server : update prompt on slot restore (#9800)

ggerganov · web-flow · commit 32da4a2f44bd · 2024-10-11T09:16:00.000+03:00
diff --git a/examples/server/server.cpp b/examples/server/server.cpp
@@ -1778,6 +1778,9 @@ struct server_context {
                     }
                     slot->cache_tokens.resize(token_count);
 
+                    // TODO: maybe detokenize the slot->cache_tokens instead?
+                    slot->prompt = string_format("[restored %d tokens from file]", (int) token_count);
+
                     const int64_t t_end = ggml_time_us();
                     const double t_restore_ms = (t_end - t_start) / 1000.0;
 

Original file line number	Diff line number	Diff line change
`@@ -1778,6 +1778,9 @@ struct server_context {`
`1778`	`1778`	`}`
`1779`	`1779`	`slot->cache_tokens.resize(token_count);`
`1780`	`1780`
	`1781`	`+ // TODO: maybe detokenize the slot->cache_tokens instead?`
	`1782`	`+ slot->prompt = string_format("[restored %d tokens from file]", (int) token_count);`
	`1783`	`+`
`1781`	`1784`	`const int64_t t_end = ggml_time_us();`
`1782`	`1785`	`const double t_restore_ms = (t_end - t_start) / 1000.0;`
`1783`	`1786`