Migrate documentation into docs/ with full history#1321

Merged

Zhichenzzz merged 346 commits into

mainfrom

docs/import-mintlify-docs

Jun 12, 2026

This pull request is big! We're only showing the most recent 250 commits

Commits on Apr 29, 2026

Commits on Apr 30, 2026

Merge pull request #52 from radixark/shi/fix-sample-filter-doc
Zhichenzzz
authored
Merge pull request #51 from radixark/shi/fix-rollout-fn-doc
Zhichenzzz
authored
docs(deepseek): replace V4 placeholder with full launch tutorial
Shi Dong
committed
docs(customization): widen --buffer-filter-path rollout_id annotation
Shi Dong
committed
docs(deepseek): address PR review on V4-Flash/V4-Pro split
Shi Dong
committed
docs(deepseek): tighten V4 quirks list and fix active-param counts
Shi Dong
committed
Merge pull request #54 from radixark/shi/fix-buffer-filter-annotation
Zhichenzzz
authored
docs(home): split Supported Hardware into NVIDIA / AMD bullets
Shi Dong
committed
docs(home): rename Latest-updates link to match CLI Reference
Shi Dong
committed
docs(home): align all Latest-updates link texts with target page titles
Shi Dong
committed
docs(blog): drop Try-it code block; defer to Quick Start
Shi Dong
committed
docs(blog): apply reviewer wording to Try-it paragraph
Shi Dong
committed
docs(contributing): refresh Repository Layout against current miles tree
Shi Dong
committed
docs(advanced): sync p2p-weight-transfer with miles upstream
Zhichenzzz
committed
Merge pull request #55 from radixark/shi/split-supported-hw
Zhichenzzz
authored
Merge pull request #56 from radixark/shi/fix-cli-reference-link
Zhichenzzz
authored
Merge pull request #57 from radixark/shi/polish-introducing-miles-try-it
Zhichenzzz
authored
Merge pull request #58 from radixark/shi/repo-layout-refresh
Zhichenzzz
authored
Merge pull request #59 from radixark/zhichen/sync-p2p-weight-transfer-doc
Zhichenzzz
authored
docs(models): add Qwen3.6 dense and MoE recipes
Zhichenzzz
committed
docs(models): add Nemotron-3-Nano dense and MoE recipes
Zhichenzzz
committed

Commits on May 1, 2026

Commits on May 8, 2026

docs: expand low-precision feature bullet (#65)

yueming-yuan
and
Shi Dong
authored
docs: highlight Efficient R3 with cache + D2H overlap (#66)

yueming-yuan
and
Shi Dong
authored
docs: add fast-and-stable model-support bullet to core features (#67)
yueming-yuan
authored
docs: add broad hardware support bullet (#68)

yueming-yuan
and
Shi Dong
authored
docs(index): align Supported models lists with Models page (#64)
Zhichenzzz
authored
docs: rewrite low-precision page (FP8 / MXFP8 / NVFP4) (#70)

yueming-yuan
and
Shi Dong
authored
docs: convert British '-ise/-isation' to American '-ize/-ization' (#71)

Shi-Dong
and
Shi Dong
authored
docs(installation): clarify H100/H200 status as 'CI guarded' (#72)

Shi-Dong
and
Shi Dong
authored
docs(usage): make parse_args import copyable as a code block (#73)

Shi-Dong
and
Shi Dong
authored
docs(usage): add Megatron parallelism compatibility section (#74)

Shi-Dong
and
Shi Dong
authored
docs: add Argument Groups page and linkify launch-script tables (#75)

Shi-Dong
and
Shi Dong
authored
docs: add Fully Async Rollout user-guide page (#76)

Shi-Dong
and
Shi Dong
authored
docs(index): replace Supported models prose with two-panel matrix (#77)

Shi-Dong
and
Shi Dong
authored
docs(deepseek): replace fictional MILES_SCRIPT_* env vars with real launcher defaults
Shi Dong
committed
docs(deepseek): correct Key highlights against DeepSeek-V4 model card
Shi Dong
committed
docs(deepseek): address PR draft-review feedback on §1 + §3.1
Shi Dong
committed
docs(deepseek): drop /cluster_public mention from §4.2 multi-node fan-out
Shi Dong
committed
docs(deepseek): address follow-up PR review comments
Shi Dong
committed
docs(usage): make page Megatron-only; move FSDP to developer/experimental-features
yueming-yuan
committed

Commits on May 9, 2026

Add Mintlify docs.json config

unseenmars
and
claude
committed
Fix docs.json: add required theme field

unseenmars
and
claude
committed
Fix docs.json: correct navigation format and navbar fields

unseenmars
and
claude
committed
docs(experimental): fix FSDP backend path
Shi Dong
committed
docs(deepseek): full DeepSeek V4 Flash launch tutorial (#53)

Shi-Dong
and
Shi Dong
authored
docs(deepseek): add GB300 docker image for V4-Flash (#62)

authored
docs(v4-flash): lead with full-train quick start; fold env setup + multi-node into Script breakdown

yueming-yuan
authored and
Shi Dong
committed
docs(v4-flash): nest Launcher path defaults under Quick start; add docker pull, drop GB300 placeholder

yueming-yuan
authored and
Shi Dong
committed
docs(v4-flash): document --model-local-dir as multi-node fan-out destination

yueming-yuan
authored and
Shi Dong
committed
docs(v4-flash): drop GB300 7-node parallelism row

yueming-yuan
authored and
Shi Dong
committed
docs(v4-flash): drop Miles Router and Fault Tolerance bullets from Pairs Well With

yueming-yuan
authored and
Shi Dong
committed
docs(v4-flash): note custom TP/EP/PP/CP combos are allowed beyond the validated layouts

yueming-yuan
authored and
Shi Dong
committed
docs(v4-flash): drop _get_parallel_config preamble before parallelism table

yueming-yuan
authored and
Shi Dong
committed
docs(v4-flash): rename Recipe Configuration to Example Recipe Configuration

yueming-yuan
authored and
Shi Dong
committed
docs(v4-flash): move Notable quirks from §5.5 to §4.4 (end of Script breakdown)

yueming-yuan
authored and
Shi Dong
committed

Commits on May 10, 2026

docs(index): add LoRA training-and-serving bullet to Core features
Shi Dong
committed

Commits on May 11, 2026

Commits on May 12, 2026

Commits on May 13, 2026

Commits on May 17, 2026

docs(lora): drop the 'This page is a stub' sentence (#124)
Shi-Dong
authored

Commits on May 20, 2026

docs(lora): four polish fixes from issue #125 (#126)
Shi-Dong
authored

Commits on Jun 8, 2026

Commits on Jun 11, 2026

Commits on Jun 12, 2026