AI4D Yorùbá Machine Translation Challenge
Helping Nigeria
$2 000 USD
Completed (almost 5 years ago)
Machine Translation
683 joined
84 active
Start
Dec 04, 20
Close
May 30, 21
Reveal
May 30, 21
ResourceExhaustedError: Out Of Memory
Help · 17 May 2021, 00:20 · 2
Hi Professionals,
Kindly help with hint of fixing this... I'm training my model on Google Colab. Attempts made by me so far is to reduce the batch size, yet the same error. I also had several crashes while attempting this. Thanks in anticipation
ResourceExhaustedError: OOM when allocating tensor with shape[3200,17108] and type float on /job:localhost/replica:0/task:0/device:GPU:0 by allocator GPU_0_bfc [[node sequential/dense/Tensordot/MatMul (defined at <ipython-input-33-b0d0cbc7a5c1>:2) ]] Hint: If you want to see a list of allocated tensors when OOM happens, add report_tensor_allocations_upon_oom to RunOptions for current allocation info. [Op:__inference_predict_function_165166]
Hello, sequence length and batch size both affect memory usage. Reduce both. On Colab, each time this error occurs and you change any of the above, reset runtime/factory reset to reallocate memory based on your new setting.
Thank you Sir.