diff options
Diffstat (limited to 'ggml.h')
| -rw-r--r-- | ggml.h | 45 |
1 files changed, 36 insertions, 9 deletions
@@ -1055,7 +1055,6 @@ extern "C" { size_t nb1, size_t offset); - // a -> b, return view(b) GGML_API struct ggml_tensor * ggml_cpy( struct ggml_context * ctx, @@ -1078,6 +1077,33 @@ extern "C" { struct ggml_context * ctx, struct ggml_tensor * a); + // make contiguous, with new shape + GGML_API struct ggml_tensor * ggml_cont_1d( + struct ggml_context * ctx, + struct ggml_tensor * a, + int64_t ne0); + + GGML_API struct ggml_tensor * ggml_cont_2d( + struct ggml_context * ctx, + struct ggml_tensor * a, + int64_t ne0, + int64_t ne1); + + GGML_API struct ggml_tensor * ggml_cont_3d( + struct ggml_context * ctx, + struct ggml_tensor * a, + int64_t ne0, + int64_t ne1, + int64_t ne2); + + GGML_API struct ggml_tensor * ggml_cont_4d( + struct ggml_context * ctx, + struct ggml_tensor * a, + int64_t ne0, + int64_t ne1, + int64_t ne2, + int64_t ne3); + // return view(a), b specifies the new shape // TODO: when we start computing gradient, make a copy instead of view GGML_API struct ggml_tensor * ggml_reshape( @@ -1225,14 +1251,15 @@ extern "C" { struct ggml_tensor * b); // rotary position embedding - // if mode & 1 == 1, skip n_past elements + // if mode & 1 == 1, skip n_past elements (DEPRECATED) // if mode & 2 == 1, GPT-NeoX style // if mode & 4 == 1, ChatGLM style - // TODO: avoid creating a new tensor every time + // + // b is an int32 vector with size a->ne[2], it contains the positions GGML_API struct ggml_tensor * ggml_rope( struct ggml_context * ctx, struct ggml_tensor * a, - int n_past, + struct ggml_tensor * b, int n_dims, int mode, int n_ctx); @@ -1241,7 +1268,7 @@ extern "C" { GGML_API struct ggml_tensor * ggml_rope_inplace( struct ggml_context * ctx, struct ggml_tensor * a, - int n_past, + struct ggml_tensor * b, int n_dims, int mode, int n_ctx); @@ -1250,7 +1277,7 @@ extern "C" { GGML_API struct ggml_tensor * ggml_rope_custom( struct ggml_context * ctx, struct ggml_tensor * a, - int n_past, + struct ggml_tensor * b, int n_dims, int mode, int n_ctx, @@ -1261,7 +1288,7 @@ extern "C" { GGML_API struct ggml_tensor * ggml_rope_custom_inplace( struct ggml_context * ctx, struct ggml_tensor * a, - int n_past, + struct ggml_tensor * b, int n_dims, int mode, int n_ctx, @@ -1272,7 +1299,7 @@ extern "C" { GGML_API struct ggml_tensor * ggml_rope_xpos_inplace( struct ggml_context * ctx, struct ggml_tensor * a, - int n_past, + struct ggml_tensor * b, int n_dims, float base, bool down); @@ -1282,7 +1309,7 @@ extern "C" { GGML_API struct ggml_tensor * ggml_rope_back( struct ggml_context * ctx, struct ggml_tensor * a, - int n_past, + struct ggml_tensor * b, int n_dims, int mode, int n_ctx, |