Fix llama3 chat_template (extra <|eot_id|> on last turn) (#1635)

* Fix llama3 chat_template (the {{eos_token}} leads to an extra <|eot_id|> being added in the last turn). Output now matches official Llama 3 Instruct model

* add tests

* chore: lint

---------

Co-authored-by: Wing Lian <[email protected]>

src/axolotl/utils/chat_templates.py

@@ -24,7 +24,7 @@ def chat_templates(user_choice: str):
         "chatml": "{% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{% for message in messages %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}",
         "gemma": "{{ bos_token }}{% if messages[0]['role'] == 'system' %}{{ raise_exception('System role not supported') }}{% 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 (message['role'] == 'assistant') %}{% set role = 'model' %}{% else %}{% set role = message['role'] %}{% endif %}{{ '<start_of_turn>' + role + '\n' + message['content'] | trim + '<end_of_turn>\n' }}{% endfor %}{% if add_generation_prompt %}{{'<start_of_turn>model\n'}}{% endif %}",
         "cohere": "{{ bos_token }}{% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% elif false == true %}{% set loop_messages = messages %}{% set system_message = 'You are Command-R, a brilliant, sophisticated, AI-assistant trained to assist human users by providing thorough responses. You are trained by Cohere.' %}{% else %}{% set loop_messages = messages %}{% set system_message = false %}{% endif %}{% if system_message != false %}{{ '<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>' + system_message + '<|END_OF_TURN_TOKEN|>' }}{% endif %}{% for message in loop_messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ '<|START_OF_TURN_TOKEN|><|USER_TOKEN|>' + content.strip() + '<|END_OF_TURN_TOKEN|>' }}{% elif message['role'] == 'assistant' %}{{ '<|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>'  + content.strip() + '<|END_OF_TURN_TOKEN|>' }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>' }}{% endif %}",
-        "llama3": "{% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% else %}{{ eos_token }}{% endif %}",
+        "llama3": "{% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}",
     }
 
     if user_choice in templates:

tests/prompt_strategies/test_chat_templates.py

@@ -0,0 +1,85 @@
+"""
+tests for chat_template prompt strategy
+"""
+import unittest
+
+import pytest
+from datasets import Dataset
+from transformers import AutoTokenizer
+
+from axolotl.prompt_strategies.chat_template import (
+    ChatTemplatePrompter,
+    ChatTemplateStrategy,
+)
+from axolotl.utils.chat_templates import chat_templates
+
+
+@pytest.fixture(name="sharegpt_dataset")
+def fixture_sharegpt_dataset():
+    # pylint: disable=duplicate-code
+    return Dataset.from_list(
+        [
+            {
+                "conversations": [
+                    {
+                        "from": "human",
+                        "value": "hello",
+                    },
+                    {
+                        "from": "gpt",
+                        "value": "hello",
+                    },
+                    {
+                        "from": "human",
+                        "value": "goodbye",
+                    },
+                    {
+                        "from": "gpt",
+                        "value": "goodbye",
+                    },
+                ]
+            }
+        ]
+    )
+
+
+@pytest.fixture(name="llama3_tokenizer")
+def fixture_llama3_tokenizer():
+    tokenizer = AutoTokenizer.from_pretrained("NousResearch/Meta-Llama-3-8B")
+    tokenizer.eos_token = "<|eot_id|>"
+
+    return tokenizer
+
+
+class TestSharegptChatTemplateLlama3:
+    """
+    Test class for ShareGPT style datasets with llama-3 prompts using the chat_template strategy.
+    """
+
+    def test_llama3(self, llama3_tokenizer, sharegpt_dataset):
+        # pylint: disable=duplicate-code
+        strategy = ChatTemplateStrategy(
+            ChatTemplatePrompter(llama3_tokenizer, chat_templates("llama3")),
+            llama3_tokenizer,
+            False,
+            512,
+        )
+        res = strategy.tokenize_prompt(sharegpt_dataset[0])
+        input_ids = res["input_ids"]
+        # fmt: off
+        assert input_ids == [
+            128000,  # bos
+            128006, 882, 128007,  # user header
+            271, 15339, 128009,  # user prompt eot
+            128006, 78191, 128007,  # assistant header
+            271, 15339, 128009,   # assistant response eot
+            128006, 882, 128007,
+            271, 19045, 29474, 128009,
+            128006, 78191, 128007,
+            271, 19045, 29474, 128009,
+        ]
+        # fmt: on
+
+
+if __name__ == "__main__":
+    unittest.main()