diff options
Diffstat (limited to 'results/classifier/gemma3:12b/hypervisor/601')
| -rw-r--r-- | results/classifier/gemma3:12b/hypervisor/601 | 21 |
1 files changed, 21 insertions, 0 deletions
diff --git a/results/classifier/gemma3:12b/hypervisor/601 b/results/classifier/gemma3:12b/hypervisor/601 new file mode 100644 index 000000000..0bef3f732 --- /dev/null +++ b/results/classifier/gemma3:12b/hypervisor/601 @@ -0,0 +1,21 @@ + +import tensorflow causes qemu: uncaught target signal 6 (Aborted) - core dumped +Description of problem: +Crashes when importing tensorflow in Docker container under --platorm linux/amd64 on M1 Mac +``` +2021-09-06 13:35:24.435613: F tensorflow/core/lib/monitoring/sampler.cc:42] Check failed: bucket_limits_[i] > bucket_limits_[i - 1] (0 vs. 10) +qemu: uncaught target signal 6 (Aborted) - core dumped +``` +Steps to reproduce: +See https://gitlab.com/ryan-feather/docker-tensorflow-qemu-bug/ for Dockerfile and description of steps repeating here. +1. Using the dockerfile +``` +FROM python:3.9-buster +RUN pip install tensorflow==2.6.0 + +``` +2. `docker buildx build --iidfile build.id --platform linux/amd64 . --progress=plain` +3. ``` docker run --platform linux/amd64 `cat build.id` python -c "import tensorflow"``` +Additional information: +See +https://github.com/docker/for-mac/issues/5342 where the Docker team suggests this is a qemu bug. I couldn't find where anyone had opened one of these here, so hopefully this isn't a duplicate. |