ensure deepseek r1 thoughts parsed even w/o tool calls

This commit is contained in:
ochafik 2025-02-04 04:48:08 +00:00
parent b6e14a4101
commit 1f5ec59809
2 changed files with 91 additions and 49 deletions

View file

@ -565,39 +565,41 @@ static common_chat_msg common_chat_parse_llama_3_1(const std::string & input, bo
static common_chat_params common_chat_params_init_deepseek_r1(const common_chat_template & tmpl, const struct common_chat_inputs & inputs) { static common_chat_params common_chat_params_init_deepseek_r1(const common_chat_template & tmpl, const struct common_chat_inputs & inputs) {
common_chat_params data; common_chat_params data;
data.grammar_lazy = inputs.tool_choice != "required"; if (!inputs.tools.is_null() && !inputs.tools.empty()) {
data.grammar = build_grammar([&](const common_grammar_builder & builder) { data.grammar_lazy = inputs.tool_choice != "required";
std::vector<std::string> tool_rules; data.grammar = build_grammar([&](const common_grammar_builder & builder) {
foreach_function(inputs.tools, [&](const json & tool) { std::vector<std::string> tool_rules;
const auto & function = tool["function"]; foreach_function(inputs.tools, [&](const json & tool) {
std::string name = function["name"]; const auto & function = tool["function"];
auto parameters = function["parameters"]; std::string name = function["name"];
auto args_rule = builder.add_schema(name + "-args", parameters); auto parameters = function["parameters"];
tool_rules.push_back(builder.add_rule(name + "-call", auto args_rule = builder.add_schema(name + "-args", parameters);
"\"<tool▁call▁begin>function<tool▁sep>" + name + "\\n" tool_rules.push_back(builder.add_rule(name + "-call",
"```json\\n\" " + args_rule + " \"```<tool▁call▁end>\"")); "\"<tool▁call▁begin>function<tool▁sep>" + name + "\\n"
}); "```json\\n\" " + args_rule + " \"```<tool▁call▁end>\""));
// Distill Qwen 7B & 32B models seem confused re/ syntax of their tool call opening tag, });
// so we accept common variants (then it's all constrained) // Distill Qwen 7B & 32B models seem confused re/ syntax of their tool call opening tag,
builder.add_rule("root", // so we accept common variants (then it's all constrained)
"( \"<tool▁calls▁begin>\" | \"<tool_calls_begin>\" | \"<tool calls begin>\" | \"<tool\\\\_calls\\\\_begin>\" ) " builder.add_rule("root",
"(" +string_join(tool_rules, " | ") + ")" + (inputs.parallel_tool_calls ? "*" : "") + " " "( \"<tool▁calls▁begin>\" | \"<tool_calls_begin>\" | \"<tool calls begin>\" | \"<tool\\\\_calls\\\\_begin>\" ) "
"\"<tool▁calls▁end>\"" "(" +string_join(tool_rules, " | ") + ")" + (inputs.parallel_tool_calls ? "*" : "") + " "
" space"); "\"<tool▁calls▁end>\""
data.grammar_triggers.push_back({"<tool▁calls▁begin>", /* .at_start = */ false}); " space");
data.grammar_triggers.push_back({"<tool_calls_begin>", /* .at_start = */ false}); data.grammar_triggers.push_back({"<tool▁calls▁begin>", /* .at_start = */ false});
data.grammar_triggers.push_back({"<tool calls begin>", /* .at_start = */ false}); data.grammar_triggers.push_back({"<tool_calls_begin>", /* .at_start = */ false});
data.grammar_triggers.push_back({"<tool\\_calls\\_begin>", /* .at_start = */ false}); data.grammar_triggers.push_back({"<tool calls begin>", /* .at_start = */ false});
data.grammar_triggers.push_back({"<tool▁call▁begin>", /* .at_start = */ false}); data.grammar_triggers.push_back({"<tool\\_calls\\_begin>", /* .at_start = */ false});
data.preserved_tokens = { data.grammar_triggers.push_back({"<tool▁call▁begin>", /* .at_start = */ false});
"<think>", data.preserved_tokens = {
"</think>", "<think>",
"<tool▁sep>", "</think>",
"<tool▁calls▁end", "<tool▁sep>",
"<tool▁call▁begin>", "<tool▁calls▁end",
"<tool▁call▁end>", "<tool▁call▁begin>",
}; "<tool▁call▁end>",
}, grammar_options); };
}, grammar_options);
}
auto prompt = apply(tmpl, inputs.messages, inputs.tools.empty() ? json() : inputs.tools, inputs.add_generation_prompt); auto prompt = apply(tmpl, inputs.messages, inputs.tools.empty() ? json() : inputs.tools, inputs.add_generation_prompt);
// Hacks to fix the official (broken) prompt. // Hacks to fix the official (broken) prompt.
@ -638,7 +640,7 @@ static common_chat_msg common_chat_parse_deepseek_r1(const std::string & input)
auto msg2 = parse_json_tool_calls(tool_calls, std::nullopt, function_regex, close_regex); auto msg2 = parse_json_tool_calls(tool_calls, std::nullopt, function_regex, close_regex);
msg.tool_calls = std::move(msg2.tool_calls); msg.tool_calls = std::move(msg2.tool_calls);
} else { } else {
msg.content = rest; msg.content = std::string(rest.begin() + rest.find_first_not_of(" \r\n"), rest.end());
} }
} else { } else {
msg.content = input; msg.content = input;
@ -970,6 +972,9 @@ common_chat_params common_chat_params_init(const common_chat_template & tmpl, co
// Firefunction v2 requires datetime and functions in the context, even w/o tools. // Firefunction v2 requires datetime and functions in the context, even w/o tools.
return common_chat_params_init_firefunction_v2(tmpl, inputs); return common_chat_params_init_firefunction_v2(tmpl, inputs);
} }
if (src.find("<tool▁calls▁begin>") != std::string::npos) {
return common_chat_params_init_deepseek_r1(tmpl, inputs);
}
if (!has_tools) { if (!has_tools) {
return common_chat_params_init_without_tools(tmpl, inputs); return common_chat_params_init_without_tools(tmpl, inputs);
@ -986,9 +991,6 @@ common_chat_params common_chat_params_init(const common_chat_template & tmpl, co
auto allow_python_tag_builtin_tools = src.find("<|python_tag|>") != std::string::npos; auto allow_python_tag_builtin_tools = src.find("<|python_tag|>") != std::string::npos;
return common_chat_params_init_llama_3_1_tool_calls(tmpl, inputs, allow_python_tag_builtin_tools); return common_chat_params_init_llama_3_1_tool_calls(tmpl, inputs, allow_python_tag_builtin_tools);
} }
if (src.find("<tool▁calls▁begin>") != std::string::npos) {
return common_chat_params_init_deepseek_r1(tmpl, inputs);
}
if (src.find("[TOOL_CALLS]") != std::string::npos) { if (src.find("[TOOL_CALLS]") != std::string::npos) {
return common_chat_params_init_mistral_nemo(tmpl, inputs); return common_chat_params_init_mistral_nemo(tmpl, inputs);
} }

View file

@ -345,20 +345,20 @@ def test_weather_tool_call(hf_repo: str, template_override: str | Tuple[str, str
@pytest.mark.slow @pytest.mark.slow
@pytest.mark.parametrize("result_override,n_predict,hf_repo,template_override", [ @pytest.mark.parametrize("result_override,n_predict,hf_repo,template_override", [
(None, 128, "bartowski/Phi-3.5-mini-instruct-GGUF:Q4_K_M", "chatml"), (None, 128, "bartowski/Phi-3.5-mini-instruct-GGUF:Q4_K_M", "chatml"),
(None, 128, "bartowski/Qwen2.5-7B-Instruct-GGUF:Q4_K_M", None), (None, 128, "bartowski/Qwen2.5-7B-Instruct-GGUF:Q4_K_M", None),
(None, 128, "bartowski/Qwen2.5-7B-Instruct-GGUF:Q4_K_M", "chatml"), (None, 128, "bartowski/Qwen2.5-7B-Instruct-GGUF:Q4_K_M", "chatml"),
(None, 128, "bartowski/Hermes-2-Pro-Llama-3-8B-GGUF:Q4_K_M", ("NousResearch/Hermes-2-Pro-Llama-3-8B", "tool_use")), (None, 128, "bartowski/Hermes-2-Pro-Llama-3-8B-GGUF:Q4_K_M", ("NousResearch/Hermes-2-Pro-Llama-3-8B", "tool_use")),
(None, 128, "bartowski/Hermes-3-Llama-3.1-8B-GGUF:Q4_K_M", ("NousResearch/Hermes-3-Llama-3.1-8B", "tool_use")), (None, 128, "bartowski/Hermes-3-Llama-3.1-8B-GGUF:Q4_K_M", ("NousResearch/Hermes-3-Llama-3.1-8B", "tool_use")),
(None, 128, "bartowski/functionary-small-v3.2-GGUF:Q8_0", ("meetkai/functionary-medium-v3.2", None)), (None, 128, "bartowski/functionary-small-v3.2-GGUF:Q8_0", ("meetkai/functionary-medium-v3.2", None)),
(None, 128, "bartowski/Meta-Llama-3.1-8B-Instruct-GGUF:Q4_K_M", None), (None, 128, "bartowski/Meta-Llama-3.1-8B-Instruct-GGUF:Q4_K_M", None),
(None, 128, "bartowski/Mistral-Nemo-Instruct-2407-GGUF:Q4_K_M", None), (None, 128, "bartowski/Mistral-Nemo-Instruct-2407-GGUF:Q4_K_M", None),
("^> 0.56$", 128, "bartowski/Mistral-Nemo-Instruct-2407-GGUF:Q4_K_M", "chatml"), ("^> 0.56$", 128, "bartowski/Mistral-Nemo-Instruct-2407-GGUF:Q4_K_M", "chatml"),
# TODO: fix these (wrong results, either didn't respect decimal instruction or got wrong value) # TODO: fix these (wrong results, either didn't respect decimal instruction or got wrong value)
("[\\s\\S\\r\\n]*?\\b0\\.55644242476$", 128, "bartowski/Phi-3.5-mini-instruct-GGUF:Q4_K_M", None), ("[\\s\\S\\r\\n]*?\\b0\\.55644242476$", 128, "bartowski/Phi-3.5-mini-instruct-GGUF:Q4_K_M", None),
("[\\s\\S\\r\\n]*?which equals 0\\.5\\.", 8192, "bartowski/DeepSeek-R1-Distill-Qwen-7B-GGUF:Q4_K_M", None), ("[\\s\\S\\r\\n]*?which equals 0\\.5\\.", 8192, "bartowski/DeepSeek-R1-Distill-Qwen-7B-GGUF:Q4_K_M", None),
("**Answer:** 0\\.25\\b", 8192, "bartowski/DeepSeek-R1-Distill-Qwen-7B-GGUF:Q4_K_M", ("llama-cpp-deepseek-r1", None)), ("[\\s\\S\\r\\n]*?\\*\\*Answer:\\*\\* 0\\.25\\b", 8192, "bartowski/DeepSeek-R1-Distill-Qwen-7B-GGUF:Q4_K_M", ("llama-cpp-deepseek-r1", None)),
]) ])
def test_calc_result(result_override: str | None, n_predict: int, hf_repo: str, template_override: str | Tuple[str, str | None] | None): def test_calc_result(result_override: str | None, n_predict: int, hf_repo: str, template_override: str | Tuple[str, str | None] | None):
global server global server
@ -435,6 +435,46 @@ def test_calc_result(result_override: str | None, n_predict: int, hf_repo: str,
f'Expected something like "The y coordinate is 0.56.", got {content}' f'Expected something like "The y coordinate is 0.56.", got {content}'
@pytest.mark.slow
@pytest.mark.parametrize("n_predict,expect_content,expect_thoughts,hf_repo,template_override", [
(128, "^The sum of 102 and 7 is 109.*", None, "bartowski/Phi-3.5-mini-instruct-GGUF:Q4_K_M", None),
(1024, "To find the sum of.*", "I need to calculate the sum of 102 and 7.*", "bartowski/DeepSeek-R1-Distill-Qwen-7B-GGUF:Q4_K_M", None),
(1024, "To find the sum of.*", "First, I need to add the tens place.*", "bartowski/DeepSeek-R1-Distill-Qwen-7B-GGUF:Q4_K_M", ("llama-cpp-deepseek-r1", None)),
])
def test_thoughts(n_predict: int, expect_content: str | None, expect_thoughts: str | None, hf_repo: str, template_override: str | Tuple[str, str | None] | None):
global server
server.n_slots = 1
server.jinja = True
server.n_ctx = 8192 * 2
server.n_predict = n_predict
server.model_hf_repo = hf_repo
server.model_hf_file = None
if isinstance(template_override, tuple):
(template_hf_repo, template_variant) = template_override
server.chat_template_file = f"../../../models/templates/{template_hf_repo.replace('/', '-') + ('-' + template_variant if template_variant else '')}.jinja"
assert os.path.exists(server.chat_template_file), f"Template file {server.chat_template_file} does not exist. Run `python scripts/get_chat_template.py {template_hf_repo} {template_variant} > {server.chat_template_file}` to download the template."
elif isinstance(template_override, str):
server.chat_template = template_override
server.start(timeout_seconds=TIMEOUT_SERVER_START)
res = server.make_request("POST", "/chat/completions", data={
"max_tokens": n_predict,
"messages": [
{"role": "user", "content": "What's the sum of 102 and 7?"},
]
}, timeout=TIMEOUT_HTTP_REQUEST)
assert res.status_code == 200, f"Expected status code 200, got {res.status_code}"
choice = res.body["choices"][0]
assert choice["message"].get("tool_calls") is None, f'Expected no tool call in {choice["message"]}'
content = choice["message"].get("content")
if expect_content is not None:
assert re.match(expect_content, content), f'Expected {expect_content}, got {content}'
thoughts = choice["message"].get("thoughts")
if expect_thoughts is not None:
assert re.match(expect_thoughts, thoughts), f'Expected {expect_thoughts}, got {thoughts}'
@pytest.mark.slow @pytest.mark.slow
@pytest.mark.parametrize("expected_arguments_override,hf_repo,template_override", [ @pytest.mark.parametrize("expected_arguments_override,hf_repo,template_override", [
(None, "bartowski/DeepSeek-R1-Distill-Qwen-7B-GGUF:Q4_K_M", None), (None, "bartowski/DeepSeek-R1-Distill-Qwen-7B-GGUF:Q4_K_M", None),