefix}.short_conv",
358
)
359
360
if layer_idx < config.num_dense_layers: