Skip to content

tfjs-4.8.0

Compare
Choose a tag to compare
@Dedongala Dedongala released this 22 Jun 00:51
· 125 commits to master since this release
74081cb

Core (4.7.0 ==> 4.8.0)

Misc

  • Update monorepo to 4.8.0. (#7766). Thanks, @Dedongala.
  • Avoid draw API warning if the current backend does not have implementation (#7744).
  • Add register name when register the class object (#7717).

Data (4.7.0 ==> 4.8.0)

Misc

Layers (4.7.0 ==> 4.8.0)

Misc

Converter (4.7.0 ==> 4.8.0)

Misc

Node (4.7.0 ==> 4.8.0)

Misc

Wasm (4.7.0 ==> 4.8.0)

Misc

Cpu (4.7.0 ==> 4.8.0)

Misc

Webgl (4.7.0 ==> 4.8.0)

Misc

WebGPU (4.7.0 ==> 4.8.0)

Misc

  • Update lockfiles branch tfjs_4.8.0_lockfiles lock files. (#7769). Thanks, @Dedongala.
  • Update monorepo to 4.8.0. (#7766). Thanks, @Dedongala.
  • webgpu: Refactor some helper functions of matmul/conv2d (#7748). Thanks, @qjia7.
  • [WebGPU] support SparseSegmentSum and SparseSegmentMean (#7746). Thanks, @xhcao.
  • webgpu: Delete benchmark_ops_test file (#7747). Thanks, @qjia7.
  • webgpu: Tighten the condition to go the plain argminmax (#7742). Thanks, @qjia7.
  • [webgpu] Update ELU_DER (#7745). Thanks, @hujiajie.
  • [webgpu] Update ATAN2,MAX,MIN (#7743). Thanks, @hujiajie.
  • [webgpu] Update ADD,COMPLEX_MULTIPLY_*,DIV,MUL,SQUARED_DIFFERENCE,SUB (#7737). Thanks, @hujiajie.dedongala@d