Mlp example#172

Draft

jianyicheng wants to merge 32 commits intomainfrom

Commits on Apr 24, 2024

Updated Docker to main for local build
Jianyi Cheng
committed
Sync scripts for updated docker setups (seperating CPU and GPU containers)
Jianyi Cheng
committed
removed docker as submodule
Jianyi Cheng
committed
Pull from Makefile to avoid repeated sync on submodule
Jianyi Cheng
committed
Added instructions to log for better readability on CI log
Jianyi Cheng
committed
Initially seperate dataflow sv out of top
Jianyi Cheng
committed
Split two files initially
Jianyi Cheng
committed
Refactored dataflow level
Jianyi Cheng
committed
Added missing dependences and updated file paths
Jianyi Cheng
committed
Rafactored memory map emit
Jianyi Cheng
committed
Added rounding at the output of relu
Jianyi Cheng
committed
Merge branch 'main' into mlp_example
Jianyi Cheng
committed

Commits on Apr 25, 2024

Fixed relu var name typos
Jianyi Cheng
committed
Updated linear layer with the right parameter names
Jianyi Cheng
committed
Remove temporary code for parallelism
Jianyi Cheng
committed
Added missing component interface
Jianyi Cheng
committed
removed extra comma in parameter map
Jianyi Cheng
committed

Commits on Apr 26, 2024

Updated parallelism parameter formats
Jianyi Cheng
committed
Updated bram width calculation with the latest parallelism parameters
Jianyi Cheng
committed
Fix quantization meta data for fixed-point quantization (#173)

ChengZhang-98
and
Jianyi Cheng
authored
reverted changes made by the quantization PR
Jianyi Cheng
committed
reduced test case
Jianyi Cheng
committed

Commits on Apr 28, 2024

Fetched previous verilog analysis pass for testing
Jianyi Cheng
committed
Fetched the latest version of draft
Jianyi Cheng
committed
Pass syntax error in Python
Jianyi Cheng
committed
refactored test pass format
Jianyi Cheng
committed
format quantize PR https://github.com/DeepWok/mase/pull/173
Jianyi Cheng
committed
Added bit truncation in bram param to avoid verilator warnings
Jianyi Cheng
committed
Fixed bitwidth error (this is temporary for the version of casting inside hardware components
Jianyi Cheng
committed
Fixed minor parallelism parameter shapes
Jianyi Cheng
committed
Get working flow for hardware testing - but need to check hardware results
Jianyi Cheng
committed

Commits on Apr 29, 2024

Mlp quantization error (#178)

Jianyi Cheng
and
ChengZhang-98
authored