mirror of
				https://github.com/ggml-org/llama.cpp.git
				synced 2025-10-31 08:51:55 +00:00 
			
		
		
		
	llama : tokenizer fixes (#2549)
* Merge tokenizer fixes into the gguf branch. * Add test vocabularies
This commit is contained in:
		| @@ -62,7 +62,7 @@ int main(int argc, char ** argv) { | ||||
|     fprintf(stderr, "\n\n"); | ||||
|  | ||||
|     for (auto id : tokens_list) { | ||||
|         fprintf(stderr, "%s", llama_token_to_str(ctx, id)); | ||||
|         fprintf(stderr, "%s", llama_token_to_str(ctx, id).c_str()); | ||||
|     } | ||||
|  | ||||
|     fflush(stderr); | ||||
| @@ -109,7 +109,7 @@ int main(int argc, char ** argv) { | ||||
|         } | ||||
|  | ||||
|         // print the new token : | ||||
|         printf("%s", llama_token_to_str(ctx, new_token_id)); | ||||
|         printf("%s", llama_token_to_str(ctx, new_token_id).c_str()); | ||||
|         fflush(stdout); | ||||
|  | ||||
|         // push this new token for next evaluation | ||||
|   | ||||
		Reference in New Issue
	
	Block a user
	 goerch
					goerch