Search results

Enhancement: Add AutoML capabilities and automated hyperparameter tuning

stat:awaiting responseStatus - Awaiting response from authorStatus - Awaiting response from author

Status: Open.

#120669 In tensorflow/tensorflow;

· anshul23102 opened on Jun 9, 2026

Bug: Memory usage grows unbounded during training, OOM crashes

type:bugBugBug

Status: Open.

#120668 In tensorflow/tensorflow;

· anshul23102 opened on Jun 9, 2026

Security: TensorFlow models vulnerable to adversarial attacks

stat:awaiting responseStatus - Awaiting response from authorStatus - Awaiting response from author

Status: Open.

#120667 In tensorflow/tensorflow;

· anshul23102 opened on Jun 9, 2026

Enhancement: Add better model compression and quantization support

stat:awaiting responseStatus - Awaiting response from authorStatus - Awaiting response from author

Status: Open.

#120666 In tensorflow/tensorflow;

· anshul23102 opened on Jun 9, 2026

Bug: Model serving performance inconsistent, latency spikes

stat:awaiting responseStatus - Awaiting response from authorStatus - Awaiting response from author

type:bugBugBug

type:performancePerformance IssuePerformance Issue

Status: Open.

#120665 In tensorflow/tensorflow;

· anshul23102 opened on Jun 9, 2026

Document Keras lazy initializer seed behavior across eager, tf.function, and XLA

2.20.0tensorflow 2.20.0tensorflow 2.20.0

comp:xlaXLAXLA

type:bugBugBug

type:docs-bugDocument issuesDocument issues

Status: Open.

#120627 In tensorflow/tensorflow;

· laolvfan opened on Jun 8, 2026

XLA changes floor(tan(atan(1.0))) result compared with eager execution

2.20.0tensorflow 2.20.0tensorflow 2.20.0

comp:xlaXLAXLA

type:bugBugBug

Status: Open.

#120622 In tensorflow/tensorflow;

· laolvfan opened on Jun 8, 2026

XLA produces different output for tf.while_loop when loop body depends on Python mutable state

2.20.0tensorflow 2.20.0tensorflow 2.20.0

comp:xlaXLAXLA

type:bugBugBug

Status: Open.

#120620 In tensorflow/tensorflow;

· laolvfan opened on Jun 8, 2026

TensorFlow MultiWorkerMirroredStrategy: XLA and non-XLA forward/backward produce divergent loss and gradient norm with Adam

comp:xlaXLAXLA

TF 2.18

type:bugBugBug

Status: Open.

#120580 In tensorflow/tensorflow;

· whitephilomel opened on Jun 7, 2026

TensorFlow MultiWorkerMirroredStrategy: XLA and non-XLA forward/backward produce divergent loss under the same distributed setup

comp:xlaXLAXLA

TF 2.18

type:bugBugBug

Status: Open.

#120579 In tensorflow/tensorflow;

· whitephilomel opened on Jun 7, 2026

tf.math.bessel_i1 gradient returns 1.0 at x=0, correct value is 0.5

2.21.0tensorflow 2.21.0tensorflow 2.21.0

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

type:bugBugBug

Status: Open.

#120578 In tensorflow/tensorflow;

· wuyii8941 opened on Jun 7, 2026

[RFC] [Feature Request] Introducing tf.llm: High-Level Keras-Compatible API for 1-Line LLM Loading and LoRA Fine-Tuning

type:featureFeature requestsFeature requests

Status: Open.

#120460 In tensorflow/tensorflow;

· Sarkar-AGI opened on Jun 5, 2026