spec : add self‑speculative decoding (no draft model required) + refactor (#18471)

* server: introduce self-speculative decoding

* server: moved self-call into speculative.cpp

* can_speculate() includes self-speculation

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

* server: can_speculate() tests self-spec

* server: replace can_speculate() with slot.can_speculate()

Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>

* common: use %zu format specifier for size_t in logging

Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>

* server: can_speculate() requires a task instance

* common: ngram map, config self-speculative decoding

* common: add enum common_speculative_type

* common: add vector of speculative states

* common: add option --spec-draftless

* server: cleanup (remove slot.batch_spec, rename)

* common: moved self-spec impl to ngram-map

* common: cleanup (use common_speculative_state_draft)

* spec : refactor

* cont : naming

* spec: remove --spec-config

* doc: (draftless) speculative decoding

* common: print performance in spec decoding

* minor : cleanup

* common : better names

* minor : cleanup + fix build

* minor: comments

* CODEOWNERS: add common/ngram-map.* (#18471)

* common : rename speculative.draftless_type -> speculative.type

* ngram-map : fix uninitialized values

* ngram-map : take into account the input can become shorter

* ngram-map : revert len check for now

* arg : change `--spec-draftless` -> `--spec-type`

* spec : add common_speculative_state::accept()

* spec : refactor + add common_speculative_begin()

* spec : fix begin() call with mtmd

* spec : additional refactor + remove common_speculative_params

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>

common/speculative.h

@@ -5,31 +5,33 @@
 
 struct common_speculative;
 
-struct common_speculative_params {
-    int n_draft = 16;  // max drafted tokens
-    int n_reuse = 256;
+// comma separated list of all types
+std::string common_speculative_type_name_str();
 
-    float p_min = 0.75f; // min probability required to accept a token in the draft
-};
+// convert string to type
+enum common_speculative_type common_speculative_type_from_name(const std::string & name);
 
-struct common_speculative * common_speculative_init(
-        struct llama_context * ctx_tgt,
-        struct llama_context * ctx_dft
-);
+// convert type to string
+std::string common_speculative_type_to_str(enum common_speculative_type type);
 
-void common_speculative_free(struct common_speculative * spec);
+common_speculative * common_speculative_init(
+        const common_params_speculative & params,
+              llama_context             * ctx_tgt);
 
-bool common_speculative_are_compatible(
-        const struct llama_context * ctx_tgt,
-        const struct llama_context * ctx_dft);
+void common_speculative_free(common_speculative * spec);
 
-void common_speculative_add_replacement_tgt_dft(
-        struct common_speculative * spec,
-        const char *source, const char *dest);
+// optionally call once at the beginning of a new generation
+void common_speculative_begin(common_speculative * spec, const llama_tokens & prompt);
 
 // sample up to n_draft tokens and add them to the batch using the draft model
-llama_tokens common_speculative_gen_draft(
-               struct common_speculative * spec,
-        struct common_speculative_params   params,
-                      const llama_tokens & prompt,
-                             llama_token   id_last);
+llama_tokens common_speculative_draft(
+                     common_speculative * spec,
+        const common_params_speculative & params,
+                     const llama_tokens & prompt,
+                            llama_token   id_last);
+
+// informs the speculative decoder that n_accepted tokens were accepted by the target model
+void common_speculative_accept(common_speculative * spec, uint16_t n_accepted);
+
+// print statistics about the speculative decoding
+void common_speculative_print_stats(const common_speculative * spec);