Code for Life
First-year PhD student at CMU / Undergrad from ACM Class, SJTU / ML Systems / Deep Learning Compilers / @apache TVM Committer
-
Carnegie Mellon University
- Pittsburgh, United States
- https://ruihanglai.com
Highlights
- Pro
Block or Report
Block or report MasterJH5574
Report abuse
Contact GitHub support about this user’s behavior. Learn more about reporting abuse.
Report abusePinned
-
apache/tvm Public
Open deep learning compiler stack for cpu, gpu and specialized accelerators
-
-
mlc-ai/web-stable-diffusion Public
Bringing stable diffusion models to web browsers. Everything runs inside the browser with no server support.
-
1,291 contributions in the last year
Less
More
Activity overview
Contribution activity
March 2023
Created 90 commits in 5 repositories
Created 1 repository
Created a pull request in apache/tvm that received 3 comments
[Unity][Transform] LiftTransformParams handling multiple functions
Previously, the LiftTransformParams pass only works on function "main". This is a bit restrictive as in our recent practice on stable diffusion, th…
+124
−18
•
3
comments
Opened 26 other pull requests in 4 repositories
apache/tvm
17
merged
- [Unity][Web] WebGPU explicit max buffer size
- [Unity][WEB] Update text prompts for syntactical correctness
- [Unity][Transform] Memory plan across the IRModule
- [Unity][Frontend] from_fx keeps parameters in order
- [Unity][Frontend] FX translator supports unwrapping unit return tuple
- [Unity][Frontend] Attach imported model weights, deprecate ImporterOutput
- [Unity][TIR][Pass] ForceNarrowIndexToInt32
- [Unity][Frontend] FX translator support torch.baddbmm
- [Unity][Fix] FX translating dtype
- [Fix][TIR] LowerCrossThreadReduction with write-back predicate
-
[Unity][Op] Legalize
round,floor,ceil,sign -
[Unity][Frontend] FX translator returning weights with
keep_params_as_input - [Unity][Frontend] FX translator supporting more ops
- [Unity][Op] Argmax and argmin
- [Unity][Op] Group normalization
- [TOPI] Group normalization
- [HotFix][MetaSchedule] Turn off database shash check
mlc-ai/web-stable-diffusion
6
merged
mlc-ai/relax
2
closed
mlc-ai/utils
1
merged
Reviewed 26 pull requests in 3 repositories
apache/tvm
20 pull requests
- [Unity][Transform] Enhance RewriteDataflowReshape transform
- [TIR][Schedule] Allow buffer name argument to Schedule.set_scope
- [Unity][Transform] Automatic Layout Conversion
- [Unity][Op] Enable special dimension value 0 in reshape
- [Unity][DEBUG] Add Instrument
-
[Unity] Fix StructInfo Infer for
vm.alloc_tensor -
[Unity][Frontend] Import
tanhand fixlayer_norm - [Unity][Transform] Let MetaScheduleApplyDatabase use DefaultSchedule when no record is found.
- [Unity][WEB] Improve ndarray cache
- [Unity] Improve implementation of FuseOps
- [Unity] ensure memory.alloc_tensor/storage roundtrippable
- [Unity][WEB] Simplify WebGPU Codegen per spec
- [MetaSchedule] preseve global_symbol attached to function after applying MS
- [Unity][WEB] Improve webgpu codegen options to skip readonly
- [TOPI] Batch Norm Training Mode
- [Unity] Introduce call_dps_packed
- [Unity] Introduce Default GPU Schedule Pass
- [Unity][WEBGPU] Codegen improvements and WebRuntime
-
[Unity][Op] Full support of Relax op
power - [Unity][Debugging] AST printer
mlc-ai/web-stable-diffusion
3 pull requests
mlc-ai/relax
3 pull requests
20
contributions
in private repositories
Mar 1 – Mar 14




