areusch commented on a change in pull request #22: URL: https://github.com/apache/tvm-rfcs/pull/22#discussion_r692378556
########## File path: rfcs/0022-tir-non-scalar-constants.md ########## @@ -0,0 +1,107 @@ + +- Feature Name: tir_non_scalar_constants +- Start Date: 2021-06-01 +- RFC PR: https://github.com/apache/tvm-rfcs/pull/22 +- GitHub Issue: TBD + +# 1. Summary + +This RFC proposes how non-scalar constants could be represented in TIR and used by passes in the lowering process. + +# 2. Motivation + +Currently, the non-scalar constants could be represented in Relay (relay.Constant) to be used by relay passes but not in TIR. Therefore, when performing lowering using TIR passes, we have to maintain a side-channel of tir::Var to constant non-scalar data mapping to perform transformations that could use the knowledge where some of the data are constants. Review comment: nit: say "non-scalar constants are represented in Relay" ########## File path: rfcs/0022-tir-non-scalar-constants.md ########## @@ -0,0 +1,107 @@ + +- Feature Name: tir_non_scalar_constants +- Start Date: 2021-06-01 +- RFC PR: https://github.com/apache/tvm-rfcs/pull/22 +- GitHub Issue: TBD + +# 1. Summary + +This RFC proposes how non-scalar constants could be represented in TIR and used by passes in the lowering process. + +# 2. Motivation + +Currently, the non-scalar constants could be represented in Relay (relay.Constant) to be used by relay passes but not in TIR. Therefore, when performing lowering using TIR passes, we have to maintain a side-channel of tir::Var to constant non-scalar data mapping to perform transformations that could use the knowledge where some of the data are constants. Review comment: flip the second sentence around: "Therefore, when a compiler pass is operating on TIR, it needs additional input to determine whether a `tir::Var` is runtime-supplied data or a compile-time constant (and what that constant data is). Should that pass want to transform the constant data, the additional input also becomes an output. In order to provide this data, the compiler has to maintain a side-channel map, complicating the Pass interface." ########## File path: rfcs/0022-tir-non-scalar-constants.md ########## @@ -0,0 +1,107 @@ + +- Feature Name: tir_non_scalar_constants +- Start Date: 2021-06-01 +- RFC PR: https://github.com/apache/tvm-rfcs/pull/22 +- GitHub Issue: TBD + +# 1. Summary + +This RFC proposes how non-scalar constants could be represented in TIR and used by passes in the lowering process. + +# 2. Motivation + +Currently, the non-scalar constants could be represented in Relay (relay.Constant) to be used by relay passes but not in TIR. Therefore, when performing lowering using TIR passes, we have to maintain a side-channel of tir::Var to constant non-scalar data mapping to perform transformations that could use the knowledge where some of the data are constants. + +Few example scenarios as further motivation : + +## Weight compression + +When lowering for accelerators (E.g. : [Arm(R) Ethos(TM)-U NPU](https://github.com/apache/tvm-rfcs/pull/11)), certain operations will need to get tiled to co-optimize performance and memory utilization. Such tiling patterns create slices of weights that need compressing that will end up with varying sizes. Therefore, the knowledge of some tir::Vars refer to constants are critical in the level of TIR to perform this. + +## Memory Planning + +The TIR program has the ability to express both inter and intra operator memory requirement, post-scheduling as explained further by [Unified Static Memory Planning RFC](https://github.com/apache/tvm-rfcs/pull/9). It would be better if the constants could be embedded to the TIR PrimFunc. Moreover, this allows various [target-dependent lowerings](https://github.com/apache/tvm-rfcs/pull/10), to produce TIR PrimFuncs with constants in it. + +## Winograd Constants + +The Winograd transformation (used for fast GEMMs) involves multiplication by a hard-coded constant tensor. This is currently accomplished in TE using a complicated TE compute expression with many nested selects. Being able to directly express a constant tensor here would significantly simplify this code. + + +# 3. Guide-level explanation + +This is not particularly a user-facing feature and this will allow constants to be 'linked' to TIR. Initially, we are planning to use this with gated on '-link-params' argument for relay.build and TVMC. + +# 4. Reference-level explanation Review comment: does the example from the other RFC still hold? if so, worth copying here? ########## File path: rfcs/0022-tir-non-scalar-constants.md ########## @@ -0,0 +1,107 @@ + +- Feature Name: tir_non_scalar_constants +- Start Date: 2021-06-01 +- RFC PR: https://github.com/apache/tvm-rfcs/pull/22 +- GitHub Issue: TBD + +# 1. Summary + +This RFC proposes how non-scalar constants could be represented in TIR and used by passes in the lowering process. + +# 2. Motivation + +Currently, the non-scalar constants could be represented in Relay (relay.Constant) to be used by relay passes but not in TIR. Therefore, when performing lowering using TIR passes, we have to maintain a side-channel of tir::Var to constant non-scalar data mapping to perform transformations that could use the knowledge where some of the data are constants. + +Few example scenarios as further motivation : + +## Weight compression + +When lowering for accelerators (E.g. : [Arm(R) Ethos(TM)-U NPU](https://github.com/apache/tvm-rfcs/pull/11)), certain operations will need to get tiled to co-optimize performance and memory utilization. Such tiling patterns create slices of weights that need compressing that will end up with varying sizes. Therefore, the knowledge of some tir::Vars refer to constants are critical in the level of TIR to perform this. + +## Memory Planning + +The TIR program has the ability to express both inter and intra operator memory requirement, post-scheduling as explained further by [Unified Static Memory Planning RFC](https://github.com/apache/tvm-rfcs/pull/9). It would be better if the constants could be embedded to the TIR PrimFunc. Moreover, this allows various [target-dependent lowerings](https://github.com/apache/tvm-rfcs/pull/10), to produce TIR PrimFuncs with constants in it. + +## Winograd Constants + +The Winograd transformation (used for fast GEMMs) involves multiplication by a hard-coded constant tensor. This is currently accomplished in TE using a complicated TE compute expression with many nested selects. Being able to directly express a constant tensor here would significantly simplify this code. + + +# 3. Guide-level explanation + +This is not particularly a user-facing feature and this will allow constants to be 'linked' to TIR. Initially, we are planning to use this with gated on '-link-params' argument for relay.build and TVMC. + +# 4. Reference-level explanation + +The proposal is quite simple and it could be explained as follows : + +``` +@tvm.script.tir +def myfunc(): + param = tir.allocate_const([1, 1, 1, 1, 1, 1, 1, 1, 1, 1], "int32", [10]) +``` + +This follows closely the semantics of tir.allocate and the difference being it represent a buffer filled with constants. + +There are mainly two ways of constants being created in the lowering : + +A1. Linking the params of the model (relay.Constants) + +A2. Creation of constants in the lowering. Review comment: though i agree lower is a function, at this level could you clarify ("when scheduling relay into TIR"). Could you also indicate that this is different from A1 in that the constants are created by the scheduling system and therefore can be directly embedded into the operator function? ########## File path: rfcs/0022-tir-non-scalar-constants.md ########## @@ -0,0 +1,107 @@ + +- Feature Name: tir_non_scalar_constants +- Start Date: 2021-06-01 +- RFC PR: https://github.com/apache/tvm-rfcs/pull/22 +- GitHub Issue: TBD + +# 1. Summary + +This RFC proposes how non-scalar constants could be represented in TIR and used by passes in the lowering process. + +# 2. Motivation + +Currently, the non-scalar constants could be represented in Relay (relay.Constant) to be used by relay passes but not in TIR. Therefore, when performing lowering using TIR passes, we have to maintain a side-channel of tir::Var to constant non-scalar data mapping to perform transformations that could use the knowledge where some of the data are constants. + +Few example scenarios as further motivation : + +## Weight compression + +When lowering for accelerators (E.g. : [Arm(R) Ethos(TM)-U NPU](https://github.com/apache/tvm-rfcs/pull/11)), certain operations will need to get tiled to co-optimize performance and memory utilization. Such tiling patterns create slices of weights that need compressing that will end up with varying sizes. Therefore, the knowledge of some tir::Vars refer to constants are critical in the level of TIR to perform this. + +## Memory Planning + +The TIR program has the ability to express both inter and intra operator memory requirement, post-scheduling as explained further by [Unified Static Memory Planning RFC](https://github.com/apache/tvm-rfcs/pull/9). It would be better if the constants could be embedded to the TIR PrimFunc. Moreover, this allows various [target-dependent lowerings](https://github.com/apache/tvm-rfcs/pull/10), to produce TIR PrimFuncs with constants in it. + +## Winograd Constants + +The Winograd transformation (used for fast GEMMs) involves multiplication by a hard-coded constant tensor. This is currently accomplished in TE using a complicated TE compute expression with many nested selects. Being able to directly express a constant tensor here would significantly simplify this code. Review comment: do you mind including a link to code for those not acquainted? ########## File path: rfcs/0022-tir-non-scalar-constants.md ########## @@ -0,0 +1,107 @@ + +- Feature Name: tir_non_scalar_constants +- Start Date: 2021-06-01 +- RFC PR: https://github.com/apache/tvm-rfcs/pull/22 +- GitHub Issue: TBD + +# 1. Summary + +This RFC proposes how non-scalar constants could be represented in TIR and used by passes in the lowering process. + +# 2. Motivation + +Currently, the non-scalar constants could be represented in Relay (relay.Constant) to be used by relay passes but not in TIR. Therefore, when performing lowering using TIR passes, we have to maintain a side-channel of tir::Var to constant non-scalar data mapping to perform transformations that could use the knowledge where some of the data are constants. + +Few example scenarios as further motivation : + +## Weight compression + +When lowering for accelerators (E.g. : [Arm(R) Ethos(TM)-U NPU](https://github.com/apache/tvm-rfcs/pull/11)), certain operations will need to get tiled to co-optimize performance and memory utilization. Such tiling patterns create slices of weights that need compressing that will end up with varying sizes. Therefore, the knowledge of some tir::Vars refer to constants are critical in the level of TIR to perform this. Review comment: nit: maybe "see ARM(R) Ethos(TM)-U NPU RFC" ########## File path: rfcs/0022-tir-non-scalar-constants.md ########## @@ -0,0 +1,107 @@ + +- Feature Name: tir_non_scalar_constants +- Start Date: 2021-06-01 +- RFC PR: https://github.com/apache/tvm-rfcs/pull/22 +- GitHub Issue: TBD + +# 1. Summary + +This RFC proposes how non-scalar constants could be represented in TIR and used by passes in the lowering process. + +# 2. Motivation + +Currently, the non-scalar constants could be represented in Relay (relay.Constant) to be used by relay passes but not in TIR. Therefore, when performing lowering using TIR passes, we have to maintain a side-channel of tir::Var to constant non-scalar data mapping to perform transformations that could use the knowledge where some of the data are constants. + +Few example scenarios as further motivation : + +## Weight compression + +When lowering for accelerators (E.g. : [Arm(R) Ethos(TM)-U NPU](https://github.com/apache/tvm-rfcs/pull/11)), certain operations will need to get tiled to co-optimize performance and memory utilization. Such tiling patterns create slices of weights that need compressing that will end up with varying sizes. Therefore, the knowledge of some tir::Vars refer to constants are critical in the level of TIR to perform this. + +## Memory Planning + +The TIR program has the ability to express both inter and intra operator memory requirement, post-scheduling as explained further by [Unified Static Memory Planning RFC](https://github.com/apache/tvm-rfcs/pull/9). It would be better if the constants could be embedded to the TIR PrimFunc. Moreover, this allows various [target-dependent lowerings](https://github.com/apache/tvm-rfcs/pull/10), to produce TIR PrimFuncs with constants in it. Review comment: > It would be better I believe you, but state why. Also state the advantage of producing TIR PrimFuncs in target-depdendent lowerings. ########## File path: rfcs/0022-tir-non-scalar-constants.md ########## @@ -0,0 +1,107 @@ + +- Feature Name: tir_non_scalar_constants +- Start Date: 2021-06-01 +- RFC PR: https://github.com/apache/tvm-rfcs/pull/22 +- GitHub Issue: TBD + +# 1. Summary + +This RFC proposes how non-scalar constants could be represented in TIR and used by passes in the lowering process. + +# 2. Motivation + +Currently, the non-scalar constants could be represented in Relay (relay.Constant) to be used by relay passes but not in TIR. Therefore, when performing lowering using TIR passes, we have to maintain a side-channel of tir::Var to constant non-scalar data mapping to perform transformations that could use the knowledge where some of the data are constants. + +Few example scenarios as further motivation : + +## Weight compression + +When lowering for accelerators (E.g. : [Arm(R) Ethos(TM)-U NPU](https://github.com/apache/tvm-rfcs/pull/11)), certain operations will need to get tiled to co-optimize performance and memory utilization. Such tiling patterns create slices of weights that need compressing that will end up with varying sizes. Therefore, the knowledge of some tir::Vars refer to constants are critical in the level of TIR to perform this. + +## Memory Planning + +The TIR program has the ability to express both inter and intra operator memory requirement, post-scheduling as explained further by [Unified Static Memory Planning RFC](https://github.com/apache/tvm-rfcs/pull/9). It would be better if the constants could be embedded to the TIR PrimFunc. Moreover, this allows various [target-dependent lowerings](https://github.com/apache/tvm-rfcs/pull/10), to produce TIR PrimFuncs with constants in it. + +## Winograd Constants + +The Winograd transformation (used for fast GEMMs) involves multiplication by a hard-coded constant tensor. This is currently accomplished in TE using a complicated TE compute expression with many nested selects. Being able to directly express a constant tensor here would significantly simplify this code. + + +# 3. Guide-level explanation + +This is not particularly a user-facing feature and this will allow constants to be 'linked' to TIR. Initially, we are planning to use this with gated on '-link-params' argument for relay.build and TVMC. + +# 4. Reference-level explanation + +The proposal is quite simple and it could be explained as follows : + +``` +@tvm.script.tir +def myfunc(): + param = tir.allocate_const([1, 1, 1, 1, 1, 1, 1, 1, 1, 1], "int32", [10]) +``` + +This follows closely the semantics of tir.allocate and the difference being it represent a buffer filled with constants. + +There are mainly two ways of constants being created in the lowering : + +A1. Linking the params of the model (relay.Constants) + +A2. Creation of constants in the lowering. + +For A1, this should only be done if the target support codegeneration of the constant data as part of the operators. Review comment: Since it's not possible to do this for graph-level constants outside AOT flow, specify the future plan ########## File path: rfcs/0022-tir-non-scalar-constants.md ########## @@ -0,0 +1,107 @@ + +- Feature Name: tir_non_scalar_constants +- Start Date: 2021-06-01 +- RFC PR: https://github.com/apache/tvm-rfcs/pull/22 +- GitHub Issue: TBD + +# 1. Summary + +This RFC proposes how non-scalar constants could be represented in TIR and used by passes in the lowering process. + +# 2. Motivation + +Currently, the non-scalar constants could be represented in Relay (relay.Constant) to be used by relay passes but not in TIR. Therefore, when performing lowering using TIR passes, we have to maintain a side-channel of tir::Var to constant non-scalar data mapping to perform transformations that could use the knowledge where some of the data are constants. + +Few example scenarios as further motivation : + +## Weight compression + +When lowering for accelerators (E.g. : [Arm(R) Ethos(TM)-U NPU](https://github.com/apache/tvm-rfcs/pull/11)), certain operations will need to get tiled to co-optimize performance and memory utilization. Such tiling patterns create slices of weights that need compressing that will end up with varying sizes. Therefore, the knowledge of some tir::Vars refer to constants are critical in the level of TIR to perform this. + +## Memory Planning + +The TIR program has the ability to express both inter and intra operator memory requirement, post-scheduling as explained further by [Unified Static Memory Planning RFC](https://github.com/apache/tvm-rfcs/pull/9). It would be better if the constants could be embedded to the TIR PrimFunc. Moreover, this allows various [target-dependent lowerings](https://github.com/apache/tvm-rfcs/pull/10), to produce TIR PrimFuncs with constants in it. + +## Winograd Constants + +The Winograd transformation (used for fast GEMMs) involves multiplication by a hard-coded constant tensor. This is currently accomplished in TE using a complicated TE compute expression with many nested selects. Being able to directly express a constant tensor here would significantly simplify this code. + + +# 3. Guide-level explanation + +This is not particularly a user-facing feature and this will allow constants to be 'linked' to TIR. Initially, we are planning to use this with gated on '-link-params' argument for relay.build and TVMC. + +# 4. Reference-level explanation + +The proposal is quite simple and it could be explained as follows : + +``` +@tvm.script.tir +def myfunc(): + param = tir.allocate_const([1, 1, 1, 1, 1, 1, 1, 1, 1, 1], "int32", [10]) +``` + +This follows closely the semantics of tir.allocate and the difference being it represent a buffer filled with constants. + +There are mainly two ways of constants being created in the lowering : + +A1. Linking the params of the model (relay.Constants) Review comment: can you briefly state how exactly these are produced (e.g. bound relay Var are translated to tir.Constant <specify the function where these are created>). ########## File path: rfcs/0022-tir-non-scalar-constants.md ########## @@ -0,0 +1,107 @@ + +- Feature Name: tir_non_scalar_constants +- Start Date: 2021-06-01 +- RFC PR: https://github.com/apache/tvm-rfcs/pull/22 +- GitHub Issue: TBD + +# 1. Summary + +This RFC proposes how non-scalar constants could be represented in TIR and used by passes in the lowering process. + +# 2. Motivation + +Currently, the non-scalar constants could be represented in Relay (relay.Constant) to be used by relay passes but not in TIR. Therefore, when performing lowering using TIR passes, we have to maintain a side-channel of tir::Var to constant non-scalar data mapping to perform transformations that could use the knowledge where some of the data are constants. + +Few example scenarios as further motivation : + +## Weight compression + +When lowering for accelerators (E.g. : [Arm(R) Ethos(TM)-U NPU](https://github.com/apache/tvm-rfcs/pull/11)), certain operations will need to get tiled to co-optimize performance and memory utilization. Such tiling patterns create slices of weights that need compressing that will end up with varying sizes. Therefore, the knowledge of some tir::Vars refer to constants are critical in the level of TIR to perform this. + +## Memory Planning + +The TIR program has the ability to express both inter and intra operator memory requirement, post-scheduling as explained further by [Unified Static Memory Planning RFC](https://github.com/apache/tvm-rfcs/pull/9). It would be better if the constants could be embedded to the TIR PrimFunc. Moreover, this allows various [target-dependent lowerings](https://github.com/apache/tvm-rfcs/pull/10), to produce TIR PrimFuncs with constants in it. + +## Winograd Constants + +The Winograd transformation (used for fast GEMMs) involves multiplication by a hard-coded constant tensor. This is currently accomplished in TE using a complicated TE compute expression with many nested selects. Being able to directly express a constant tensor here would significantly simplify this code. + + +# 3. Guide-level explanation + +This is not particularly a user-facing feature and this will allow constants to be 'linked' to TIR. Initially, we are planning to use this with gated on '-link-params' argument for relay.build and TVMC. + +# 4. Reference-level explanation + +The proposal is quite simple and it could be explained as follows : + +``` +@tvm.script.tir +def myfunc(): + param = tir.allocate_const([1, 1, 1, 1, 1, 1, 1, 1, 1, 1], "int32", [10]) +``` + +This follows closely the semantics of tir.allocate and the difference being it represent a buffer filled with constants. + +There are mainly two ways of constants being created in the lowering : + +A1. Linking the params of the model (relay.Constants) + +A2. Creation of constants in the lowering. + +For A1, this should only be done if the target support codegeneration of the constant data as part of the operators. + +For A2, the lowering for targets that support constant as part of the operators, there can be new (differently sized) constants could be created due to optimizations such as weight compression as required by the target. Review comment: can you explain how the compiler knows which targets support this? ########## File path: rfcs/0022-tir-non-scalar-constants.md ########## @@ -0,0 +1,107 @@ + +- Feature Name: tir_non_scalar_constants +- Start Date: 2021-06-01 +- RFC PR: https://github.com/apache/tvm-rfcs/pull/22 +- GitHub Issue: TBD + +# 1. Summary + +This RFC proposes how non-scalar constants could be represented in TIR and used by passes in the lowering process. + +# 2. Motivation + +Currently, the non-scalar constants could be represented in Relay (relay.Constant) to be used by relay passes but not in TIR. Therefore, when performing lowering using TIR passes, we have to maintain a side-channel of tir::Var to constant non-scalar data mapping to perform transformations that could use the knowledge where some of the data are constants. + +Few example scenarios as further motivation : + +## Weight compression + +When lowering for accelerators (E.g. : [Arm(R) Ethos(TM)-U NPU](https://github.com/apache/tvm-rfcs/pull/11)), certain operations will need to get tiled to co-optimize performance and memory utilization. Such tiling patterns create slices of weights that need compressing that will end up with varying sizes. Therefore, the knowledge of some tir::Vars refer to constants are critical in the level of TIR to perform this. + +## Memory Planning + +The TIR program has the ability to express both inter and intra operator memory requirement, post-scheduling as explained further by [Unified Static Memory Planning RFC](https://github.com/apache/tvm-rfcs/pull/9). It would be better if the constants could be embedded to the TIR PrimFunc. Moreover, this allows various [target-dependent lowerings](https://github.com/apache/tvm-rfcs/pull/10), to produce TIR PrimFuncs with constants in it. + +## Winograd Constants + +The Winograd transformation (used for fast GEMMs) involves multiplication by a hard-coded constant tensor. This is currently accomplished in TE using a complicated TE compute expression with many nested selects. Being able to directly express a constant tensor here would significantly simplify this code. + + +# 3. Guide-level explanation + +This is not particularly a user-facing feature and this will allow constants to be 'linked' to TIR. Initially, we are planning to use this with gated on '-link-params' argument for relay.build and TVMC. Review comment: say "Intially, tir.Constant will only be created during scheduling when `-link-params` is included in the Target (e.g. to `relay.build` and to TVMC)" -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@tvm.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
