happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

cont : remove assistant start prefix altogether

This commit is contained in:
Alde Rojas 2026-01-02 08:41:17 -06:00
parent 980c772d44
commit c41d8f8c34
No known key found for this signature in database
2 changed files with 8 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
common/chat.cpp
View File

@ -2568,9 +2568,6 @@ static common_chat_params common_chat_params_init_solar_open(const common_chat_t
auto lit_end = p.atomic(p.literal("<|end|>"));
auto parser_until_end = p.until("<|end|>");
// prefix <- "<|begin|>assistant"?
auto parser_prefix = p.optional(lit_assistant_begin);
// reasoning <- "<|think|>" (!"<|end|>" .)*
auto parser_reasoning = p.rule("reasoning", lit_think + p.reasoning(parser_until_end));
@ -2601,7 +2598,7 @@ static common_chat_params common_chat_params_init_solar_open(const common_chat_t
// Response format parser
if (inputs.json_schema.is_object() && !inputs.json_schema.empty()) {
auto parser_response_format = lit_content + p.content(p.schema(p.json(), "response-format", inputs.json_schema));
return parser_prefix + p.choice({
return p.choice({
wrap_seq({parser_reasoning, parser_response_format}),
wrap_seq({parser_response_format})
});
@ -2648,7 +2645,7 @@ static common_chat_params common_chat_params_init_solar_open(const common_chat_t
if (min_calls == 1) {
// If required, then try any combination of the reasoning, content, and tool call
return parser_prefix + p.choice({
return p.choice({
wrap_seq({parser_reasoning, parser_content, parser_tool_calls}),
wrap_seq({parser_reasoning, parser_tool_calls}),
wrap_seq({parser_content, parser_tool_calls}),
@ -2656,12 +2653,12 @@ static common_chat_params common_chat_params_init_solar_open(const common_chat_t
});
}
return parser_prefix + wrap_choice({parser_reasoning, parser_content, parser_tool_calls});
return wrap_choice({parser_reasoning, parser_content, parser_tool_calls});
}
// Content only parser
include_grammar = false;
return parser_prefix + wrap_choice({parser_reasoning, parser_content});
return wrap_choice({parser_reasoning, parser_content});
});
data.parser = parser.save();

8
tests/test-chat.cpp
View File

@ -3789,14 +3789,14 @@ static void test_template_output_peg_parsers() {
// Test basic message and reasoning_effort = low
test_peg_parser(tmpls.get(), [&](auto & t) {
t.input = "<|begin|>assistant<|content|>Hello, world!\nWhat's up?";
t.input = "<|content|>Hello, world!\nWhat's up?";
t.params.chat_template_kwargs["reasoning_effort"] = "\"low\"";
t.expect = message_assist;
});
// Test tool call
test_peg_parser(tmpls.get(), [&](auto & t) {
t.input = "<|begin|>assistant<|tool_calls|>"
t.input = "<|tool_calls|>"
"<|tool_call:begin|>123456789"
"<|tool_call:name|>special_function"
"<|tool_call:args|>{\"arg1\":1}"
@ -3836,7 +3836,7 @@ static void test_template_output_peg_parsers() {
// Test tool call without reasoning and tool_choice = required
test_peg_parser(tmpls.get(), [&](auto & t) {
t.input = "<|begin|>assistant<|tool_calls|>"
t.input = "<|tool_calls|>"
"<|tool_call:begin|>0"
"<|tool_call:name|>special_function"
"<|tool_call:args|>{\"arg1\":1}"
@ -3890,7 +3890,7 @@ static void test_template_output_peg_parsers() {
// Test response format no reasoning
test_peg_parser(tmpls.get(), [&](auto & t) {
t.input = "<|begin|>assistant<|content|>"
t.input = "<|content|>"
R"({"amount": 123.45, "date": "2025-12-03"})";
t.params.chat_template_kwargs["reasoning_effort"] = "\"low\"";