Fy/versioning #1
Conversation
- Add Docusaurus versioning infrastructure - Create version snapshots for 0.8.0 (stable) and 0.7.0 - Set up Latest (dev) documentation track - Add GitHub Actions workflows for automated doc syncing - Add custom version banner components - Update configuration and sidebar structure
There was a problem hiding this comment.
Pull Request Overview
This PR adds versioned documentation for ServerlessLLM versions 0.8.0 and 0.7.0. The changes include complete documentation sets for both versions covering getting started guides, deployment instructions, feature documentation, and store-specific guides.
Key Changes:
- Added version tracking in
versions.jsonfor versions 0.8.0 and 0.7.0 - Created versioned sidebar configurations for both versions
- Added comprehensive documentation including quickstart guides, deployment guides, feature documentation, and ServerlessLLM Store documentation
Reviewed Changes
Copilot reviewed 65 out of 92 changed files in this pull request and generated 6 comments.
Show a summary per file
| File | Description |
|---|---|
| versions.json | Defines available documentation versions |
| versioned_sidebars/version-0.8.0-sidebars.json | Sidebar navigation for v0.8.0 docs |
| versioned_sidebars/version-0.7.0-sidebars.json | Sidebar navigation for v0.7.0 docs |
| versioned_docs/version-0.8.0/store/*.md | Store documentation including quickstart, ROCm guide, and quantization |
| versioned_docs/version-0.8.0/*.md | Core documentation including intro, getting started, and features |
| versioned_docs/version-0.8.0/images/* | Image assets for documentation |
| docs/store/category.json | Category configuration for store documentation |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
|
|
||
| Our api aims to be compatible with the `sharded_state` load format in vLLM. Thus, due to the model modifications about the model architecture done by vLLM, the model format for vLLM is **not** the same as we used in transformers. Thus, the `ServerlessLLM format` mentioned in the subsequent sections means the format integrated with vLLM, which is different from the `ServerlessLLM format` used in the previous sections. | ||
|
|
||
| Thus, for fist-time users, you have to load the model from other backends and then converted it to the ServerlessLLM format. |
There was a problem hiding this comment.
Corrected spelling of 'fist-time' to 'first-time'.
|
|
||
| Our api aims to be compatible with the `sharded_state` load format in vLLM. Thus, due to the model modifications about the model architecture done by vLLM, the model format for vLLM is **not** the same as we used in transformers. Thus, the `ServerlessLLM format` mentioned in the subsequent sections means the format integrated with vLLM, which is different from the `ServerlessLLM format` used in the previous sections. | ||
|
|
||
| Thus, for fist-time users, you have to load the model from other backends and then converted it to the ServerlessLLM format. |
There was a problem hiding this comment.
Verb tense inconsistency: 'converted' should be 'convert' to match 'load'.
| print(tokenizer.decode(outputs[0], skip_special_tokens=True)) | ||
| ``` | ||
|
|
||
| 4. Clean up by "Ctrl+C" the server process. |
There was a problem hiding this comment.
Inconsistent formatting: Use backticks for the instruction. Should be: 'Clean up by pressing Ctrl+C on the server process.' or 'Clean up by stopping (Ctrl+C) the server process.'
| print(tokenizer.decode(outputs[0], skip_special_tokens=True)) | ||
| ``` | ||
|
|
||
| 4. Clean up by `Ctrl+C` the server process. |
There was a problem hiding this comment.
Inconsistent with Step 4 later in the document which uses proper formatting: 'Clean up by Ctrl+C the server process.' The later version correctly states 'Clean up by pressing Ctrl+C...'
| model = AutoModelForCausalLM.from_pretrained('facebook/opt-1.3b', torch_dtype=torch.float16) --> | ||
|
|
||
| # Replace './models' with your local path. | ||
| save_lora(adapter, './models/facebook/opt-1.3b') |
There was a problem hiding this comment.
Missing explanation for where the 'adapter' variable comes from. The TODO comment above indicates this section is incomplete.
| import clsx from 'clsx'; | ||
| import useDocusaurusContext from '@docusaurus/useDocusaurusContext'; | ||
| import Link from '@docusaurus/Link'; | ||
| import Translate from '@docusaurus/Translate'; |
There was a problem hiding this comment.
Unused import Translate.
This pr adds versioning for the document site.
This pr also creates two snapshots (0.7 and 0.8) as past and stable document.