Skip to content

Archive RPAv2 and Delete Torchax-pt path. #504

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Aug 27, 2025
Merged

Archive RPAv2 and Delete Torchax-pt path. #504

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 0 additions & 22 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -113,7 +113,6 @@ To switch different backends (default is jax):

```
TPU_BACKEND_TYPE=jax
TPU_BACKEND_TYPE=torchax
TPU_BACKEND_TYPE=pytorch_xla
```

Expand Down Expand Up @@ -209,27 +208,6 @@ docker run \
--max_model_len=1024 \
```

## Torchax Guide

**NOTE**: This is under development so the run may fail.

### Install dependencies

#### Install `vLLM`

Follow the above [step](#install-vllm-tpu) to install vllm for TPU backend.

#### Install `tpu_commons`

Follow the above step to install [tpu_commons](#install-tpu_commons)

### Run example script

```
cd vllm
TPU_BACKEND_TYPE=torchax VLLM_TORCHAX_ENABLED=1 VLLM_USE_V1=1 python examples/offline_inference/tpu.py
```

## How to test kernel?

Install dependencies:
Expand Down
126 changes: 0 additions & 126 deletions tests/kernels/flash_attention_kernel_test.py
View file
Open in desktop

This file was deleted.

3 changes: 2 additions & 1 deletion tests/kernels/ragged_kv_cache_update_test.py → ...kernels/ragged_kv_cache_update_v2_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,7 +6,8 @@
from jax.sharding import Mesh, NamedSharding
from jax.sharding import PartitionSpec as P

from tpu_commons.kernels.ragged_kv_cache_update import kv_cache_update
from tpu_commons.kernels.ragged_paged_attention.v2.ragged_kv_cache_update import \
kv_cache_update


def kv_cache_update_ref(new_kv, slot_mapping, kv_cache):
Expand Down
Loading
Loading