From a56760695716c027f368c907556018f65cdb40e7 Mon Sep 17 00:00:00 2001 From: Csaba Kecskemeti Date: Thu, 31 Jul 2025 07:22:48 -0700 Subject: [PATCH 1/2] MODEL_TENSOR.SSM_DT_NORM has defined twice, and sencond overwritten the jamba model's layername --- gguf-py/gguf/tensor_mapping.py | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/gguf-py/gguf/tensor_mapping.py b/gguf-py/gguf/tensor_mapping.py index 15adbfa781845..480760d817d7e 100644 --- a/gguf-py/gguf/tensor_mapping.py +++ b/gguf-py/gguf/tensor_mapping.py @@ -615,10 +615,6 @@ class TensorNameMap: "model.layers.layers.{bid}.mixer.dt_proj", # plamo2 ), - MODEL_TENSOR.SSM_DT_NORM: ( - "model.layers.{bid}.mamba.dt_layernorm", # jamba - ), - MODEL_TENSOR.SSM_A: ( "model.layers.{bid}.A_log", # mamba-hf "backbone.layers.{bid}.mixer.A_log", # mamba @@ -647,6 +643,7 @@ class TensorNameMap: MODEL_TENSOR.SSM_DT_NORM: ( "model.layers.layers.{bid}.mixer.dt_norm.weight", # plamo2 + "model.layers.{bid}.mamba.dt_layernorm", # jamba ), MODEL_TENSOR.SSM_NORM: ( From 272a174c8a3028732f41e6623875d4b068f2136a Mon Sep 17 00:00:00 2001 From: Csaba Kecskemeti Date: Thu, 31 Jul 2025 07:53:21 -0700 Subject: [PATCH 2/2] correct order --- gguf-py/gguf/tensor_mapping.py | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/gguf-py/gguf/tensor_mapping.py b/gguf-py/gguf/tensor_mapping.py index 480760d817d7e..df490fc80e9b7 100644 --- a/gguf-py/gguf/tensor_mapping.py +++ b/gguf-py/gguf/tensor_mapping.py @@ -615,6 +615,11 @@ class TensorNameMap: "model.layers.layers.{bid}.mixer.dt_proj", # plamo2 ), + MODEL_TENSOR.SSM_DT_NORM: ( + "model.layers.layers.{bid}.mixer.dt_norm.weight", # plamo2 + "model.layers.{bid}.mamba.dt_layernorm", # jamba + ), + MODEL_TENSOR.SSM_A: ( "model.layers.{bid}.A_log", # mamba-hf "backbone.layers.{bid}.mixer.A_log", # mamba @@ -641,11 +646,6 @@ class TensorNameMap: "model.layers.layers.{bid}.mixer.D", # plamo2 ), - MODEL_TENSOR.SSM_DT_NORM: ( - "model.layers.layers.{bid}.mixer.dt_norm.weight", # plamo2 - "model.layers.{bid}.mamba.dt_layernorm", # jamba - ), - MODEL_TENSOR.SSM_NORM: ( "model.layers.{bid}.mamba.norm", # falcon-h1 granite-hybrid "backbone.layers.{bid}.mixer.norm", # mamba2