20.11 release candidate 3.
  1. e3de4e5 MLBEDSW-3633: SplitV incorrectly placed on CPU by Jacob Bohlin · 3 years, 6 months ago
  2. 14e8a20 vela: Rename --keep-scale-placement CLI by Tim Hall · 3 years, 6 months ago
  3. 061eeb4 MLBEDSW-3424 Arch cache by Louis Verhaard · 3 years, 6 months ago
  4. d266580 MLBEDSW-3562: Improve blockdep calculation by Louis Verhaard · 3 years, 6 months ago
  5. c21d1c7 MLBEDSW-3583: Prevent DMA for converted FullyConnected by Louis Verhaard · 3 years, 6 months ago
  6. 0dd21c7 MLBEDSW-3558: Put FC on CPU when OFM != 2D by Dwight Lidman · 3 years, 6 months ago
  7. 933f55e MLBEDSW-3599: Added API for finding block configs by Louis Verhaard · 3 years, 6 months ago
  8. 34b9dc1 MLBEDSW-3352 Fix ifm end_coord for upsampling by Patrik Gustavsson · 3 years, 6 months ago
  9. 11831ce MLBEDSW-3424: Added API.md by Louis Verhaard · 3 years, 6 months ago
  10. ef0c7fe MLBEDSW-3530: Fix performance issue by Diqing Zhong · 3 years, 6 months ago
  11. 9785cc7 MLBEDSW-3352 Avoid ifm streaming for some cases by Patrik Gustavsson · 3 years, 6 months ago
  12. 34d2917 vela: Improve printing of sets by Michael McGeagh · 3 years, 6 months ago
  13. 54a6111 vela: Fixed formatting of SUPPORTED_OPS by Michael McGeagh · 3 years, 6 months ago
  14. 298e383 MLBEDSW-3352 Fix incorrectly set default value by Michael McGeagh · 3 years, 6 months ago
  15. 5207830 MLBEDSW-3425: Added external API for driver actions by Louis Verhaard · 3 years, 6 months ago
  16. aeae567 MLBEDSW-3424: Expose API through separate file by Louis Verhaard · 3 years, 7 months ago
  17. 27d36f0 MLBEDSW-3468: Move of scale tensors to SRAM after weight compressor by Andreas Nevalainen · 3 years, 6 months ago
  18. 43f8f64 Revert "MLMBED-3450: Do not convert batched fully connected to conv" by Patrik Gustavsson · 3 years, 6 months ago 2.0.0.rc2
  19. d059d8b MLBEDSW-3157: Add test for broadcast shapes by Andreas Nevalainen · 3 years, 6 months ago
  20. d13bd06 vela: Update tool description text by Tim Hall · 3 years, 6 months ago
  21. 907cb03 vela: Tanh and Sigmoid broken in fixup_act_reorder by Tim Hall · 3 years, 6 months ago
  22. b9b515c vela: Remove and change CLI options by Tim Hall · 3 years, 7 months ago
  23. 1bd531d MLBEDSW-3249: Vela config file examples by Tim Hall · 3 years, 7 months ago
  24. c8a7386 vela: Rename Yoda to Ethos-U65 by Tim Hall · 3 years, 7 months ago
  25. c718743 MLBEDSW-3302: Reject per-channel scaling for unsupported ops by Dwight Lidman · 3 years, 6 months ago
  26. 8956761 vela: Improve the scaling is equal check by Tim Hall · 3 years, 7 months ago
  27. 083f103 MLBEDSW-3346: Add index check during padding by Andreas Nevalainen · 3 years, 6 months ago
  28. 986e319 MLBEDSW-3476: Fix performance regression by Diqing Zhong · 3 years, 6 months ago
  29. c8a22f1 MLBEDSW-3251 Add version to external API by Patrik Gustavsson · 3 years, 6 months ago
  30. bdf09f9 [MLBEDSW-3300] Fix DepthwiseConv2D fails when bias tensor quant_values are None by Fredrik Svedberg · 3 years, 6 months ago
  31. 4bb989b [MLBEDSW-3348] Fix Reshape operator fails with TypeError during deserialization by Fredrik Svedberg · 3 years, 6 months ago
  32. a648aa9 vela: Remove ExpandDims from supported ops list by Michael McGeagh · 3 years, 6 months ago
  33. ed67b88 MLMBED-3468: Update scale tensors SRAM size calculation by Andreas Nevalainen · 3 years, 6 months ago
  34. b0ca274 MLBEDSW-3494 Fix rounding of fused Quantized by Patrik Gustavsson · 3 years, 6 months ago
  35. 837dc1b MLBEDSW-3403 Generate supported op report by Michael McGeagh · 3 years, 6 months ago
  36. 69b3176 MLBEDSW-3491: Fix index out of range in code gen by Louis Verhaard · 3 years, 6 months ago
  37. 0411edb MLBEDSW-3493: bug fixes in mark_tensors by Louis Verhaard · 3 years, 6 months ago
  38. 15a8e80 MLMBED-3450: Do not convert batched fully connected to conv by Andreas Nevalainen · 3 years, 6 months ago
  39. c86a9d2 MLBEDSW-3483, KeyError "fused_activation_function" by Louis Verhaard · 3 years, 7 months ago
  40. 2fa1588 MLBEDSW-3350 Put softmax on CPU if beta < 0 by Patrik Gustavsson · 3 years, 6 months ago
  41. e05de45 MLBEDSW-3301: Vela fails ungracefully when reading string buffers by Dwight Lidman · 3 years, 6 months ago
  42. e8a5a78 MLBEDSW-839: Code generation using external API by Louis Verhaard · 3 years, 7 months ago 2.0.0.rc1
  43. dda21af MLBEDSW-3463: StridedSlice fixup function causes infinite recursion by Dwight Lidman · 3 years, 6 months ago
  44. 7aa5aa4 MLBEDSW-3380 Update readme with build flags by Michael McGeagh · 3 years, 6 months ago
  45. e6ccd87 MLBEDSW-3019: Add profiling debug database by Tim Hall · 3 years, 6 months ago
  46. e168b96 Vela: estimate memory transfer efficiency by Diqing Zhong · 3 years, 6 months ago
  47. e5204a6 Vela: Fix perf estimation for conv 1D reshape by Diqing Zhong · 3 years, 7 months ago
  48. 42e833d MLBEDSW-3146: memory transfers cycle estimation by Diqing Zhong · 3 years, 8 months ago
  49. 09387e2 MLBEDSW-3146: Cycle estimation for conv/pooling ops by Diqing Zhong · 3 years, 8 months ago
  50. 897cc14 MLBEDSW-3222: Bias tensors in fast storage by Andreas Nevalainen · 3 years, 7 months ago
  51. 73320a4 MLBEDSW-3377: fixup_stridedslice_output may silently change CPU ops by Dwight Lidman · 3 years, 6 months ago
  52. 6ae0e42 MLBEDSW-2868 Refactor separation of scale + bias tensors by Patrik Gustavsson · 3 years, 7 months ago
  53. fd31428 [MLBEDSW-3227] Improve u65 softmax performance by Fredrik Svedberg · 3 years, 6 months ago
  54. 219ec07 MLBEDSW-3402 SupportedOp now returns external name by Michael McGeagh · 3 years, 6 months ago
  55. fad90c2 MLBEDSW-3212 Remove CLI opt ifm-ofm-overlap by Patrik Gustavsson · 3 years, 7 months ago
  56. 65fd998 MLBEDSW-2412 All constraints have been refactored by Michael McGeagh · 3 years, 7 months ago
  57. 37ce38c MLBEDSW-3275: Added infinity check for Relu scaling values by Jacob Bohlin · 3 years, 7 months ago
  58. 268394d MLBEDSW-1974: Set Scratch buffers size by Jacob Bohlin · 3 years, 9 months ago
  59. 3435958 MLBEDSW-2868 Separate scale+bias tensors by Patrik Gustavsson · 3 years, 7 months ago
  60. 7e1d1d1 Vela: Fix wrong bandwidth by Diqing Zhong · 3 years, 7 months ago
  61. a151f59 MLBEDSW-3212 Enable overlap of elementwise input/output by Patrik Gustavsson · 3 years, 7 months ago
  62. c7c0b1b MLBEDSW-3283: Bug fix: StridedSlice Op is placed on CPU by Diqing Zhong · 3 years, 7 months ago
  63. 7b1654b MLBEDSW-3285: AttributeError Tensor has no attribute by Tim Hall · 3 years, 7 months ago
  64. 4ed38bc vela: Refactor operators to use Kernel objects by Tim Hall · 3 years, 7 months ago
  65. 9358296 vela: Improve the scaling is equal check by Tim Hall · 3 years, 8 months ago
  66. e8887a3 MLBEDSW-603: Improve cycle estimation in elementwise ops by Diqing Zhong · 3 years, 8 months ago
  67. 17afa28 MLBEDSW-3268: Refactor mark_tensors by Louis Verhaard · 3 years, 7 months ago
  68. 6e82708 MLBEDSW-3194: Updated elementwise IFM banks count by Andreas Nevalainen · 3 years, 7 months ago
  69. 1f951fc MLBEDSW-2412 Refactor constraints for conv ops by Michael McGeagh · 3 years, 7 months ago
  70. c6ac194 MLBEDSW-3004: UnpackReshaped can't be serialised by Dwight Lidman · 3 years, 8 months ago
  71. 5f2ea2f MLBEDSW-3219: Suppress CPU info Const/Placeholder by Louis Verhaard · 3 years, 7 months ago
  72. 1068362 Revert "MLBEDSW-3219: Suppress CPU info for Const/Placeholder" by patrik.gustavsson · 3 years, 7 months ago
  73. 04986c0 MLBEDSW-3219: Suppress CPU info for Const/Placeholder by Louis Verhaard · 3 years, 7 months ago
  74. 184b250 vela: Improve extra info in constraint checks by Michael McGeagh · 3 years, 7 months ago
  75. 5554bbe MLBEDSW-3219 Added info print for unsupported operator by Patrik Gustavsson · 3 years, 7 months ago
  76. 23f8716 MLBEDSW-3230 Remove restriction of batching 16 for FC by Patrik Gustavsson · 3 years, 7 months ago
  77. 7ec09e7 MLBEDSW-3220: missing "fused_activation_function" caused crash by Louis Verhaard · 3 years, 7 months ago
  78. 2e186c7 MLBEDSW-3154 Bug fix for LUT ops with IFM from SplitSliceRead by Louis Verhaard · 3 years, 7 months ago
  79. 8359a47 MLBEDSW-3061: Update supported_operators.py by Dwight Lidman · 3 years, 8 months ago
  80. 04f8c00 MLBEDSW-3218: Added operator indices Quantize/Dequantize by Louis Verhaard · 3 years, 7 months ago
  81. e2dbed2 MLBEDSW-2985 Avoid reshape if input when there are no values by Patrik Gustavsson · 3 years, 7 months ago
  82. aee5d75 MLBEDSW-3148: Refactor Operation by Louis Verhaard · 3 years, 8 months ago
  83. 36ad73a MLBEDSW-3154 Fix issue for checking axis in concat by Patrik Gustavsson · 3 years, 7 months ago
  84. d5044a4 Vela: Fix issue with elementwise block config validation by Tim Hall · 3 years, 7 months ago
  85. 37ded34 MLBEDSW-2412 Replace generic restrictions by Michael McGeagh · 3 years, 8 months ago
  86. 1eeea51 vela: SupportedOperators promote to class instance by Michael McGeagh · 3 years, 8 months ago
  87. 3010d9b MLBEDSW-3060 Adjust check if weights fit in sram by Patrik Gustavsson · 3 years, 8 months ago
  88. 8912c53 MLBEDSW-3153: Fix overflow in sigmoid LUT by Louis Verhaard · 3 years, 8 months ago
  89. 530992a MLBEDSW-3001 Fix Min Max OPs not properly checked by Patrik Gustavsson · 3 years, 8 months ago
  90. 0f98b36 [MLBEDSW-2802] Fix 5D tensor crash by Fredrik Svedberg · 3 years, 8 months ago
  91. 0265f40 MLBEDSW-3025: Remove SHRAM from reports by Louis Verhaard · 3 years, 8 months ago
  92. 7ad862a MLBEDSW-3032 Fixed issue in removal of reshapes by Patrik Gustavsson · 3 years, 8 months ago
  93. f03bad3 MLBEDSW-2031: LUT support tanh/sigmoid by Louis Verhaard · 3 years, 8 months ago
  94. eadb166 MLBEDSW-2885: Fix overflow from inf number by Andreas Nevalainen · 3 years, 9 months ago
  95. fa2f92a MLBEDSW-3035: Updated StridedSlice checks by Louis Verhaard · 3 years, 8 months ago
  96. f3d737e MLBEDSW-2811: Add rescaling to ReLus with different scaling by Andreas Nevalainen · 3 years, 8 months ago
  97. 0b9c9a3 MLBEDSW-2337: Intermediate feature maps in fast storage by Louis Verhaard · 3 years, 8 months ago
  98. 8854dc9 MLBEDSW-2788 Fix crash on non-constant weight tensors by Andreas Nevalainen · 3 years, 8 months ago
  99. 6c97e9a MLBEDSW-3018 Change to check for avoiding NHCWB16 by Patrik Gustavsson · 3 years, 8 months ago
  100. 9db529a MLBEDSW-3070: Fix addressing of weights by Louis Verhaard · 3 years, 8 months ago