ggml : add ggml_row_size() (fixes llama out of space) (#4461)

* Fixes "Not enough space in the context's memory pool" encountered on certain models, which seems to be caused by some imprecision related to the automatic casting of floating point values

* do not cast to size_t, instead just use doubles

* ggml : add ggml_row_size(), deprecate ggml_type_sizef()

* ggml : fix row size compute to avoid overflows

* tests : fix sizey -> sizez

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

1 files changed, 7 insertions, 2 deletions

diff --git a/ggml.c b/ggml.c
index 7e127281..f0a97269 100644
--- a/ggml.c
+++ b/ggml.c
@@ -2011,8 +2011,13 @@ size_t ggml_type_size(enum ggml_type type) {
     return type_traits[type].type_size;
 }
 
-float ggml_type_sizef(enum ggml_type type) {
-    return ((float)(type_traits[type].type_size))/type_traits[type].blck_size;
+size_t ggml_row_size(enum ggml_type type, int64_t ne) {
+    assert(ne % ggml_blck_size(type) == 0);
+    return ggml_type_size(type)*ne/ggml_blck_size(type);
+}
+
+double ggml_type_sizef(enum ggml_type type) {
+    return ((double)(type_traits[type].type_size))/type_traits[type].blck_size;
 }
 
 const char * ggml_type_name(enum ggml_type type) {