Clean up fp8 support (#4230)

2025-03-09 21:46:35 -07:00
parent fbd560028a
commit e8a69e4d0c
5 changed files with 86 additions and 110 deletions
--- a/python/sglang/srt/layers/vocab_parallel_embedding.py
+++ b/python/sglang/srt/layers/vocab_parallel_embedding.py
@@ -264,7 +264,6 @@ class VocabParallelEmbedding(torch.nn.Module):
        quant_method = None
        if quant_config is not None:
            quant_method = quant_config.get_quant_method(self, prefix=prefix)
-            print("quant_method", quant_method)
        if quant_method is None:
            quant_method = UnquantizedEmbeddingMethod()