Update unit tests status badge in readme#1
Merged
Conversation
khatwanimohit
force-pushed
the
mohit/fix-readme
branch
from
bee760d to
86f442f
Compare
rwitten
approved these changes
Apr 18, 2023
Collaborator
rwitten
left a comment
rwitten
left a comment
There was a problem hiding this comment.
Congrats on the first commit!
Merged
4 tasks
copybara-service Bot
pushed a commit
that referenced
this pull request
Jun 18, 2025
JAX is undergoing a rename of the contents of `jax.experimental.layouts` in preparation for its graduation from experimental: 1. "Formats" are replacing "layouts", and specifically `Layout` -> `Format` 2. "Layouts" are replacing "device local layouts", and specifically `DeviceLocalLayout` -> `Layout` This is an incremental update carrying out #1. PiperOrigin-RevId: 773086710
copybara-service Bot
pushed a commit
that referenced
this pull request
Jun 19, 2025
…s` and `.output_layouts` respectively. JAX is undergoing a rename of the contents of jax.experimental.layouts in preparation for its graduation from experimental: "Formats" are replacing "layouts", and specifically Layout -> Format "Layouts" are replacing "device local layouts", and specifically DeviceLocalLayout -> Layout This is an incremental update carrying out #1. PiperOrigin-RevId: 773141876
copybara-service Bot
pushed a commit
that referenced
this pull request
Jun 19, 2025
JAX is undergoing a rename of the contents of jax.experimental.layouts in preparation for its graduation from experimental: "Formats" are replacing "layouts", and specifically Layout -> Format "Layouts" are replacing "device local layouts", and specifically DeviceLocalLayout -> Layout This is an incremental update carrying out #1. PiperOrigin-RevId: 773163981
![github-actions[bot]](https://avatars.githubusercontent.com/in/15368?s=60&v=4){kind=small}
There was a problem hiding this comment.
📋 Review Summary
This pull request updates the URL for the unit tests status badge in the README.md file. The change correctly points the badge to the google/maxtext repository, which appears to be the upstream source, ensuring the test status is accurate.
🔍 General Feedback
- This is a good maintenance update that improves the quality of the project's documentation.
There was a problem hiding this comment.
📋 Review Summary
I was unable to perform a code review for this pull request. The tools required to fetch the pull request data, such as mcp__github__get_pull_request_diff, were not available in my execution environment.
🔍 General Feedback
- The instructions provided for my operation are contradictory. They refer to data-gathering tools (
mcp__github__get_pull_request,mcp__github__get_pull_request_files,mcp__github__get_pull_request_diff) that are not present in the list of available tools. - My attempts to work around this limitation by using shell commands to find a diff file were blocked by security restrictions.
- Without access to the code changes, I cannot provide any feedback on correctness, security, or other review criteria.
Please check the configuration of the code review agent and ensure that all necessary tools are available and that the instructions are consistent with the execution environment.
There was a problem hiding this comment.
📋 Review Summary
This pull request appears to intend to make train.py importable, but the primary change has the opposite effect. The review identifies a critical issue that prevents safe importation of the module.
🔍 General Feedback
- The goal of making
train.pyimportable is good for code structure and reusability. - However, the current implementation will cause side effects when the module is imported. My comment includes a suggested fix for this critical issue.
amitsrivastava78
added a commit
to amitsrivastava78/maxtext
that referenced
this pull request
Feb 5, 2026
OPTIMIZATION RESULTS from comprehensive TPU testing: KEY FINDING AI-Hypercomputer#1: block_kv_compute=64 (not 128!) - Config: block_q=1024, block_kv_sparse=256, block_kv_compute=64 - Result: 1.095× speedup (9.5% FASTER than JAX) - Previous: 0.916× (9.2% slower) - Improvement: +19.5% performance gain KEY FINDING AI-Hypercomputer#2: Scales exceptionally well on large inputs - Config: 64 heads, 2048 q_len, 512 sparse_len - Result: 3.306× speedup on larger problems! - Shows kernel is production-ready for real workloads UPDATED: kascade_kernel.py defaults to use block_kv_compute=64 STATUS: ✅ EXCEEDS 1.2× threshold for production integration Next: Integrate into MaxText attention layer
geeningwang
pushed a commit
to geeningwang/maxtext
that referenced
this pull request
Apr 17, 2026
- Benchmark history: add row AI-Hypercomputer#7 (scan=true, 71.1 ms / 450 tok/s, +28% vs AI-Hypercomputer#6) - Add 'scan_layers=true Analysis' section explaining the 15.4 ms overhead: root cause is loss of XLA inter-layer weight-prefetch pipelining (lax.while_loop prevents cross-iteration scheduling); ~0.32 ms/layer consistent with memory-bandwidth-bound workload losing prefetch overlap - Quantify sparse dispatch break-even bar: must recover >22% to beat 55.7 ms dense baseline; rough estimate ~34 ms achievable with ragged_all_to_all - Update 'Most Impactful Next Fixes' section: scan fix done, ragged_all_to_all is now AI-Hypercomputer#1 priority; update HEAD ref to 539cc04 - Re-rank optimisation table: ragged_all_to_all moved to rank 2 (highest unrun), scan_layers=true added as rank 3 (done, prerequisite)
geeningwang
pushed a commit
to geeningwang/maxtext
that referenced
this pull request
Apr 20, 2026
Key insight: AR decode at batch=32/T=32 is weight-bandwidth-bound. Sparse MoE dispatch does not help (proven: 101.5ms, 83% regression). New ranking: AI-Hypercomputer#1 Batch size > 1 (near-linear throughput scaling) AI-Hypercomputer#2 Int8/FP8 weight quantisation (halves dominant cost) AI-Hypercomputer#3 Speculative decoding AI-Hypercomputer#4 Remove effects_barrier / async EOS Demoted: ragged_all_to_all for decode → bottom (failed) scan_layers=true + 4-phase decoder → AI-Hypercomputer#12 (prefill prereq only) Re-scoped ragged_all_to_all as prefill-only optimisation (valid at T=512 where intermediates 2.68 GB/layer dominate).
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Update unit tests status badge in readme