failed running "basenji_sat_bed.py" in tutorials

Question

failed running "basenji_sat_bed.py" in tutorials

yal054 opened this issue 2 years ago · 4 comments

Hi, thank you for maintaining this great package.

I have error when running the "basenji_sat_bed.py" in tutorials

model_strides [128]
target_lengths [1024]
target_crops [0]
Predicting 0
2023-03-13 13:25:22.633776: I tensorflow/stream_executor/cuda/cuda_dnn.cc:368] Loaded cuDNN version 8202
2023-03-13 13:25:25.229024: I tensorflow/stream_executor/cuda/cuda_blas.cc:1786] TensorFloat-32 will be used for the matrix multiplication. This will only be logged once.
ERROR: Sequence 0 failed
Waiting for threads to finish.

Is that a communicate error? Could you let me know how to fix it?
Thanks!

Answer 1 · 2023-03-16T16:25:00.000Z

I think I figured, the "ScoreWorker" is expecting 4 "seq_pred_stats" items to be passed to it, but "seq_preds_js" is missing in the script.

Answer 2 · 2023-03-21T02:49:07.000Z

I encountered the same problem, did you solve the problem?

Answer 3 · 2023-03-21T14:23:52.000Z

After removing "seq_pred_stats", it seems to work fine

Answer 4 · 2023-03-23T21:59:32.000Z

You're right. I added a new score to basenji_sat_vcf without updating basenji_sat_bed. I pushed a fix to the master branch. Thanks!