graph : fix reuse check for recurrent inputs

Author: ggerganov

src/llama-graph.cpp

@@ -263,6 +263,9 @@ bool llm_graph_input_rs::can_reuse(const llm_graph_params & params) {
     res &= s_copy_main->ne[0]  == params.ubatch.n_seqs;
     res &= s_copy_extra->ne[0] == mctx->get_n_rs() - params.ubatch.n_seqs;
 
+    res &= head == mctx->get_head();
+    res &= rs_z == mctx->get_rs_z();
+
     return res;
 }
 
@@ -487,6 +490,9 @@ bool llm_graph_input_mem_hybrid::can_reuse(const llm_graph_params & params) {
     res &= inp_rs->s_copy_main->ne[0]  == params.ubatch.n_seqs;
     res &= inp_rs->s_copy_extra->ne[0] == mctx->get_recr()->get_n_rs() - params.ubatch.n_seqs;
 
+    res &= inp_rs->head == mctx->get_recr()->get_head();
+    res &= inp_rs->rs_z == mctx->get_recr()->get_rs_z();
+
     return res;
 }
 
@@ -1827,6 +1833,9 @@ static std::unique_ptr<llm_graph_input_rs> build_rs_inp_impl(
     inp->s_copy_main  = ggml_view_1d(ctx0, inp->s_copy, n_seqs, 0);
     inp->s_copy_extra = ggml_view_1d(ctx0, inp->s_copy, n_rs - n_seqs, n_seqs * inp->s_copy->nb[0]);
 
+    inp->head = mctx_cur->get_head();
+    inp->rs_z = mctx_cur->get_rs_z();
+
     return inp;
 }
 
@@ -1895,7 +1904,7 @@ ggml_tensor * llm_graph_context::build_rwkv_token_shift_store(
 llm_graph_input_mem_hybrid * llm_graph_context::build_inp_mem_hybrid() const {
     const auto * mctx_cur = static_cast<const llama_memory_hybrid_context *>(mctx);
 
-    auto inp_rs   = build_rs_inp_impl(ctx0, ubatch, mctx_cur->get_recr());
+    auto inp_rs   = build_rs_inp_impl     (ctx0, ubatch, mctx_cur->get_recr());
     auto inp_attn = build_attn_inp_kv_impl(ctx0, ubatch, hparams, cparams, mctx_cur->get_attn());
 
     auto inp = std::make_unique<llm_graph_input_mem_hybrid>(cparams, std::move(inp_attn), std::move(inp_rs), mctx_cur);

src/llama-graph.h

@@ -234,6 +234,10 @@ class llm_graph_input_rs : public llm_graph_input_i {
     ggml_tensor * s_copy_extra;  // I32 [n_rs - n_seqs]
 
     const llama_memory_recurrent_context * mctx;
+
+    // used in view offsets, need to match for valid graph reuse
+    uint32_t head;
+    int32_t rs_z;
 };
 
 class llm_graph_input_cross_embd : public llm_graph_input_i {