I have an issue when loading a Qwen chat model with LlamaCpp

[Abdechakour-Mc]

I'm trying to load this model codeqwen-1_5-7b-chat-q8_0.gguf using models.LlamaCpp but it throughs a User warning as shown below, I tried to load the model using the native Llama.cpp and it worked just fine!

here is the used code:

from guidance import models, gen

path = "codeqwen-1_5-7b-chat-q8_0.gguf"

# Initialize the model
llm = models.LlamaCpp(path, n_ctx=4096, n_gpu_layers=-1)

UserWarning: Chat template {% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system
You are a helpful assistant<|im_end|>
' }}{% endif %}{{'<|im_start|>' + message['role'] + '
' + message['content'] + '<|im_end|>' + '
'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant
' }}{% endif %} was unable to be loaded directly into guidance.
Defaulting to the ChatML format which may not be optimal for the selected model.
For best results, create and pass in a guidance.ChatTemplate subclass for your model.
warnings.warn(f"""Chat template {chat_template} was unable to be loaded directly into guidance.

[yileitu]

Same here for llama 3.1. Any solutions?

[KJKR73]

import time
import guidance
from guidance.models import Transformers
from guidance import gen, user, assistant, select

import torch
from datasets import load_dataset
from guidance.chat import ChatTemplate
from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig

bnb_config = BitsAndBytesConfig(
    load_in_4bit=True,
    bnb_4bit_use_double_quant=True,
    bnb_4bit_quant_type="nf4",
    bnb_4bit_compute_dtype=torch.bfloat16
)

# Load the tokenizer
model_id = "meta-llama/Meta-Llama-3.1-8B-Instruct"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, quantization_config=bnb_config, device_map={"": 0})

# The custom chat template
llama31_template = tokenizer.chat_template
class Llama31ChatTemplate(ChatTemplate):
    template_str = llama31_template
    def get_role_start(self, role_name):
        if role_name == "system":
            return "<|begin_of_text|><|start_header_id|>system<|end_header_id|>\n\n"
        elif role_name == "user":
            return "<|start_header_id|>user<|end_header_id|>\n\n"
        elif role_name == "assistant":
            return "<|start_header_id|>assistant<|end_header_id|>\n\n"
        else:
            raise UnsupportedRoleException(role_name, self)

    def get_role_end(self, role_name=None):
        return "<|eot_id|>"

# Load the model with the chat template
model = Transformers(model=model, tokenizer=tokenizer, chat_template=Llama31ChatTemplate)

Here is a working fix for the Transformer variants. You can load any model and its chat template with this. Hope this helps :)

[prnvbn]

Is there anyway to achieve this with the LLamaCpp model? I tried the following where I got the chat_template from huggingface - https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct/commit/8c22764a7e3675c50d4c7c9a4edb474456022b16

    llama31_template = "{{- bos_token }}\n{%- if custom_tools is defined %}\n    {%- set tools = custom_tools %}\n{%- endif %}\n{%- if not tools_in_user_message is defined %}\n    {%- set tools_in_user_message = true %}\n{%- endif %}\n{%- if not date_string is defined %}\n    {%- set date_string = \"26 Jul 2024\" %}\n{%- endif %}\n{%- if not tools is defined %}\n    {%- set tools = none %}\n{%- endif %}\n\n{#- This block extracts the system message, so we can slot it into the right place. #}\n{%- if messages[0]['role'] == 'system' %}\n    {%- set system_message = messages[0]['content']|trim %}\n    {%- set messages = messages[1:] %}\n{%- else %}\n    {%- set system_message = \"\" %}\n{%- endif %}\n\n{#- System message + builtin tools #}\n{{- \"<|start_header_id|>system<|end_header_id|>\\n\\n\" }}\n{%- if builtin_tools is defined or tools is not none %}\n    {{- \"Environment: ipython\\n\" }}\n{%- endif %}\n{%- if builtin_tools is defined %}\n    {{- \"Tools: \" + builtin_tools | reject('equalto', 'code_interpreter') | join(\", \") + \"\\n\\n\"}}\n{%- endif %}\n{{- \"Cutting Knowledge Date: December 2023\\n\" }}\n{{- \"Today Date: \" + date_string + \"\\n\\n\" }}\n{%- if tools is not none and not tools_in_user_message %}\n    {{- \"You have access to the following functions. To call a function, please respond with JSON for a function call.\" }}\n    {{- 'Respond in the format {\"name\": function name, \"parameters\": dictionary of argument name and its value}.' }}\n    {{- \"Do not use variables.\\n\\n\" }}\n    {%- for t in tools %}\n        {{- t | tojson(indent=4) }}\n        {{- \"\\n\\n\" }}\n    {%- endfor %}\n{%- endif %}\n{{- system_message }}\n{{- \"<|eot_id|>\" }}\n\n{#- Custom tools are passed in a user message with some extra guidance #}\n{%- if tools_in_user_message and not tools is none %}\n    {#- Extract the first user message so we can plug it in here #}\n    {%- if messages | length != 0 %}\n        {%- set first_user_message = messages[0]['content']|trim %}\n        {%- set messages = messages[1:] %}\n    {%- else %}\n        {{- raise_exception(\"Cannot put tools in the first user message when there's no first user message!\") }}\n{%- endif %}\n    {{- '<|start_header_id|>user<|end_header_id|>\\n\\n' -}}\n    {{- \"Given the following functions, please respond with a JSON for a function call \" }}\n    {{- \"with its proper arguments that best answers the given prompt.\\n\\n\" }}\n    {{- 'Respond in the format {\"name\": function name, \"parameters\": dictionary of argument name and its value}.' }}\n    {{- \"Do not use variables.\\n\\n\" }}\n    {%- for t in tools %}\n        {{- t | tojson(indent=4) }}\n        {{- \"\\n\\n\" }}\n    {%- endfor %}\n    {{- first_user_message + \"<|eot_id|>\"}}\n{%- endif %}\n\n{%- for message in messages %}\n    {%- if not (message.role == 'ipython' or message.role == 'tool' or 'tool_calls' in message) %}\n        {{- '<|start_header_id|>' + message['role'] + '<|end_header_id|>\\n\\n'+ message['content'] | trim + '<|eot_id|>' }}\n    {%- elif 'tool_calls' in message %}\n        {%- if not message.tool_calls|length == 1 %}\n            {{- raise_exception(\"This model only supports single tool-calls at once!\") }}\n        {%- endif %}\n        {%- set tool_call = message.tool_calls[0].function %}\n        {%- if builtin_tools is defined and tool_call.name in builtin_tools %}\n            {{- '<|start_header_id|>assistant<|end_header_id|>\\n\\n' -}}\n            {{- \"<|python_tag|>\" + tool_call.name + \".call(\" }}\n            {%- for arg_name, arg_val in tool_call.arguments | items %}\n                {{- arg_name + '=\"' + arg_val + '\"' }}\n                {%- if not loop.last %}\n                    {{- \", \" }}\n                {%- endif %}\n                {%- endfor %}\n            {{- \")\" }}\n        {%- else  %}\n            {{- '<|start_header_id|>assistant<|end_header_id|>\\n\\n' -}}\n            {{- '{\"name\": \"' + tool_call.name + '\", ' }}\n            {{- '\"parameters\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- \"}\" }}\n        {%- endif %}\n        {%- if builtin_tools is defined %}\n            {#- This means we're in ipython mode #}\n            {{- \"<|eom_id|>\" }}\n        {%- else %}\n            {{- \"<|eot_id|>\" }}\n        {%- endif %}\n    {%- elif message.role == \"tool\" or message.role == \"ipython\" %}\n        {{- \"<|start_header_id|>ipython<|end_header_id|>\\n\\n\" }}\n        {%- if message.content is mapping or message.content is iterable %}\n            {{- message.content | tojson }}\n        {%- else %}\n            {{- message.content }}\n        {%- endif %}\n        {{- \"<|eot_id|>\" }}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|start_header_id|>assistant<|end_header_id|>\\n\\n' }}\n{%- endif %}\n",
    class Llama31ChatTemplate(guidance.chat.ChatTemplate):
        template_str = llama31_template
        def get_role_start(self, role_name):
            if role_name == "system":
                return "<|begin_of_text|><|start_header_id|>system<|end_header_id|>\n\n"
            elif role_name == "user":
                return "<|start_header_id|>user<|end_header_id|>\n\n"
            elif role_name == "assistant":
                return "<|start_header_id|>assistant<|end_header_id|>\n\n"
            else:
                raise UnsupportedRoleException(role_name, self)

        def get_role_end(self, role_name=None):
            return "<|eot_id|>"


    # when using llama.cppm ensure that the llama.cpp was installed with BLAS ENABLED
    # an alternative to using llama.cpp is using huggingface transformers library
    llama2 = models.LlamaCpp(
        model_path,
        n_gpu_layers=NUM_LAYERS,
        n_batch=512,
        n_ctx=N_CONTEXT,
        echo=False,
        temperature=0.5,
        chat_template=Llama31ChatTemplate(),
        verbose=True, 
    )

[develCuy]

guidance has a nice set of chat templates already

guidance/guidance/chat.py

Line 334 in 87bc0a8

class Llama3dot2ChatTemplate(ChatTemplate):

It was doing a great job previously but latest versions sometimes can't find a match. Yet, you can pass a template manually. Like this:

from guidance.chat import Llama3dot2ChatTemplate;

Then, pass that class as chat_template param:

llama32 = models.LlamaCpp(
  ...
  chat_template=Llama3dot2ChatTemplate(),
  ...
)