Skip to content

LLM inference in C/C++

C++ ggml

convert: fix GLM regression in index_tensors (#27655)

jacekpoplawski71cc86fa41ff
diff --git a/conversion/glm.py b/conversion/glm.py
index 23fdbca881..7544f850cb 100644
--- a/conversion/glm.py
+++ b/conversion/glm.py
@@ -122,7 +122,9 @@ class Glm4MoeModel(TextModel):
self.tensor_map = gguf.get_tensor_name_map(self.model_arch, self.block_count)
def index_tensors(self, remote_hf_model_id: str | None = None):
- type(self)._n_main_layers = self.hparams["num_hidden_layers"]
+ hparams = {**self.hparams, **self.hparams.get("text_config", {})}
+ key = next((k for k in ["n_layers", "num_hidden_layers", "n_layer", "num_layers"] if k in hparams), None)
+ type(self)._n_main_layers = hparams.get(key)
return super().index_tensors(remote_hf_model_id=remote_hf_model_id)
@classmethod