Docs Content Part 1: Homepage and getting started #448
Conversation
…rehensive getting started with dependency details, and expanded concepts page with Monarch, Services, TorchStore, and RL workflows
Co-authored-by: Svetlana Karslioglu <[email protected]>
Co-authored-by: Svetlana Karslioglu <[email protected]>
Co-authored-by: Svetlana Karslioglu <[email protected]>
Co-authored-by: Svetlana Karslioglu <[email protected]>
Co-authored-by: Svetlana Karslioglu <[email protected]>
Co-authored-by: Svetlana Karslioglu <[email protected]>
- Enhanced homepage with Monarch foundation emphasis, technology stack highlights, validated examples, and clear navigation paths - Expanded getting started with detailed dependency explanations (Monarch, vLLM, TorchTitan, TorchStore, PyTorch Nightly) - Converted installation and verification steps to numbered lists for better readability - Removed FAQ references as FAQ page has been removed - Fixed GPU/process terminology in code examples
meta-cla
bot
added
the
CLA Signed
AlannaBurke
force-pushed
the
docs/pr1-homepage-getting-started-only
branch
from
5151f64 to
f9b136a
Compare
|
Need to clean up those references to usage.md: https://github.com/meta-pytorch/forge/actions/runs/18579946858/job/52972683886?pr=448#step:11:103 |
| conda activate forge | ||
| ``` | ||
|
|
||
| 3. **Run Installation Script** |
There was a problem hiding this comment.
this may be ok for now, but even possibly as soon as EOD today we may have different instructions cc @joecummings
If we keep a script, what the script does will be different. I think we can ship this for now, and update this once we're done
| Fine-tune Llama 3 8B on your data. **Requires: 2+ GPUs** | ||
|
|
||
| 1. **Download the Model** | ||
|
|
There was a problem hiding this comment.
@ebsmothers @daniellepintz - could you two please review these commands for SFT and ensure this is what we want?
There was a problem hiding this comment.
@AlannaBurke let's use this command
hf download meta-llama/Meta-Llama-3.1-8B-Instruct --local-dir /tmp/Meta-Llama-3.1-8B-Instruct --exclude "original/consolidated.00.pth"
There was a problem hiding this comment.
Actually there is no need to download the model anymore, we simplified that in the configs. users just need to run this cmd once they have acc to the model:
python -m apps.sft.main --config apps/sft/llama3_8b.yaml
this is an easier workflow for users
| - Steps to reproduce | ||
| - Expected vs actual behavior | ||
|
|
||
| **Diagnostic command:** |
There was a problem hiding this comment.
this is a really good idea - let's keep this, and I think we should come up with a script for this in our issue templates..
cc @joecummings @daniellepintz ? not sure who to tag here
| Fine-tune Llama 3 8B on your data. **Requires: 2+ GPUs** | ||
|
|
||
| 1. **Download the Model** | ||
|
|
There was a problem hiding this comment.
@AlannaBurke let's use this command
hf download meta-llama/Meta-Llama-3.1-8B-Instruct --local-dir /tmp/Meta-Llama-3.1-8B-Instruct --exclude "original/consolidated.00.pth"
4 participants
I split the docs work into 2 PRs, this is the first. Let's get this merged ASAP.