-
Notifications
You must be signed in to change notification settings - Fork 76k
All issues
Issue creation is restricted in this repository
Issues
is:issue state:open
is:issue state:open
Search results
- Status: Open.#125359 In tensorflow/tensorflow;
Heap-buffer-overflow (WRITE) from
.tflitecrafted model2.21.0tensorflow 2.21.0tensorflow 2.21.0comp:liteTF Lite related issuesTF Lite related issuestype:bugBugBugStatus: Open.#124982 In tensorflow/tensorflow;tf.raw_ops.BlockLSTMGrad crashes with fatal check failure (SIGABRT) in Tensor::Slice when seq_len_max exceeds tensor dimension size
2.21.0tensorflow 2.21.0tensorflow 2.21.0awaiting PR mergeawaiting PR mergeawaiting PR mergecomp:opsOPs related issuesOPs related issuestype:bugBugBugStatus: Open.#124974 In tensorflow/tensorflow;@tf.function re-tracing causes unbounded memory growth — traced graphs are never garbage collected
2.21.0tensorflow 2.21.0tensorflow 2.21.0comp:coreissues related to core part of tensorflowissues related to core part of tensorflowtype:bugBugBugStatus: Open.#124972 In tensorflow/tensorflow;tf.raw_ops.CropAndResizeGradBoxes causes SIGSEGV when boxes tensor contains NaN values
2.21.0tensorflow 2.21.0tensorflow 2.21.0comp:opsOPs related issuesOPs related issuestype:bugBugBugStatus: Open.#124955 In tensorflow/tensorflow;tf.distribute.Server crashes with fatal check failure (SIGABRT) when inter_op_parallelism_threads is set to a negative value
2.21.0tensorflow 2.21.0tensorflow 2.21.0comp:opsOPs related issuesOPs related issuestype:bugBugBugStatus: Open.#124954 In tensorflow/tensorflow;tf.raw_ops.QuantizedAvgPool / tf.raw_ops.QuantizedMaxPool cause process abort (SIGABRT) in MklLayoutRewritePass when ksize or strides length is less than 4
2.21.0tensorflow 2.21.0tensorflow 2.21.0comp:opsOPs related issuesOPs related issuestype:bugBugBugStatus: Open.#124953 In tensorflow/tensorflow;tf.nn.separable_conv2d causes process abort (SIGABRT) in MklLayoutRewritePass when strides contains INT64_MAX
2.21.0tensorflow 2.21.0tensorflow 2.21.0comp:coreissues related to core part of tensorflowissues related to core part of tensorflowtype:bugBugBugStatus: Open.#124952 In tensorflow/tensorflow;Broadcasting binary ops give CUDA_ERROR_ILLEGAL_ADDRESS above 2^31 output elements on GPU
2.20.0tensorflow 2.20.0tensorflow 2.20.0awaiting PR mergeawaiting PR mergeawaiting PR mergecomp:gpuGPU related issuesGPU related issuescomp:opsOPs related issuesOPs related issuestype:bugBugBugStatus: Open.#124949 In tensorflow/tensorflow;Explicit tf.device() placement on a nonexistent device is correctly rejected by eager execution and autoclustering, but silently ignored under jit_compile=True [Recent/StarCoder]
comp:xlaXLAXLAtype:bugBugBugtype:supportSupport issuesSupport issuesStatus: Open.#124880 In tensorflow/tensorflow;