[Migration] Migrate GGUF quantization support to plugin (#39612)

Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
This commit is contained in:
Isotr0py
2026-06-12 12:02:21 -07:00
committed by GitHub
parent d6fd7ce8da
commit 6635279d8a
57 changed files with 72 additions and 9048 deletions
+2
View File
@@ -279,12 +279,14 @@ class OlmoeModel(nn.Module):
super().__init__()
config = vllm_config.model_config.hf_config
quant_config = vllm_config.quant_config
self.vocab_size = config.vocab_size
self.config = config
self.embed_tokens = VocabParallelEmbedding(
config.vocab_size,
config.hidden_size,
quant_config=quant_config,
)
self.start_layer, self.end_layer, self.layers = make_layers(
config.num_hidden_layers,