CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-03 09:22:01 +00:00
Code Issues Packages Projects Releases Wiki Activity

llama : support running Mamba-Codestral-7B-v0.1

Browse Source
This commit is contained in:
Francis Couture-Harpin
2024-08-18 22:43:39 -04:00
parent dceff23fae
commit 2bfe9de6d3
2 changed files with 5 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
convert_hf_to_gguf.py
View File

@@ -2843,6 +2843,10 @@ class Mamba2Model(Model):
def modify_tensors(self, data_torch: Tensor, name: str, bid: int | None) -> Iterable[tuple[str, Tensor]]:
del bid # unused
if name.startswith("model.backbone") or name.startswith("model.lm_head"):
# map Mamba-Codestral-7B-v0.1 tensor names to the names used by Mamba-2
name = name.removeprefix("model.")
if name.endswith(".dt_bias"):
name = name.rpartition(".dt_bias")[0] + ".dt_proj.bias"