ONNX Runtime improvements #1306
xenova
merged
49 commits into
v4
from
ort-improvements
ONNX Runtime improvements (experimental native webgpu; fix iOS) (#1231)
747a04de
customize the wasm paths (#1250)
bdaead04
Merge branch 'main' into ort-improvements
e025b7be
[internal] Add is_decoder option to session retrieval for preferred o…
573e5dc8
Update tests
4d3a28d9
Formatting
80f58ce7
Bump ort versions
071d728a
Bump onnxruntime-node version
cfb8ff3c
Bump versions
7ea7718d
Bump ORT versions
29728286
Merge branch 'main' into ort-improvements
ca9765c3
Bump versions
ce7dfc08
Only check webgpu fp16 for non-node environments
87cdd0a0
Fix
51c4bda2
Assume node supports webgpu
0f58976e
Update ORT node support comment
00f177b3
Relax test strictness
11b59ed9
Update conversion script versions
99f62ee1
sroussey
commented
on 2025-06-18
Downgrade onnxslim
654a4534
Merge branch 'main' into ort-improvements
c7d47ce9
cleanup
36b83817
Merge branch 'main' into ort-improvements
67568491
Update package-lock.json
9a950976
Update onnxruntime versions
37976d25
Update post-build script
6096b015
Use built-in session release function
93a8d47a
Call garbage collection after each tokenizer test
194044f6
Do not double-throw error
a0751bf8
Merge branch 'main' into ort-improvements
e575d6d2
Fix race-condition in build process with file removal
b4506dcb
Update versions
df4a2c13
Bump jinja version
70890bf4
[version] Update to 3.6.3
467f59c2
Add support for LFM2 models (#1367)
1d08e91e
Merge branch 'main' into ort-improvements
abbcd89c
Use prefix in lfm2 output location (#1369)
2c32e1d9
Merge branch 'main' into ort-improvements
66ffdb04
Update package-lock.json
fc042ada
Run `npm audit fix`
8dd74314
Add special tokens in text-generation pipeline if tokenizer requires …
11fdae56
Add support for ModernBERT Decoder (#1371)
9407ed88
Use from/to buffer instead of string
255753aa
Add support for Voxtral (#1373)
beb1d170
Support longform voxtral processing (#1375)
635bff5b
[version] Update to 3.7.0
0feb5b76
Merge branch 'main' into ort-improvements
0a86afde
Add support for Arcee (#1377)
82206db5
Optimize tensor.slice() (#1381)
bd2449ef
Merge branch 'main' into ort-improvements
effca640
Login to write a write a comment.
Login via GitHub