Repository navigation
Update to transformers v5 #30566
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.
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
Update to transformers v5 #30566
Changes from 4 commits
Commits
Show all changes
201 commits
Select commit
Hold shift + click to select a range
990f522
update to transformers v5
hmellor 42bc6a1
Merge branch 'main' into transformers-v5
hmellor dd261ff
Merge branch 'main' into transformers-v5
hmellor 048a32c
Merge branch 'main' into transformers-v5
hmellor 933bef9
Allow Transformer v5 in `common.txt`
hmellor 12f6195
Merge branch 'main' into transformers-v5
hmellor 769d436
Update PEFT pin to avoid bad import
hmellor 214c373
Update lm-eval
hmellor ec4ffa9
`HF_HUB_ENABLE_HF_TRANSFER` -> `HF_XET_HIGH_PERFORMANCE`
hmellor 94e1429
Skip custom model which uses old imports
hmellor fbb843a
Update some more lm-eval pins
hmellor 954ea46
Merge branch 'main' into transformers-v5
hmellor 352a274
Fix timtout issues from `huggingface-hub` v1
hmellor 7c81a9c
Add `HF_HUB_DOWNLOAD_TIMEOUT` to other test images
hmellor 1aa6c01
Merge branch 'main' into transformers-v5
hmellor eea0d7c
Update missed ROCM pin
hmellor 30d8b3d
Install transformers from main temporarily
hmellor 1767434
Merge branch 'main' into transformers-v5
hmellor 17ad8ca
new main pin
hmellor 3052b14
Merge branch 'main' into transformers-v5
hmellor 6a8954e
Merge branch 'main' into transformers-v5
hmellor 489d5d9
Add backward compatibility test as copy of nightly test
hmellor c3abbd7
Skip `MiniCPMV`
hmellor 7b9016f
Merge branch 'main' into transformers-v5
hmellor 97bdae0
bump huggingface-hub
hmellor ede39e6
Bump accelerate version
hmellor 21917f4
Merge branch 'main' into transformers-v5
hmellor 113b5ee
bump transformers main pin
hmellor 8563dcc
Merge branch 'main' into transformers-v5
hmellor 9ee40ac
bump main
hmellor 57d0aad
Merge branch 'main' into transformers-v5
hmellor 84447bd
bump main
hmellor ccc8b3e
Skip experimental Transformers backend features, fix later
hmellor ee4c25c
bump hf hub
hmellor d7dd270
bumpm hf experimental version
hmellor 4da0a83
OpenCUA should be fixed now
hmellor 873adc7
Merge branch 'main' into transformers-v5
hmellor 6074afe
Merge branch 'main' into transformers-v5
hmellor fdacb70
Merge branch 'main' into transformers-v5
hmellor f7ac9c2
bump treansformers main
hmellor a6544bf
Merge branch 'main' into transformers-v5
hmellor 093999b
bump transformers main
hmellor 06a569f
Skip Molmo2
hmellor af97153
Skip openpangu
hmellor 7682651
Merge branch 'main' into transformers-v5
hmellor c0ac4cd
bump transformers main
hmellor 6e6fa6f
glmasr is no longer remote code in v5
hmellor ced047c
skip OpenCUA
hmellor 148c40e
Skip HCXVisionForCausalLM
hmellor 3daa6d8
Merge branch 'main' into transformers-v5
hmellor c46b56d
bump transformers main
hmellor f0f00aa
bump transformers main
hmellor 12810a0
Merge branch 'main' into transformers-v5
hmellor afdf43f
Merge branch 'main' into transformers-v5
hmellor 2e9fb90
Merge branch 'main' into transformers-v5
hmellor 37c707d
Skip broken custom models for processor tests
hmellor 567e00f
bump transformers main
hmellor c0f2e1b
Leave these version limits alone
hmellor a7d580a
Merge branch 'main' into transformers-v5
hmellor 4159b7f
bump main
hmellor 311ea04
Merge branch 'main' into transformers-v5
hmellor a1fb41b
bump transformers main
hmellor 1f7f5fa
Merge branch 'main' into transformers-v5
hmellor b0d99c9
Fix Flamingo min versions
hmellor 5f1d9f9
Fix Qwen3.5 min version and availability of checkpoints
hmellor a2fc272
Skip Plamo2 for HF (vLLM should still run ok)
hmellor 6b563d4
Leave tensor schema skip alone and add another for hf reasons
hmellor fe32e65
Merge branch 'main' into transformers-v5
hmellor 64fa2e2
Remove hf skip for tensor schema test
hmellor 773ad0e
`MiniCPMV` version reason should stop it working in vLLM, not just HF
hmellor 11672ab
Merge branch 'main' into transformers-v5
hmellor 445c7fe
bump main
hmellor bf217bc
Merge branch 'main' into transformers-v5
hmellor cfaa2ed
Unskip models which should now work
hmellor 04692c2
Ovis doesn't work in vLLM actually
hmellor 743ff55
Merge branch 'main' into transformers-v5
hmellor 3045933
Merge branch 'main' into transformers-v5
hmellor f7c7f5e
bump main
hmellor d99f3b5
Skip InternS1 properly
hmellor a7f676c
InternS1Pro can work
hmellor 44b7504
Update OpenCUA skip
hmellor a6d4100
Update OpenPanguVL skip
hmellor 6f6ee9e
Skip `ExaoneMoeMTP` because it's not compatible with the test harness...
hmellor 27c7f10
Merge branch 'main' into transformers-v5
hmellor 05cb1bc
Merge branch 'main' into transformers-v5
hmellor d35c05d
bump main
hmellor b3f8f01
Merge branch 'main' into transformers-v5
hmellor b0d6bb3
bump main
hmellor 3176096
Merge branch 'main' into transformers-v5
hmellor 65021f7
Merge branch 'main' of https://github.com/vllm-project/vllm into tran…
hmellor bd8cc8b
bump transformers
hmellor d19703c
Merge branch 'main' into transformers-v5
hmellor db2c800
bump transformers
hmellor 2a36e1d
Merge branch 'main' into transformers-v5
hmellor 91f54ac
bump transformers
hmellor 121b681
Put ExaoneMoe back, we'll fix it another way
hmellor 407fde0
Merge branch 'main' into transformers-v5
hmellor 489aeda
bump transformers
hmellor aad386e
Merge branch 'main' into transformers-v5
hmellor fdbf94d
Merge branch 'main' into transformers-v5
hmellor 4c138ee
bump transformers
hmellor 1216a62
Merge branch 'main' into transformers-v5
hmellor b99bedc
bump transformers
hmellor 07e1b40
Merge branch 'main' into transformers-v5
hmellor 0c515b0
Bump main
hmellor b03a868
Merge branch 'main' into transformers-v5
hmellor 1786f7f
bump main
hmellor 698f4b1
Merge branch 'main' into transformers-v5
hmellor 4da6603
bump main
hmellor f4b5c7d
Merge branch 'main' into transformers-v5
hmellor 5b28c7d
Merge branch 'main' into transformers-v5
hmellor 36460f8
bump main
hmellor 3a2b517
bump main
hmellor 1b81af3
Merge branch 'main' into transformers-v5
hmellor d978d72
Merge branch 'main' of https://github.com/vllm-project/vllm into tran…
hmellor b0fb9ec
bump main
hmellor 43bbda5
bump main
hmellor e9e48f0
Merge branch 'main' of https://github.com/vllm-project/vllm into tran…
hmellor 7b05415
bump main
hmellor 740533f
skip broken models in VLM tests
hmellor 5894f1d
More models not compatible with v5
hmellor 3c393aa
Merge branch 'main' of https://github.com/vllm-project/vllm into tran…
hmellor 3d46d90
bump main
hmellor 49b5b2f
Merge branch 'main' into transformers-v5
hmellor ead016b
Merge branch 'main' of https://github.com/vllm-project/vllm into tran…
hmellor 18dd0bd
Try try timeout fix
hmellor 5c6f97a
bump main
hmellor b99d67d
Revert "Try try timeout fix"
hmellor 19dd32d
Explicitly call `huggingface_hub.close_session` on shutdown
hmellor 4dc0c85
Move close_session earlier
hmellor 9887e8e
Merge branch 'main' of https://github.com/vllm-project/vllm into tran…
hmellor 552e9e2
bump main
hmellor c48b942
Merge branch 'main' of https://github.com/vllm-project/vllm into tran…
hmellor 77ca5a9
bump main
hmellor 202134f
Merge branch 'main' of https://github.com/vllm-project/vllm into tran…
hmellor f9d42e1
pin to 5.5.1 and 0.15.0
hmellor 59cfed0
Merge branch 'main' into transformers-v5
hmellor 8877940
bump compressed tensors
hmellor cd78122
remove `--pre` from dockerfile installs
hmellor e9b8698
Revert change to rocm-test-in
hmellor 7482524
pip-compile
hmellor 139a83f
update all hf libs for best hub support
hmellor 2da5970
Revert timeout change that didn't work
hmellor 093aca6
test push
khluu a39173e
Merge branch 'main' into transformers-v5
hmellor 6c8d30e
upgrade to transformers 5.5.3
khluu a6f6084
skip phi4 test
khluu b8463a2
skip sarvam
khluu eaa1e54
gemma4 fix
khluu 1545c11
skip tarsier2
khluu 5c3f5a5
skip minicpmv
khluu c6a4292
fix step3p5
khluu eb0479b
fix gguf loader
khluu 24f77bf
fix music flamingo
khluu 6ff178d
set shutdown timeout to 150s
khluu b5c68a3
lint
khluu 5e8afc0
Merge branch 'main' into transformers-v5
khluu d9e6625
step3p5 fix
khluu 86bc3f8
skip mteb tests
khluu 1b51036
fix TransformersMultiModalMoEForCausalLM
khluu dae2db3
skip paddleocr, nemotron, voxtral
khluu 4ce8ba8
fix gemma4 duplicate arg limit_mm_per_prompt
khluu 407fc73
gemma4 video placement fix
khluu e648579
fix gemma4
khluu 09f7c26
Update vllm/model_executor/models/gemma4_mm.py
khluu ffe85f5
fix ext pooling mm test
khluu 814e130
Update vllm/model_executor/models/gemma4_mm.py
DarkLight1337 2393c1e
revert timeout change as it didn't fix the issue
hmellor 16eb5f1
simpler test skip
hmellor bce473a
fix pre-commit
hmellor 1b0635d
fix hf runner using vllm configs
hmellor 57e7949
skip other phi4 tests
hmellor cfe4e32
skip failing ultravox test
hmellor 9eb2d21
skip transformers backend eagle3 test because it's not urgent
hmellor b79aec0
Merge branch 'main' into transformers-v5
khluu 3bedcc2
fix gemma4 image placeholder
khluu d42fd89
gemma4 tensor shape fix
khluu 48a3199
skip fireredasr2 asr-nano-2512 fireredlid
khluu 4c6cac1
use full gpu for basic models init test
khluu 2d7903e
register custom config
khluu 875c012
gc collect llm delete test
khluu 6776f31
add todo comment
hmellor 8f551d0
alternative fix for vllm config in get_tokenizer
hmellor e67530c
revert step3p5 test changes now that get_tokenizer is fixed
hmellor 87f3a14
Bump `huggingface-hub` and remove delete workaround
hmellor 40742ca
temp fix for tinymixtral test
khluu ea58ae3
Revert "temp fix for tinymixtral test"
khluu 3693a95
fix tiny-mixtral CPU test: reduce bfloat16 rounding error
khluu f50bb9d
add back firered and funasr model back to registry
khluu 6d40ca7
skip XverseForCausalLM tests on transformers v5
khluu e187e72
claude fix pretokenizer for step3p5 and tool parser
khluu cb03f5d
fix LoRA dual-stream defs guarded by import-time env check
khluu cc19a1b
fix get_tokenizer crash when tokenizer path has no model config
khluu d894c4b
Revert "claude fix pretokenizer for step3p5 and tool parser"
hmellor 816db8b
better fix for bad tokenizer_class config
hmellor 410ae69
Revert "fix LoRA dual-stream defs guarded by import-time env check"
hmellor 962976d
test side fix for lora dual stream
hmellor 2cba808
Merge branch 'main' into transformers-v5
hmellor f48f8ce
skip FireRedASR2, FireRedLID, FunASR tests on transformers >= 5.2
khluu 79e9772
Merge branch 'main' into transformers-v5
khluu 75efe07
fix FireRedASR2/FireRedLID/FunASR skip reason: hf -> vllm
khluu File filter
Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
There are no files selected for viewing
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
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
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
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.
Uh oh!
There was an error while loading. Please reload this page.