[android] Add support for float16 tensor #15479
Merged
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
The Android binding exposes helpers for feeding IEEE-754 half-precision (FP16) inputs directly. Use `Tensor.fromBlob(shortArray, shape)` or reuse a direct `ShortBuffer` created via `Tensor.allocateHalfBuffer(numElements)` to avoid extra copies: ```kotlin val shape = longArrayOf(24, 4096) val halfData: ShortArray = buildHalfEncodedData() val tensor = Tensor.fromBlob(halfData, shape) val buffer = Tensor.allocateHalfBuffer(halfData.size) buffer.put(halfData) buffer.rewind() val tensorNoCopy = Tensor.fromBlob(buffer, shape) ``` All buffers must be direct and use the native byte order; the helper above takes care of this.
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/15479
Note: Links to docs will display an error until the docs builds have been completed. ❌ 1 New FailureAs of commit 7665b71 with merge base 9d68039 ( NEW FAILURE - The following job has failed:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
bot
added
the
CLA Signed
kirklandsign
approved these changes
Oct 30, 2025
larryliu0820
added
ciflow/android
abhinaykukkadapu
pushed a commit
to abhinaykukkadapu/executorch
that referenced
this pull request
Nov 6, 2025
The Android binding exposes helpers for feeding IEEE-754 half-precision (FP16) inputs directly. Use `Tensor.fromBlob(shortArray, shape)` or reuse a direct `ShortBuffer` created via `Tensor.allocateHalfBuffer(numElements)` to avoid extra copies: ```kotlin val shape = longArrayOf(24, 4096) val halfData: ShortArray = buildHalfEncodedData() val tensor = Tensor.fromBlob(halfData, shape) val buffer = Tensor.allocateHalfBuffer(halfData.size) buffer.put(halfData) buffer.rewind() val tensorNoCopy = Tensor.fromBlob(buffer, shape) ``` All buffers must be direct and use the native byte order; the helper above takes care of this.
jirioc
pushed a commit
to nxp-upstream/executorch
that referenced
this pull request
Dec 19, 2025
The Android binding exposes helpers for feeding IEEE-754 half-precision (FP16) inputs directly. Use `Tensor.fromBlob(shortArray, shape)` or reuse a direct `ShortBuffer` created via `Tensor.allocateHalfBuffer(numElements)` to avoid extra copies: ```kotlin val shape = longArrayOf(24, 4096) val halfData: ShortArray = buildHalfEncodedData() val tensor = Tensor.fromBlob(halfData, shape) val buffer = Tensor.allocateHalfBuffer(halfData.size) buffer.put(halfData) buffer.rewind() val tensorNoCopy = Tensor.fromBlob(buffer, shape) ``` All buffers must be direct and use the native byte order; the helper above takes care of this.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
The Android binding exposes helpers for feeding IEEE-754 half-precision (FP16) inputs directly.
Use
Tensor.fromBlob(shortArray, shape)or reuse a directShortBuffercreated viaTensor.allocateHalfBuffer(numElements)to avoid extra copies:All buffers must be direct and use the native byte order; the helper above takes care of this.