PydanticSerializationError: ... <class 'numpy.int64'>, an error caused by weave.

[Talonz-z]

I encountered this error during the running of both train and evaluation.
Task failed: PydanticSerializationError: Unable to serialize unknown type: <class 'numpy.int64'>
As this error specially doesn't come with traceback, so it's hard to debug and took me few hours. Finally I found that the problem is about weave, after disabling weave, actor_rollout_ref.rollout.trace.backend=weave \, set this line to 'none' instead of 'weave', the problem can be fixed,

I don't know is this only happened to me(maybe environment error) or not, and I also wonder is there any print info during the evaluation, as it takes hours to finish the training and get result(8x L40). I don't know if closing weave will block the print info or not.

[love-cfall]

he

I encountered this error during the running of both train and evaluation. Task failed: PydanticSerializationError: Unable to serialize unknown type: <class 'numpy.int64'> As this error specially doesn't come with traceback, so it's hard to debug and took me few hours. Finally I found that the problem is about weave, after disabling weave, actor_rollout_ref.rollout.trace.backend=weave \, set this line to 'none' instead of 'weave', the problem can be fixed,

I don't know is this only happened to me(maybe environment error) or not, and I also wonder is there any print info during the evaluation, as it takes hours to finish the training and get result(8x L40). I don't know if closing weave will block the print info or not.

Hello, I’d like to ask for some clarification.

I’m training Qwen3-14B-base on 8×H100, and during the early stage of training I encounter a large number of errors like:

Failed to decode tool call: Expecting property name enclosed in double quotes
Failed to decode tool call: Expecting ',' delimiter
Failed to decode tool call: Extra data
Error when executing tool: 'str' object has no attribute 'get'
Error when executing tool: 'WolframAlpha'
Error when executing tool: 'math'

I’m wondering whether this behavior is normal during the initial phase of training.