dev : add reminders to update backends

ggml-ci

Author: ggerganov

ggml-kompute.cpp

@@ -1680,6 +1680,9 @@ static void ggml_vk_graph_compute(struct ggml_kompute_context * ctx, struct ggml
 #pragma message("      https://github.com/ggerganov/llama.cpp/pull/7225")
                         GGML_ASSERT(dst->src[2] == nullptr && "phi3 frequency factors not implemented yet");
 
+#pragma message("TODO: update rope NORM mode to match NEOX mode")
+#pragma message("      https://github.com/ggerganov/llama.cpp/pull/7634")
+
                         GGML_ASSERT(ne10 == ne02);
                         GGML_ASSERT(src0t == dstt);
                         // const int n_past = ((int32_t *) dst->op_params)[0];

ggml-sycl.cpp

@@ -14029,6 +14029,9 @@ inline void ggml_sycl_op_rope(const ggml_tensor *src0, const ggml_tensor *src1,
 
     const bool is_neox = mode & 2;
 
+#pragma message("TODO: update rope NORM mode to match NEOX mode")
+#pragma message("      https://github.com/ggerganov/llama.cpp/pull/7634")
+
     if (is_neox) {
         pos = (const int32_t *) src1_dd;
 

ggml-vulkan.cpp

@@ -4303,6 +4303,9 @@ static void ggml_vk_rope(ggml_backend_vk_context * ctx, vk_context * subctx, con
 
     const bool is_neox = mode & 2;
 
+#pragma message("TODO: update rope NORM mode to match NEOX mode")
+#pragma message("      https://github.com/ggerganov/llama.cpp/pull/7634")
+
     float corr_dims[2];
     ggml_rope_yarn_corr_dims(n_dims, n_ctx_orig, freq_base, beta_fast, beta_slow, corr_dims);