LLM-Inference/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-08-12 22:31:11 +04:00
Code Issues Packages Projects Releases Wiki Activity

convert : fix DeepseekV4 rope parameters with transformers 5.x (#26673)

Browse Source
This commit is contained in:
Sigbjørn Skjæret
2026-08-06 16:06:52 +03:00
committed by GitHub
parent 3db4ff877d
commit eb5667a169
1 changed files with 7 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
conversion/deepseek.py
+7
View File
@@ -533,6 +533,13 @@ class DeepseekV4Model(TextModel):
for key, value in raw_hparams.items():
self.hparams.setdefault(key, value)
# workaround for special rope_parameters (main/compress) in transformers 5.x
if self.rope_parameters.get("full_attention", self.rope_parameters).get("rope_type") is None:
if (rope_scaling := raw_hparams.get("rope_scaling")) is not None:
if "rope_type" not in rope_scaling and (rope_type := rope_scaling.get("type")) is not None:
rope_scaling["rope_type"] = rope_type
self.rope_parameters.update(**rope_scaling)
self.block_count = self.hparams["num_hidden_layers"]
if self.mtp_only:
self.block_count += self.hparams.get("num_nextn_predict_layers", 0)