Why am I generating a lot of endoftext here

[ZhengChenYang]

Why am I generating a lot of endoftext here

[ZhengChenYang]

like this：
156 - 147 + 9 to get the final answer 12.<|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|>

[patrickstar-sjh]

I also have the same problem

(main_task pid=2436808) --------------------------------
(main_task pid=2436808) Target: 16 | Numbers: [92 48 11 49]
(main_task pid=2436808) Extracted equation: None
(main_task pid=2436808) Solution string: A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer.
(main_task pid=2436808) User: Using the numbers [92, 48, 11, 49], create an equation that equals 16. You can use basic arithmetic operations (+, -, *, /) and each number can only be used once. Show your work in tags. And return the final answer in tags, for example (1 + 2) / 3 .
(main_task pid=2436808) Assistant: Let me solve this step by step.
(main_task pid=2436808) </ </</ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </ </</th
(main_task pid=2436808) No equation found
(main_task pid=2436808) epoch 0, step 57
(main_task pid=2436808) --------------------------------
(main_task pid=2436808) Target: 18 | Numbers: [86 34 70]
(main_task pid=2436808) Extracted equation:
(main_task pid=2436808) Solution string: A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer.
(main_task pid=2436808) User: Using the numbers [86, 34, 70], create an equation that equals 18. You can use basic arithmetic operations (+, -, *, /) and each number can only be used once. Show your work in tags. And return the final answer in tags, for example (1 + 2) / 3 .
(main_task pid=2436808) Assistant: Let me solve this step by step.
(main_task pid=2436808) using basic arithmetic operations and each number can only be used once. The final answer is in </
(main_task pid=2436808) Invalid equation
(main_task pid=2436808) epoch 0, step 58
(main_task pid=2436808) epoch 0, step 59
(main_task pid=2436808) --------------------------------
(main_task pid=2436808) Target: 86 | Numbers: [41 46 89 15]
(main_task pid=2436808) Extracted equation: </ </</
(main_task pid=2436808) Solution string: A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer.
(main_task pid=2436808) User: Using the numbers [41, 46, 89, 15], create an equation that equals 86. You can use basic arithmetic operations (+, -, *, /) and each number can only be used once. Show your work in tags. And return the final answer in tags, for example (1 + 2) / 3 .
(main_task pid=2436808) Assistant: Let me solve this step by step.
(main_task pid=2436808) using basic arithmetic operations and the numbers [41, 46, 89, 15].
(main_task pid=2436808) The final answer is in </ </</ </</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</</
(main_task pid=2436808) Invalid equation
(main_task pid=2436808) epoch 0, step 60
(main_task pid=2436808) epoch 0, step 61
(main_task pid=2436808) epoch 0, step 62
(main_task pid=2436808) epoch 0, step 63
(main_task pid=2436808) epoch 0, step 64
(main_task pid=2436808) epoch 0, step 65
(main_task pid=2436808) epoch 0, step 66
(main_task pid=2436808) --------------------------------
(main_task pid=2436808) Target: 20 | Numbers: [14 54 12 64]
(main_task pid=2436808) Extracted equation: None
(main_task pid=2436808) Solution string: A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer.
(main_task pid=2436808) User: Using the numbers [14, 54, 12, 64], create an equation that equals 20. You can use basic arithmetic operations (+, -, *, /) and each number can only be used once. Show your work in tags. And return the final answer in tags, for example (1 + 2) / 3 .
(main_task pid=2436808) Assistant: Let me solve this step by step.
(main_task pid=2436808) using basic arithmetic operations, I can come up with the final answer in </
(main_task pid=2436808) No equation found
(main_task pid=2436808) epoch 0, step 67
(main_task pid=2436808) --------------------------------

[Molri19]

like this： 156 - 147 + 9 to get the final answer 12.<|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|><|endoftext|>

I've encountered the same problem, and my response_length remains unchanged all the time.

[ccl-private]

me too

[yaxundai]

Could it be solved? I have this problem

[yaxundai]

Could it be solved? I have this problem
The base model should use instruct, and the prompt_template should use the same format:
python examples/data_preprocess/countdown.py --template_type=qwen-instruct --local_dir={path_to_your_dataset}

I use the instruct model, but the template doesn't use the corresponding format to get an error like this

[AvisP]

I was having this issue also while using the Qwen 2.5 0.5B Instruct model Then I noticed that the countdown.py has an argument of template for this model. You need to pass --template qwen-instruct when running that script. After that when I tried the endless text stopped. Also make sure you have the Instruct model.

[xuzekun02]

I was having this issue also while using the Qwen 2.5 0.5B Instruct model Then I noticed that the countdown.py has an argument of template for this model. You need to pass --template qwen-instruct when running that script. After that when I tried the endless text stopped. Also make sure you have the Instruct model.

Excuse me, I also encountered the same problem when using the Qwen2.5 3B base model. A lot of <|endoftext|> are output at the end, and the response length is always the maximum value. Why is this?

[AvisP]

I think you should use the Qwen2.5 3B Instruct model and did you parse the data as I suggested by passing --template qwen-instruct to the countdown.py ?