|
About the MAX category
|
|
0
|
148
|
December 4, 2024
|
|
Help with max serve performance on H100
|
|
6
|
46
|
April 14, 2026
|
|
Overhead of Multi‑Dimensional Reduction
|
|
1
|
68
|
March 10, 2026
|
|
Constraint failed: copy_from_async only allows 4, 8, 16 bytes element
|
|
0
|
31
|
February 25, 2026
|
|
Help compiling MLP in MAX
|
|
4
|
80
|
February 23, 2026
|
|
Initial support for running MAX models on AMD RDNA GPUs
|
|
0
|
93
|
February 17, 2026
|
|
MAX models can now use customized Mojo kernels and standard library
|
|
0
|
79
|
February 17, 2026
|
|
Question on register custom ops for max on different device
|
|
6
|
51
|
February 13, 2026
|
|
Failed to resolve module path for `MOGGKernelAPI`
|
|
1
|
53
|
February 10, 2026
|
|
Build an LLM in MAX from scratch đź“–
|
|
10
|
646
|
February 6, 2026
|
|
All MAX API tests can now be run via Bazel in the `modular` repository
|
|
7
|
177
|
February 6, 2026
|
|
Mypy type checking workflow for new model development
|
|
3
|
79
|
January 30, 2026
|
|
Using @F.functional decorator for forward method
|
|
2
|
101
|
January 23, 2026
|
|
MAX Graph Python API built-in ops fail to compile for GPU - what's the correct pattern?
|
|
3
|
59
|
January 19, 2026
|
|
Model Correctness Check
|
|
1
|
57
|
January 14, 2026
|
|
Tensor Creation Allocates Memory on All GPUs and Fails with CUDA_VISIBLE_DEVICES
|
|
4
|
82
|
January 14, 2026
|
|
Slow tensor creation with experimental Tensor
|
|
2
|
75
|
January 12, 2026
|
|
Encoder-Decoder (T5) model serving support
|
|
4
|
175
|
January 4, 2026
|
|
Sparse Matrix/Array Formats for Roadmap Consideration 🙏
|
|
8
|
271
|
December 28, 2025
|
|
DeepSeek-OCR infer at CPU
|
|
4
|
195
|
December 22, 2025
|
|
Using max from another bazel project
|
|
1
|
68
|
December 7, 2025
|
|
MAX can now be used in Google Colab
|
|
4
|
606
|
December 4, 2025
|
|
After startup MAX server, curl access feedback "curl: (56) Recv failure: Connection reset by peer"
|
|
0
|
41
|
November 19, 2025
|
|
Conv2d error with non-unit dilation in MAX
|
|
1
|
75
|
November 5, 2025
|
|
Inference performance issue
|
|
6
|
306
|
October 31, 2025
|
|
Support GPTQ quant in MoE layer
|
|
0
|
89
|
October 26, 2025
|
|
[kernel] convolution on gpu
|
|
1
|
109
|
October 18, 2025
|
|
Open-sourcing all of the MAX Python API
|
|
0
|
222
|
October 16, 2025
|
|
Open source the max.{driver,dtype,engine,experimental,interfaces,mlir,profiler,support,torch} modules
|
|
3
|
250
|
April 13, 2026
|
|
MAX kernel launch overhead
|
|
2
|
129
|
March 22, 2026
|