- 618f66a Define TOSA versioning and compatibility by Eric Kunze · 7 weeks ago
- 0e121c0 Clarify error bound for non-normal values by Eric Kunze · 8 weeks ago
- aa162aa Switch fp8 to use non-saturating mode when converting by Eric Kunze · 7 weeks ago
- 0d7d001 Clarify wording on level definitions by Eric Kunze · 10 weeks ago
- 6bf6ac8 Remove Main Training profile by Eric Kunze · 2 months ago
- 4f95cd8 Main Conformance: Update SIN,COS error bounds by Dominic Symes · 2 months ago
- 0afe61f Modify convolution operators to improve bias handling by Eric Kunze · 4 months ago
- 6dd3410 Modify TOSA profiles by Eric Kunze · 3 months ago
- 2266c7a Accuracy requirements for RESIZE operator by Eric Kunze · 7 months ago
- 1f05883 Add SIN/COS to elementwise operator set by Eric Kunze · 4 months ago
- 74e2ceb Initial FP8 support by Eric Kunze · 8 months ago
- 0e87924 Main Conformance: Update POW error bound by Dominic Symes · 4 months ago
- 526f6c7 Add section of shape operators by Eric Kunze · 5 months ago v0.90.0
- 6085883 Handle CAST from fp-type to narrow integer types by Eric Kunze · 4 months ago
- f9fcb61 pseudocode: use true/false consistently for boolean values by Kevin Petit · 4 months ago
- 8754ec2 Main Conformance: TANH change, SCATTER clarification by Dominic Symes · 6 months ago
- 2f3f4a2 Fix typo in i48_t description by Eric Kunze · 5 months ago
- 83e79b5 Main Conformance: Clarify REDUCE_PRODUCT conformance by Dominic Symes · 5 months ago
- 18acfe3 Add note to POW accuracy by Eric Kunze · 5 months ago
- 2bc6c57 Main Conformance: Update SIGMOID, TANH precision by Dominic Symes · 6 months ago
- a46cf1d Main Conformance: Update RSQRT precision by Dominic Symes · 7 months ago
- f791b44 Main Conformance: Update EXP, POW precision by Dominic Symes · 7 months ago
- c237b7e Main conformance: clarify floating point accuracy by Dominic Symes · 9 months ago
- 0ae7fd6 Define missing accuracy bounds by Eric Kunze · 8 months ago
- 8bf6e71 Main Conformance: Add local_bound to RFFT2D by Dominic Symes · 9 months ago
- b203512 Main Conformance: Update dot product error bound by Dominic Symes · 9 months ago v0.90a0
- fb0284e Change TOSA specification to signless types by Eric Kunze · 11 months ago
- b5b0678 Main conformance: Add local bound flag for convolutions by Dominic Symes · 10 months ago
- 830b43b Add DIM operator and operations on shape_t values by Dominic Symes · 1 year, 1 month ago
- 7b0f1c9 COND_IF/WHILE_LOOP: Add nesting level limit by Dominic Symes · 11 months ago
- 52c89f4 Add ERF activation function by Eric Kunze · 1 year, 1 month ago
- 5333c25 Formalise the description of operator argument types by Kevin Petit · 1 year, 1 month ago
- 277a4f1 Update descriptions for activation functions. by Eric Kunze · 1 year, 1 month ago
- 98b3e33 Replace uint6_t with int8_t by Kevin Petit · 1 year, 1 month ago
- 5b936a3 Main inference compliance testing updates by Dominic Symes · 1 year, 3 months ago
- bc72ba8 Move the tensor size limit into the level range definition by Dominic Symes · 1 year, 1 month ago
- 211c5f5 Generate level maximums from the XML by Kevin Petit · 1 year, 1 month ago
- e0fee3d Fix description of tensor size limit by Kevin Petit · 1 year, 2 months ago
- 8f7688d Main inference conformance additions by Dominic Symes · 1 year, 3 months ago
- c386a05 Main inference conformance draft by Dominic Symes · 1 year, 4 months ago
- e4d6a1b Add Levels defintion by Dominic Symes · 1 year, 7 months ago
- 0205d99 Define the index type for tensor co-ordinates by Dominic Symes · 1 year, 8 months ago
- b8522ba Relax limit on input value for RESCALE by Eric Kunze · 1 year, 11 months ago
- 42229d0 Initial work on floating-point type definition by Eric Kunze · 2 years, 2 months ago
- f9e5ba9 Rework the introduction by Eric Kunze · 2 years ago
- eef012e Add the uint16_t data type by Eric Kunze · 2 years, 1 month ago
- cb6c6b3 Specify width of integer types by Dominic Symes · 2 years, 1 month ago
- ce6e02c Update floating point edge cases by Eric Kunze · 2 years, 3 months ago
- 2ff79fe TABLE: Clarify the interpolation slope range by Dominic Symes · 2 years, 4 months ago
- 3cb7535 apply_scale_32: adjust range checking by Dominic Symes · 2 years, 4 months ago
- 2a3d560 Ensure ranks match in apply_broadcast by Eric Kunze · 2 years, 5 months ago
- a3eded0 Add rudimentary spell check for specification by Eric Kunze · 2 years, 6 months ago
- 3170439 Remove zp subtraction from tensor_read pseudocode by Eric Kunze · 2 years, 7 months ago
- c949f8a Allow PAD operator to pad with non-zero by Eric Kunze · 2 years, 9 months ago
- 173fc16 Clarify range limitations for tensors by Eric Kunze · 2 years, 10 months ago
- ca2a854 Add definition of TOSA compliance by Dominic Symes · 3 years, 3 months ago
- a910153 Replace assert with REQUIRE() by Eric Kunze · 3 years ago
- 839830a Make pseudocode block tags consistent by Eric Kunze · 3 years, 3 months ago
- 8e4a9d3 Adjust pseudocode types to account for zero point by Eric Kunze · 3 years, 3 months ago
- 1e9ba65 Consistency cleanup by Eric Kunze · 3 years, 4 months ago
- f8bd586 Update elementwise operator overview by Eric Kunze · 3 years, 4 months ago
- 2dce0d0 Remove aint8 type by Eric Kunze · 3 years, 5 months ago
- e5d22a7 Disambiguate scale_t in RESIZE command by Eric Kunze · 3 years, 6 months ago
- fa1b324 Update introduction sections by Eric Kunze · 3 years, 7 months ago
- 57e79c0 Add int48 to supported number formats by Eric Kunze · 3 years, 7 months ago
- f7179b5 MUL: Add right shift on 32x32 multiply by Dominic Symes · 3 years, 7 months ago
- 298a029 Make rank() consistent across files by Dominic Symes · 3 years, 7 months ago
- 1d84a1e Use rank() instead of dimensions() in pseudocode by Eric Kunze · 3 years, 8 months ago
- 3309a53 TOSA spec 0.20.0, initial public version by Eric Kunze · 3 years, 8 months ago v0.20.0