server : fix sampling call

ggml-ci

Author: ggerganov

examples/server/server.cpp

@@ -2308,7 +2308,7 @@ struct server_context {
                 llama_decode(ctx, slot.batch_spec);
 
                 // the accepted tokens from the speculation
-                const auto ids = common_sampler_sample_n(slot.smpl, ctx, draft);
+                const auto ids = common_sampler_sample_and_accept_n(slot.smpl, ctx, draft);
 
                 slot.n_past    += ids.size();
                 slot.n_decoded += ids.size();
@@ -2323,8 +2323,6 @@ struct server_context {
 
                     id = ids[i];
 
-                    common_sampler_accept(slot.smpl, id, true);
-
                     result.tok = id;
 
                     if (!process_token(result, slot)) {