Fix code to work with latest GGML.

bert.cpp

@@ -458,7 +458,7 @@ struct bert_ctx * bert_load_from_file(const char *fname)
         model_mem_req += n_layer * (n_intermediate * ggml_type_sizef(GGML_TYPE_F32)); // ff_i_b
         model_mem_req += n_layer * (n_embd * ggml_type_sizef(GGML_TYPE_F32)); // ff_o_b
 
-        model_mem_req += (5 + 16 * n_layer) * 256; // object overhead
+        model_mem_req += (5 + 16 * n_layer) * 512; // object overhead
 
         printf("%s: ggml ctx size = %6.2f MB\n", __func__, model_mem_req / (1024.0 * 1024.0));
     }
@@ -678,8 +678,8 @@ struct bert_ctx * bert_load_from_file(const char *fname)
     // Calculate space requirements for setting up context buffers later
     {
         bert_vocab_id tokens[] = {0, 1, 2, 3};
-        // TODO: We set the initial buffer size to 16MB and hope it's enough. Maybe there is a better way to do this?
-        new_bert->buf_compute.resize(16 * 1024 * 1024);
+        // TODO: We set the initial buffer size to 32MB and hope it's enough. Maybe there is a better way to do this?
+        new_bert->buf_compute.resize(32 * 1024 * 1024);
         bert_eval(new_bert, 1, tokens, 4, nullptr);
         new_bert->max_batch_n = 0;
 
@@ -688,7 +688,7 @@ struct bert_ctx * bert_load_from_file(const char *fname)
         new_bert->mem_per_input = 1.1 * (new_bert->mem_per_token * N); // add 10% to account for ggml object overhead
 
     }
-    printf("%s: mem_per_token %zd KB, mem_per_input %lld MB\n", __func__, new_bert->mem_per_token / (1 << 10), new_bert->mem_per_input / (1 << 20));
+    printf("%s: mem_per_token %d KB, mem_per_input %lld MB\n", __func__, new_bert->mem_per_token / (1 << 10), new_bert->mem_per_input / (1 << 20));
 
     return new_bert;
 }
@@ -779,7 +779,6 @@ void bert_eval_batch(
 
         struct ggml_context *ctx0 = ggml_init(params);
         struct ggml_cgraph gf = {};
-        gf.n_threads = n_threads;
 
         // Embeddings. word_embeddings + token_type_embeddings + position_embeddings
         struct ggml_tensor *token_layer = ggml_new_tensor_1d(ctx0, GGML_TYPE_I32, N);
@@ -916,7 +915,7 @@ void bert_eval_batch(
         ggml_tensor *output = inpL;
         // run the computation
         ggml_build_forward_expand(&gf, output);
-        ggml_graph_compute(ctx0, &gf);
+        ggml_graph_compute_with_ctx(ctx0, &gf, n_threads);
 
 
         // float *dat = ggml_get_data_f32(output);