OPEN: Snitch support integration by tahaelbayad · Pull Request #14 · pulp-platform/Deeploy

tahaelbayad · 2024-11-15T13:08:17Z

This PR adds the support for the Snitch Cluster and its testing infrastructure already integrated in the GitLab repository.

Added

cMake Flow for the Snitch Cluster
Added snitch_cluster to Makefile
New Snitch platform with testing application
Testrunner for tiled and untiled execution (testRunner_snitch.py, testRunner_tiled_snitch.py)
Minimal library with CycleCounter and utility function (deeploy_log)

add snitch support snitch repo changed snitch repo changed switching to main branch of snitch add snitch and banshee commit hash

fixing typo change image in yml files cahnge image in yml files fix typo

linting

lukamac

Not bad but needs a few changes here and there.
I still haven't looked properly at the Dockerfile so there might be some comments there too.

Xeratec · 2024-12-03T09:54:57Z

 runs-on: ubuntu-22.04
 container:
- image: ghcr.io/pulp-platform/deeploy:main
+ image: ghcr.io/tahaelbayad/deeploy:main


At what point do we change this? Right before merging and ensuring we rebuild the pulp-platform docker container, right?
(Same for the other occurrences)

Exactly. First, we check that the modified CI passes using the modified Docker Container. Then, we reverted the container link to point at the main repo container. Finally, we merge and manually trigger the docker build with the new Dockerfile, then restart the CI on the main repo that would use the new docker.

NIT, but then I suggest adding a todo in the description (using the "- [ ]" notation. This will also ensure that the PR is not merged before all open tasks are completed.

- cMake Flow for the Snitch Cluster - Added `snitch_cluster` to Makefile - New Snitch platform with testing application - Testrunner for tiled and untiled execution (`testRunner_snitch.py`, `testRunner_tiled_snitch.py`) - Minimal library with CycleCounter and utility function - Update the Banshee's commit to include a recent PR.

This release contains major architectural changes, new platform support, enhanced simulation workflows, floating-point kernel support, training infrastructure for CCT models, memory allocation strategies, and documentation improvements. After merging this into `main`, the release process will proceed with: - Pushing a Git tag for the release after merging this PR - Creating a GitHub release with the prepared tag. Note: Since the release tag references the Docker container tagged with the release tag (`ghcr.io/pulp-platform/deeploy:v0.2.0`), the CI will initially fail. The Deeploy Docker image must be built after the release PR is merged and the CI restarted. ### List of Pull Requests - Prepare v0.2.0 release [#102](#102) - Add Luka as Code Owner [#101](#101) - Fix CI, Docker Files, and Documentation Workflow [#100](#100) - Chimera Platform Integration [#96](#96) - Add Tutorial and Refactor README [#97](#97) - Reduce Mean Float Template [#92](#92) - Reshape Memory Freeing and Generic Float GEMM Fixes [#91](#91) - Prepare for Release and Separate Dependencies [#90](#90) - Fix input offsets calculation [#89](#89) - Move PULP SDK to main branch/fork [#88](#88) - Finite Lifetime for IO Tensors [#51](#51) - Improved Memory Visualization and Multi-Layer Tiling Profiling [#56](#56) - Fix Linting in CI and Reformat C Files [#86](#86) - Fix Broken CMake Flow For pulp-sdk [#87](#87) - Refactor Changelog For Release [#85](#85) - ARM Docker Container and Minor Bug Fix [#84](#84) - Added Kernel for Generic Float DW Conv2D [#63](#63) - Autoselect Self-Hosted Runners if the Action is on Upstream [#81](#81) - TEST_RECENT linking on MacOS [#78](#78) - Add RV32IMF Picolibc support for Siracusa platform [#66](#66) - Improve Documentation and VSCode Support [#76](#76) - Debug Print Topology Pass and Code Transformation [#75](#75) - Find all subdirectories of Deeploy when installing with pip install [#70](#70) - Add milestone issue template [#71](#71) - Bunch of fixes and changes [#58](#58) - Add SoftHier platform [#65](#65) - rv32imf_xpulpv2 ISA support for Siracusa platform [#64](#64) - One LLVM To Compile Them All [#60](#60) - One GVSoC to Simulate Them All [#59](#59) - Add Support for CCT Last Layer Training with Embedding Dim 8-128 [#55](#55) - Add CCT Classifier Training Support [#53](#53) - L3 Bugs: DMA Struct Datatype and Maxpool Margin Error [#45](#45) - DeepQuant Quantized Linear Support [#54](#54) - Implemented Dequant Layer for Generic and Siracusa [#52](#52) - Infinite Lifetime Buffers Considered in Tiling & Memory Allocation (+ Visualization) [#44](#44) - Implemented Quant Layer for Generic and Siracusa [#49](#49) - Increase maximal Mchan DMA transfer sizes from 64KiB to 128KiB [#47](#47) - Add MiniMalloc and Decouple Memory Allocation and Tiling [#40](#40) - Float CCT Bugs on L3 [#37](#37) - Memory Allocation Strategies and Visualization [#36](#36) - Add CODEOWNERS [#42](#42) - Add Tiling Support to All CCT Kernels and Fix CCT Operators on Siracusa Platform for L2 [#35](#35) - Add Fp gemm and Softmax for Snitch platform [#31](#31) - Add Float Kernels for CCT [#29](#29) - documentation deployment [#34](#34) - main.c Float Cast Bugs [#28](#28) - Add Float GEMM on PULP with Tiling [#26](#26) - Add Float Support & Float GEMM for Generic [#25](#25) - GVSOC support for the Snitch Cluster platform [#23](#23) - Snitch Cluster Tiling Support [#22](#22) - Snitch support integration [#14](#14) - Update bibtex citation [#20](#20) - the PR template location, bump min python to 3.10, change install command [#17](#17) - Add pre-commit for python formatting [#15](#15) - FP integration (v2) [#12](#12) - shell for sequential tests of Generic, Cortex, and Mempool platforms [#11](#11) - Add issue templates [#10](#10) - Minor CI and Readme Improvements [#8](#8) - Fix GHCR Link for Docker Build [#7](#7) - neureka's ccache id [#6](#6) - GitHub-based CI/CD Flow [#4](#4) - Generic Softmax Kernel [#2](#2) - Port GitLab CI [#1](#1)

… CCT fixtures (pulp-platform#14) * fix(gelu): use exact erf-based GELU forward to match PyTorch nn.GELU() The previous tanh approximation (0.5*x*(1+tanh(sqrt(2/pi)*(x+0.044715*x^3)))) diverges from PyTorch's default nn.GELU(), which uses the exact formula 0.5*x*(1+erf(x/sqrt(2))). Switch to the erf form so that the PULP forward pass is numerically consistent with the reference and with the erf-based GELU backward (GELU_fp32_fp32_sigmoid_grad_chunk). * fix(cct): correct GeluGrad data_in in CCT and CCT-LoRA training fixtures The BiasGeluGrad_dX → GeluGrad optimizer pass was emitting GeluGrad(dY, X) instead of GeluGrad(dY, X+bias). Both cct_train and cct_lora_train vendor the already-optimized network.onnx, so the bug was baked in. Patch both ONNX graphs directly: the Add(X, bias) → X_add_bias forward node already exists in each graph (from run_optmization_remove_biasgelu), so updating the GeluGrad second input to X_add_bias requires no new nodes. Loss comparison (outputs.npz) is unaffected: the reference loss is a forward-pass quantity that does not depend on the backward GeluGrad path.

Victor-Jung requested a review from lukamac November 28, 2024 10:20

Victor-Jung marked this pull request as ready for review November 28, 2024 13:27

Victor-Jung changed the title ~~DRAFT: Snitch support integration~~OPEN: Snitch support integrationNov 28, 2024

Victor-Jung assigned tahaelbayad Nov 28, 2024

Victor-Jung added the Feature Addition of new features label Nov 28, 2024

add snitch support
b663435
add snitch support snitch repo changed snitch repo changed switching to main branch of snitch add snitch and banshee commit hash

tahaelbayad force-pushed the snitch_support_integration branch from d66d7ae to dca8455Compare November 28, 2024 18:58

adding snitch to CI/CD flow
1bb7d83
fixing typo change image in yml files cahnge image in yml files fix typo

tahaelbayad force-pushed the snitch_support_integration branch from dca8455 to 57ca63eCompare November 28, 2024 19:05

linting
da662e1
linting

tahaelbayad force-pushed the snitch_support_integration branch from 57ca63e to da662e1Compare November 28, 2024 19:08

lukamac reviewed Nov 28, 2024
View reviewed changes

Comment threadContainer/Dockerfile Outdated
Comment threadMakefile Outdated
Comment threadTargetLibraries/Snitch/CMakeLists.txt Outdated
Comment threadTargetLibraries/Snitch/cmake/snitch-runtime-precompiled.cmake
Comment threadTargetLibraries/Snitch/scripts/run_clang_format.py Outdated

tahaelbayad added 3 commits December 2, 2024 09:26

apply Luka's comment
7f8fc3e

apply Luka's comment
6bd719e

apply Luka's comment
f87c6b5

Xeratec reviewed Dec 4, 2024
View reviewed changes

tahaelbayad force-pushed the snitch_support_integration branch from f7cfa7f to f87c6b5Compare December 5, 2024 13:49

apply snitch patch and removing old snitch cmake scripts
1877fe9

lukamac reviewed Dec 10, 2024
View reviewed changes

Comment threadContainer/Dockerfile Outdated
Comment threadContainer/Dockerfile Outdated
Comment threadContainer/Dockerfile Outdated
Comment threadTargetLibraries/Snitch/CMakeLists.txt Outdated

tahaelbayad added 5 commits December 11, 2024 12:14

apply Luka's comment
5c70d65

apply Luka's comment
532c493

apply Luka's comment
d223dc5

apply Luka's comment
cd01169

change nodeRep to operatorRepresentation
1ac4c9a

Victor-Jung reviewed Dec 12, 2024
View reviewed changes

Comment threadDeeployTest/Platforms/Snitch/main.c

tahaelbayad mentioned this pull request Dec 12, 2024
[BUG] Unexpected warnings when removing compilation flags #21
Closed

tahaelbayad force-pushed the snitch_support_integration branch from 357ee6c to 1ac4c9aCompare December 12, 2024 13:18

remove snitch compile flags
98fc59d

Victor-Jung approved these changes Dec 12, 2024
View reviewed changes

changing CHANGELOG and docker's tag
7cde9a7

Victor-Jung merged commit f69868b into pulp-platform:develDec 12, 2024

Xeratec mentioned this pull request Jul 8, 2025
Release v0.2.0 #103
Merged

Navigation Menu

Uh oh!

OPEN: Snitch support integration - #14

OPEN: Snitch support integration#14
Victor-Jung merged 14 commits into
pulp-platform:develfrom
tahaelbayad:snitch_support_integration

tahaelbayad commented Nov 15, 2024

Uh oh!

lukamac left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

XeratecDec 3, 2024

Uh oh!

Victor-JungDec 5, 2024•
edited
Loading

Uh oh!

XeratecDec 5, 2024

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Uh oh!

Conversation

tahaelbayad commented Nov 15, 2024

Added

Uh oh!

lukamac left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

XeratecDec 3, 2024

Choose a reason for hiding this comment

Uh oh!

Victor-JungDec 5, 2024• edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

XeratecDec 5, 2024

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Victor-JungDec 5, 2024•
edited
Loading