Search results

[BUG] SetGetattributeCallback corrupts module state when ParseFunc fails

type:bugBugBug

Status: Open.

#121925 In tensorflow/tensorflow;

· Fuyugithub opened on Jun 24, 2026

[BUG] FastModuleType tp_setattro crashes on attribute deletion (Py_INCREF on NULL)

type:bugBugBug

Status: Open.

#121923 In tensorflow/tensorflow;

· Fuyugithub opened on Jun 24, 2026

tf.linalg.expm fails for documented float16 input because MatrixSolve has no DT_HALF CPU/GPU kernel

type:bugBugBug

Status: Open.

#121912 In tensorflow/tensorflow;

· tingPetty opened on Jun 24, 2026

Enhance tools/tf_env_collect.sh to provide better diagnostics for modern TensorFlow setups and improve bug report quality

2.21.0tensorflow 2.21.0tensorflow 2.21.0

awaiting PR mergeawaiting PR mergeawaiting PR merge

type:bugBugBug

type:build/installBuild and install issuesBuild and install issues

Status: Open.

#121865 In tensorflow/tensorflow;

· Mattral opened on Jun 23, 2026

tf.make_ndarray raises TypeError for scalar fnuz float8 dtypes (e4m3fnuz / e4m3b11fnuz / e5m2fnuz)

2.21.0tensorflow 2.21.0tensorflow 2.21.0

awaiting PR mergeawaiting PR mergeawaiting PR merge

comp:coreissues related to core part of tensorflowissues related to core part of tensorflow

type:bugBugBug

Status: Open.

#121711 In tensorflow/tensorflow;

· wilmerdooley opened on Jun 21, 2026

Review request: 3 TFLite memory safety fix PRs (open 2+ months)

comp:liteTF Lite related issuesTF Lite related issues

Status: Open.

#121631 In tensorflow/tensorflow;

· TristanInSec opened on Jun 19, 2026

tf.math.erfinv eager fp32 loses ~4 digits of precision near x = ±1 while XLA jit_compile=True is accurate

2.21.0tensorflow 2.21.0tensorflow 2.21.0

awaiting PR mergeawaiting PR mergeawaiting PR merge

comp:xlaXLAXLA

type:bugBugBug

Status: Open.

#121629 In tensorflow/tensorflow;

· wuyii8941 opened on Jun 19, 2026

Implement GDS(NVIDIA) / hipFile(AMD) for directly accessing file from storage to GPU and back

type:featureFeature requestsFeature requests

Status: Open.

#121518 In tensorflow/tensorflow;

· maifeeulasad opened on Jun 18, 2026

# tf.math.erfinv returns inf for largest float32 value below 1.0

2.21.0tensorflow 2.21.0tensorflow 2.21.0

awaiting PR mergeawaiting PR mergeawaiting PR merge

comp:coreissues related to core part of tensorflowissues related to core part of tensorflow

type:bugBugBug

Status: Open.

#121502 In tensorflow/tensorflow;

· ALinrunrun opened on Jun 18, 2026

GPU tf.math.zeta returns NaN for finite s and infinite q

2.21.0tensorflow 2.21.0tensorflow 2.21.0

awaiting PR mergeawaiting PR mergeawaiting PR merge

comp:gpuGPU related issuesGPU related issues

type:bugBugBug

Status: Open.

#121501 In tensorflow/tensorflow;

· ALinrunrun opened on Jun 18, 2026

tf.math.lgamma returns inaccurate result for smallest positive float32 subnormal

2.21.0tensorflow 2.21.0tensorflow 2.21.0

comp:opsOPs related issuesOPs related issues

type:bugBugBug

Status: Open.

#121499 In tensorflow/tensorflow;

· ALinrunrun opened on Jun 18, 2026

tf.math.digamma returns NaN for signed zero and smallest positive float32 subnormal

2.21.0tensorflow 2.21.0tensorflow 2.21.0

comp:opsOPs related issuesOPs related issues

type:bugBugBug

Status: Open.

#121498 In tensorflow/tensorflow;

· ALinrunrun opened on Jun 18, 2026