vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

simpler hacky fixes for original broken template (+ fix minja example syntax polyfill)

Browse source
This commit is contained in:
Olivier Chafik 2025-02-03 20:34:44 +00:00
parent c6214ee9d6
commit 1c302e18ba
2 changed files with 24 additions and 10 deletions
Download patch file Download diff file Expand all files Collapse all files

24
common/chat.cpp
View file

@ -569,11 +569,25 @@ static common_chat_params common_chat_params_init_deepseek_r1(const common_chat_
}; };
}, grammar_options); }, grammar_options);
auto prompt = tmpl.apply(inputs.messages, inputs.tools.empty() ? json() : inputs.tools, inputs.add_generation_prompt); auto prompt = tmpl.apply(inputs.messages, inputs.tools.empty() ? json() : inputs.tools, inputs.add_generation_prompt);
// Hack to fix the official prompt, which leaves the chat dangling after tool results.
if (string_ends_with(prompt, "<｜tool▁outputs▁end｜>")) { // Hacks to fix the official (broken) prompt.
prompt += "<｜end▁of▁sentence｜>"; // It is advisable to use --chat-template-file models/templates/llama-cpp-deepseek-r1.jinja instead,
if (inputs.add_generation_prompt) { // until the official template is fixed.
prompt += "<｜Assistant｜>"; if (tmpl.source().find("{% if ns.is_tool %}{{'<｜tool▁outputs▁end｜>'}}") != std::string::npos) {
// Don't leave the chat dangling after tool results
if (string_ends_with(prompt, "<｜tool▁outputs▁end｜>")) {
prompt += "<｜end▁of▁sentence｜>";
if (inputs.add_generation_prompt) {
prompt += "<｜Assistant｜>";
}
}
// Fix up tool call delta example added by Minja
std::string marker = "<｜tool▁call▁end｜>\n";
auto pos = prompt.rfind(marker);
if (pos != std::string::npos) {
prompt.insert(pos + marker.size() - 1, "<｜tool▁calls▁end｜>");
} else {
LOG_WRN("Failed to find expected broken tool call example marker in prompt\n");
} }
} }
data.prompt = prompt; data.prompt = prompt;

10
models/templates/llama-cpp-deepseek-r1.jinja
View file

@ -1,7 +1,7 @@
{%- if not add_generation_prompt is defined -%} {%- if not add_generation_prompt is defined -%}
{%- set add_generation_prompt = false -%} {%- set add_generation_prompt = false -%}
{%- endif -%} {%- endif -%}
{%- set ns = namespace(is_first=false, is_tool=false, is_output_first=true, system_prompt='') -%} {%- set ns = namespace(is_first=false, is_tool_outputs=false, is_output_first=true, system_prompt='') -%}
{%- for message in messages -%} {%- for message in messages -%}
{%- if message['role'] == 'system' -%} {%- if message['role'] == 'system' -%}
{%- set ns.system_prompt = message['content'] -%} {%- set ns.system_prompt = message['content'] -%}
@ -25,9 +25,9 @@ Example function tool call syntax:
{% endif -%} {% endif -%}
{{ns.system_prompt}} {{ns.system_prompt}}
{%- macro flush_tool_outputs() -%} {%- macro flush_tool_outputs() -%}
{%- if ns.is_tool -%} {%- if ns.is_tool_outputs -%}
{{- '<｜tool▁outputs▁end｜><｜end▁of▁sentence｜>' -}} {{- '<｜tool▁outputs▁end｜><｜end▁of▁sentence｜>' -}}
{%- set ns.is_tool = false -%} {%- set ns.is_tool_outputs = false -%}
{%- endif -%} {%- endif -%}
{%- endmacro -%} {%- endmacro -%}
{{- flush_tool_outputs() -}} {{- flush_tool_outputs() -}}
@ -62,7 +62,7 @@ Example function tool call syntax:
{{- '<｜Assistant｜>' + content + '<｜end▁of▁sentence｜>'}} {{- '<｜Assistant｜>' + content + '<｜end▁of▁sentence｜>'}}
{%- endif -%} {%- endif -%}
{%- if message['role'] == 'tool' -%} {%- if message['role'] == 'tool' -%}
{%- set ns.is_tool = true -%} {%- set ns.is_tool_outputs = true -%}
{%- if ns.is_output_first -%} {%- if ns.is_output_first -%}
{{- '<｜tool▁outputs▁begin｜>' -}} {{- '<｜tool▁outputs▁begin｜>' -}}
{%- set ns.is_output_first = false -%} {%- set ns.is_output_first = false -%}
@ -71,6 +71,6 @@ Example function tool call syntax:
{%- endif -%} {%- endif -%}
{%- endfor -%} {%- endfor -%}
{{- flush_tool_outputs() -}} {{- flush_tool_outputs() -}}
{%- if add_generation_prompt and not ns.is_tool -%} {%- if add_generation_prompt and not ns.is_tool_outputs -%}
{{- '<｜Assistant｜>' -}} {{- '<｜Assistant｜>' -}}
{%- endif -%} {%- endif -%}