From: Georgi Gerganov <redacted>
Date: Thu, 27 Nov 2025 14:34:13 +0000 (+0200)
Subject: arch : add description about LLM_TENSOR_INFOS (#17550)
X-Git-Tag: upstream/0.0.7446~268
X-Git-Url: https://git.djapps.eu/?a=commitdiff_plain;h=c386114922dcc54a041ed99f855defda3fa7f225;p=pkg%2Fggml%2Fsources%2Fllama.cpp

arch : add description about LLM_TENSOR_INFOS (#17550)
---

diff --git a/src/llama-arch.cpp b/src/llama-arch.cpp
index 6da9e0a3..f6e26245 100644
--- a/src/llama-arch.cpp
+++ b/src/llama-arch.cpp
@@ -2487,6 +2487,16 @@ static const std::map<llm_arch, std::map<llm_tensor, const char *>> LLM_TENSOR_N
     },
 };
 
+// declare information about the model weight tensors:
+// - the layer in which the tensor is going to be used. this is needed in order to assign the correct buffer type for the weight
+// - the operator which is going to use the weight. this is needed to determine if the respective backend supports the operator
+//
+// for example, input layers are usually assigned to CPU/host buffer types
+//
+// a mismatch between the declared information and the actual layer/op in which the tensor is used can lead to sub-optimal
+//   assignment of the buffer types and extra overhead during computation
+// example: https://github.com/ggml-org/llama.cpp/pull/17548
+//
 static const std::map<llm_tensor, llm_tensor_info> LLM_TENSOR_INFOS = {
     {LLM_TENSOR_TOKEN_EMBD,                 {LLM_TENSOR_LAYER_INPUT, GGML_OP_GET_ROWS}},
     {LLM_TENSOR_POS_EMBD,                   {LLM_TENSOR_LAYER_INPUT, GGML_OP_GET_ROWS}},