add chat template directly

Signed-off-by: jerryzhuang <[email protected]>

docker/presets/models/tfs/Dockerfile

@@ -24,13 +24,7 @@ COPY kaito/presets/inference/${MODEL_TYPE}/inference_api.py \
 COPY kaito/presets/inference/vllm/inference_api.py /workspace/vllm/inference_api.py
 
 # Chat template
-RUN apt update && apt install -y git && \
-    rm /var/lib/apt/lists/* -r
-RUN git clone https://github.com/chujiezheng/chat_templates /tmp/chat_templates && \
-    cd /tmp/chat_templates && \
-    git reset --hard 670a2eb && \
-    cp -r ./chat_templates/ /workspace/ && \
-    rm -rf /tmp/chat_templates
+ADD kaito/presets/inference/chat_templates /workspace/chat_templates
 
 # Model weights
 COPY ${WEIGHTS_PATH} /workspace/weights

presets/inference/chat_templates/falcon-instruct.jinja

@@ -0,0 +1,20 @@
+{% if messages[0]['role'] == 'system' %}
+    {% set system_message = messages[0]['content'] %}
+    {% set messages = messages[1:] %}
+{% else %}
+    {% set system_message = '' %}
+{% endif %}
+
+{{ system_message | trim }}
+{% for message in messages %}
+    {% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}
+        {{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}
+    {% endif %}
+
+    {% set content = message['content'].replace('\r\n', '\n').replace('\n\n', '\n') %}
+    {{ '\n\n' + message['role'] | capitalize + ': ' + content | trim }}
+{% endfor %}
+
+{% if add_generation_prompt %}
+    {{ '\n\nAssistant:' }}
+{% endif %}

presets/inference/chat_templates/llama-2-chat.jinja

@@ -0,0 +1,24 @@
+{% if messages[0]['role'] == 'system' %}
+    {% set system_message = '<<SYS>>\n' + messages[0]['content'] | trim + '\n<</SYS>>\n\n' %}
+    {% set messages = messages[1:] %}
+{% else %}
+    {% set system_message = '' %}
+{% endif %}
+
+{% for message in messages %}
+    {% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}
+        {{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}
+    {% endif %}
+
+    {% if loop.index0 == 0 %}
+        {% set content = system_message + message['content'] %}
+    {% else %}
+        {% set content = message['content'] %}
+    {% endif %}
+
+    {% if message['role'] == 'user' %}
+        {{ bos_token + '[INST] ' + content | trim + ' [/INST]' }}
+    {% elif message['role'] == 'assistant' %}
+        {{ ' ' + content | trim + ' ' + eos_token }}
+    {% endif %}
+{% endfor %}

presets/inference/chat_templates/llama-3-instruct.jinja

@@ -0,0 +1,18 @@
+{% if messages[0]['role'] == 'system' %}
+    {% set offset = 1 %}
+{% else %}
+    {% set offset = 0 %}
+{% endif %}
+
+{{ bos_token }}
+{% for message in messages %}
+    {% if (message['role'] == 'user') != (loop.index0 % 2 == offset) %}
+        {{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}
+    {% endif %}
+
+    {{ '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n' + message['content'] | trim + '<|eot_id|>' }}
+{% endfor %}
+
+{% if add_generation_prompt %}
+    {{ '<|start_header_id|>' + 'assistant' + '<|end_header_id|>\n\n' }}
+{% endif %}

presets/inference/chat_templates/mistral-instruct.jinja

@@ -0,0 +1,19 @@
+{% if messages[0]['role'] == 'system' %}
+    {% set system_message = messages[0]['content'] | trim + '\n\n' %}
+    {% set messages = messages[1:] %}
+{% else %}
+    {% set system_message = '' %}
+{% endif %}
+
+{{ bos_token + system_message}}
+{% for message in messages %}
+    {% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}
+        {{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}
+    {% endif %}
+
+    {% if message['role'] == 'user' %}
+        {{ '[INST] ' + message['content'] | trim + ' [/INST]' }}
+    {% elif message['role'] == 'assistant' %}
+        {{ ' ' + message['content'] | trim + eos_token }}
+    {% endif %}
+{% endfor %}

presets/inference/chat_templates/phi-3-small.jinja

@@ -0,0 +1,18 @@
+{% if messages[0]['role'] == 'system' %}
+    {% set offset = 1 %}
+{% else %}
+    {% set offset = 0 %}
+{% endif %}
+
+{{ bos_token }}
+{% for message in messages %}
+    {% if (message['role'] == 'user') != (loop.index0 % 2 == offset) %}
+        {{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}
+    {% endif %}
+
+    {{ '<|' + message['role'] + '|>\n' + message['content'] | trim + '<|end|>' + '\n' }}
+{% endfor %}
+
+{% if add_generation_prompt %}
+    {{ '<|assistant|>\n' }}
+{% endif %}

presets/inference/chat_templates/phi-3.jinja

@@ -0,0 +1,17 @@
+{% if messages[0]['role'] == 'system' %}
+    {% set offset = 1 %}
+{% else %}
+    {% set offset = 0 %}
+{% endif %}
+
+{% for message in messages %}
+    {% if (message['role'] == 'user') != (loop.index0 % 2 == offset) %}
+        {{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}
+    {% endif %}
+
+    {{ '<|' + message['role'] + '|>\n' + message['content'] | trim + '<|end|>' + '\n' }}
+{% endfor %}
+
+{% if add_generation_prompt %}
+    {{ '<|assistant|>\n' }}
+{% endif %}