Update TorchAO README inference section before PTC #3206
Merged
+57
−98
Conversation
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
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/3206
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Active SEVsThere are 1 currently active SEVs. If your PR is affected, please view them below: This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
bot
added
the
CLA Signed
jerryzh168
added
topic: documentation
jerryzh168
force-pushed
the
update-readme-10-2025
branch
from
b0fc829 to
081118f
Compare
meta-cla
bot
added
the
CLA Signed
jerryzh168
commented
Oct 17, 2025
| from torchao.quantization import Int4WeightOnlyConfig, quantize_ | ||
| quantize_(model, Int4WeightOnlyConfig(group_size=32, version=1)) | ||
| ``` | ||
| Compared to a `torch.compiled` bf16 baseline, your quantized model should be significantly smaller and faster on a single A100 GPU: |
There was a problem hiding this comment.
removing these since toy model memory/latency is not meaningful, to make our README shorter
jerryzh168
commented
Oct 17, 2025
|
|
||
| TorchAO is integrated into some of the leading open-source libraries including: | ||
|
|
||
| * HuggingFace transformers with a [builtin inference backend](https://huggingface.co/docs/transformers/main/quantization/torchao) and [low bit optimizers](https://github.com/huggingface/transformers/pull/31865) |
There was a problem hiding this comment.
reordered a bit to put more commonly used ones earlier
jerryzh168
force-pushed
the
update-readme-10-2025
branch
from
081118f to
d45a249
Compare
supriyar
reviewed
Oct 17, 2025
jerryzh168
force-pushed
the
update-readme-10-2025
branch
from
d45a249 to
f7762ad
Compare
andrewor14
approved these changes
Oct 20, 2025
docs/source/quick_start.rst
Outdated
| command instead:: | ||
|
|
||
| pip install --pre torchao --index-url https://download.pytorch.org/whl/nightly/cu121 | ||
| pip install --pre torchao --index-url https://download.pytorch.org/whl/nightly/cu126 |
There was a problem hiding this comment.
should this be cu128 or cu129?
There was a problem hiding this comment.
sure, I can change to 128
| * Integration with [FBGEMM](https://github.com/pytorch/FBGEMM/tree/main/fbgemm_gpu/experimental/gen_ai) for SOTA kernels on server GPUs | ||
| * Integration with [ExecuTorch](https://github.com/pytorch/executorch/) for edge device deployment | ||
| * Axolotl for [QAT](https://docs.axolotl.ai/docs/qat.html) and [PTQ](https://docs.axolotl.ai/docs/quantize.html) | ||
| * TorchTitan for [float8 pre-training](https://github.com/pytorch/torchtitan/blob/main/docs/float8.md) |
There was a problem hiding this comment.
should we add unsloth too? Or are we still waiting for the blog post to link?
There was a problem hiding this comment.
yeah I was waiting for you to add this
| ### PyTorch-Native Training-to-Serving Model Optimization | ||
| - Pre-train Llama-3.1-70B **1.5x faster** with float8 training | ||
| - Recover **77% of quantized perplexity degradation** on Llama-3.2-3B with QAT | ||
| - Quantize Llama-3-8B to int4 for **1.89x faster** inference with **58% less memory** |
There was a problem hiding this comment.
can you also update this Jerry?
There was a problem hiding this comment.
yeah I'm not sure about the latest, was planning for someone more familiar to update this
jerryzh168
force-pushed
the
update-readme-10-2025
branch
from
f7762ad to
6fdadba
Compare
jainapurva
reviewed
Oct 20, 2025
docs/source/serving.rst
Outdated
|
|
||
| .. code-block:: bash | ||
| pip install vllm --pre --extra-index-url https://wheels.vllm.ai/nightly |
There was a problem hiding this comment.
@jerryzh168 Should we update this to use pytorch's vllm build: https://download.pytorch.org/whl/nightly/vllm ?
jerryzh168
force-pushed
the
update-readme-10-2025
branch
from
6fdadba to
8e4e53f
Compare
Summary: att Test Plan: visual inspection Reviewers: Subscribers: Tasks: Tags:
jerryzh168
force-pushed
the
update-readme-10-2025
branch
from
8e4e53f to
30f2dd8
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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.
Summary:
att
Test Plan:
visual inspection
Reviewers:
Subscribers:
Tasks:
Tags: