Commit Graph

19476 Commits

Author SHA1 Message Date
Lysandre
4eb99fed12 Scaffolding 2025-07-01 18:23:16 +02:00
Lysandre
feb1eb7531 Scaffolding 2025-07-01 18:22:55 +02:00
Lysandre
733bcb4fed Fix CB 2025-07-01 13:08:36 +02:00
Lysandre Debut
e8f90b5397
Split transformers chat and transformers serve (#38443)
* Next token

* Split chat and serve

* Support both generation methods

* Style

* Generation Config

* temp

* temp

* Finalize serving.py

Co-authored-by: =?UTF-8?q?c=C3=A9lina?= <hanouticelina@gmail.com>

* Finalize chat.py

* Update src/transformers/commands/serving.py

Co-authored-by: célina <hanouticelina@gmail.com>

* Lucain's comments

Co-authored-by: Lucain <lucain@huggingface.co>

* Update

* Last comments on PR

* Better error handling

* Better error handling

* CI errors

* CI errors

* Add tests

* Fix tests

* Fix tests

* [chat] Split chat/serve (built on top of lysandre's PR) (#39031)

* Next token

* Split chat and serve

* Support both generation methods

* Style

* Generation Config

* temp

* temp

* Finalize serving.py

Co-authored-by: =?UTF-8?q?c=C3=A9lina?= <hanouticelina@gmail.com>

* Finalize chat.py

* Update src/transformers/commands/serving.py

Co-authored-by: célina <hanouticelina@gmail.com>

* Lucain's comments

Co-authored-by: Lucain <lucain@huggingface.co>

* Update

* Last comments on PR

* Better error handling

* Better error handling

* CI errors

* CI errors

* Add tests

* Fix tests

* Fix tests

* streaming tool call

* abstract tool state; set tool start as eos

* todos

* server working on models without tools

* rm chat's deprecated flags

* chat defaults

* kv cache persists across calls

* add server docs

* link

* Update src/transformers/commands/serving.py

* Apply suggestions from code review

* i love merge conflicts

* solve multi turn with tiny-agents

* On the fly switching of the models

* Remove required positional arg

---------

Co-authored-by: Lysandre <hi@lysand.re>
Co-authored-by: =?UTF-8?q?c=C3=A9lina?= <hanouticelina@gmail.com>
Co-authored-by: Lucain <lucain@huggingface.co>

* Protect names

* Fix tests

---------

Co-authored-by: =?UTF-8?q?c=C3=A9lina?= <hanouticelina@gmail.com>
Co-authored-by: Lucain <lucain@huggingface.co>
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>
2025-06-30 15:10:53 +02:00
Yih-Dar
539c6c2fa8
All CI jobs with A10 (#39119)
all a10

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
2025-06-30 14:23:27 +02:00
Ryan Mullins
ed9f252608
docs: Gemma 3n audio encoder (#39087)
Updating Gemma 3n docs and docstrings to clarify the relationship
between the newly trained audio encoder used in Gemma 3n and the USM
model from the original paper.
2025-06-30 14:10:51 +02:00
Yuxuan Zhang
4a79bf947d
Fix some bug for finetune and batch infer For GLM-4.1V (#39090)
* update

* 1
2025-06-30 12:16:22 +02:00
Yao Matrix
2100ee6545
fix UT failures on XPU w/ stock PyTorch 2.7 & 2.8 (#39116)
* fix UT failures on XPU w/ stock PyTorch 2.7 & 2.8

Signed-off-by: YAO Matrix <matrix.yao@intel.com>

* zamba2

Signed-off-by: YAO Matrix <matrix.yao@intel.com>

* xx

Signed-off-by: YAO Matrix <matrix.yao@intel.com>

* internvl

Signed-off-by: YAO Matrix <matrix.yao@intel.com>

* tp cases

Signed-off-by: YAO Matrix <matrix.yao@intel.com>

---------

Signed-off-by: YAO Matrix <matrix.yao@intel.com>
2025-06-30 11:49:03 +02:00
Yih-Dar
ccf2ca162e
skip some test_sdpa_can_dispatch_on_flash (#39092)
Some checks failed
Self-hosted runner (benchmark) / Benchmark (aws-g5-4xlarge-cache) (push) Has been cancelled
Build documentation / build (push) Has been cancelled
Slow tests on important models (on Push - A10) / Get all modified files (push) Has been cancelled
Self-hosted runner (push-caller) / Check if setup was changed (push) Has been cancelled
Secret Leaks / trufflehog (push) Has been cancelled
Update Transformers metadata / build_and_package (push) Has been cancelled
Slow tests on important models (on Push - A10) / Slow & FA2 tests (push) Has been cancelled
Self-hosted runner (push-caller) / build-docker-containers (push) Has been cancelled
Self-hosted runner (push-caller) / Trigger Push CI (push) Has been cancelled
* fix

* fix

* fix

---------

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
2025-06-27 23:08:14 +02:00
st81
a11f692895
Fixes the failing test test_is_split_into_words in test_pipelines_token_classification.py (#39079)
Some checks failed
Self-hosted runner (benchmark) / Benchmark (aws-g5-4xlarge-cache) (push) Waiting to run
Build documentation / build (push) Waiting to run
Slow tests on important models (on Push - A10) / Get all modified files (push) Waiting to run
Slow tests on important models (on Push - A10) / Slow & FA2 tests (push) Blocked by required conditions
Self-hosted runner (push-caller) / Check if setup was changed (push) Waiting to run
Self-hosted runner (push-caller) / build-docker-containers (push) Blocked by required conditions
Self-hosted runner (push-caller) / Trigger Push CI (push) Blocked by required conditions
Secret Leaks / trufflehog (push) Waiting to run
Update Transformers metadata / build_and_package (push) Waiting to run
New model PR merged notification / Notify new model (push) Has been cancelled
* Fix test pipelines token classification for is_split_into_words

* Fix incorrect import format
2025-06-27 19:25:32 +01:00
Sandeep Yadav
18143c76bf
Sandeepyadav1478/2025 06 19 deberta v2 model card update (#38895)
* [docs]: update deberta-v2.md model card

* chore: req updates

* chore: address code review feedback and update docs

* chore: review feedback and updates

* chore: model selection updates

* chores: quantizations review updates
2025-06-27 10:35:30 -07:00
Steven Liu
02a769b058
[fix] Add FastSpeech2ConformerWithHifiGan (#38207)
* add to mapping

* oops

* oops

* add to config_mapping_names

* revert

* fix?

* config-mapping-names

* fix?

* fix?
2025-06-27 09:38:21 -07:00
Benjamin Bossan
c2dc72bb5f
TST Fix PEFT integration test bitsandbytes config (#39082)
TST Fix PEFT integration test bitsandbytes config

The PEFT integration tests still used load_in_{4,8}_bit, which is
deprecated, moving to properly setting BitsAndBytesConfig. For 4bit,
also ensure that nf4 is being used to prevent

> RuntimeError: quant_type must be nf4 on CPU, got fp4
2025-06-27 18:33:11 +02:00
Matej Sirovatka
c8064bea9a
Fix: unprotected import of tp plugin (#39083) 2025-06-27 17:28:05 +02:00
farrosalferro
dd7dc4a4a2
Add Fast Image Processor for Chameleon (#37140)
* Add Fast Image Processor for Chameleon

* add warning to resize and move blend_rgba to convert_to_rgb

* Remove unrelated files

* Update image_processing_chameleon_fast to use auto_docstring

* fix equivalence test

---------

Co-authored-by: Yoni Gozlan <74535834+yonigozlan@users.noreply.github.com>
Co-authored-by: yonigozlan <yoni.gozlan@huggingface.co>
2025-06-27 15:26:57 +00:00
Yih-Dar
6d773fc3bc
fix dots1 tests (#39088)
fix

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
2025-06-27 16:54:11 +02:00
Tijana Vukovic
c8764ab935
guard torch distributed check (#39057)
* guard torch distributed check

* Update src/transformers/pipelines/base.py

---------

Co-authored-by: Matt <Rocketknight1@users.noreply.github.com>
2025-06-27 14:49:47 +00:00
MinJu-Ha
49d9fd49bd
Add Fast Image Processor for mobileViT (#37143)
* Add image_processing_mobilevit_fast.py

* Fix copies

* update _preprocess for channel_flip

* Update for batched image processing

* Resolve merge conflicts with main

* Fix import order and remove trailing whitespace (ruff clean-up)

* Fix copy inconsistencies

* Add NotImplementedError for post_process_semantic_segmentation to satisfy repo checks

* Add auto_docstring

* Adjust style

* Update docs/source/en/model_doc/mobilevit.md

Co-authored-by: Yoni Gozlan <74535834+yonigozlan@users.noreply.github.com>

* Update src/transformers/models/mobilevit/image_processing_mobilevit_fast.py

Co-authored-by: Yoni Gozlan <74535834+yonigozlan@users.noreply.github.com>

* Update src/transformers/models/mobilevit/image_processing_mobilevit_fast.py

Co-authored-by: Yoni Gozlan <74535834+yonigozlan@users.noreply.github.com>

* Delete not used function

* test: add missing tests for  and

* Add post_process_semantic_segmentation to mobilevit_fast.py

* Add preprocess function to image_processing_mobilebit_fast.py

* ruff check for formatting

* fix: modify preprocess method to handle BatchFeature correctly

* Remove logic for default value assignment

Co-authored-by: Yoni Gozlan <74535834+yonigozlan@users.noreply.github.com>

* Remove normalization adn RGB conversion logic not used in slow processor

Co-authored-by: Yoni Gozlan <74535834+yonigozlan@users.noreply.github.com>

* Simplify return_tensors logic using one-liner conditional expression

Co-authored-by: Yoni Gozlan <74535834+yonigozlan@users.noreply.github.com>

* Remove unused normalization and format parameters

Co-authored-by: Yoni Gozlan <74535834+yonigozlan@users.noreply.github.com>

* add **kwargs and remove default values in _preprocess

* add slow_fast equivalence tests for segmentation

* style: autoformat code with ruff

* Fix slow_fast equivalence test

* merge + remove skipped test

---------

Co-authored-by: Yoni Gozlan <74535834+yonigozlan@users.noreply.github.com>
Co-authored-by: yonigozlan <yoni.gozlan@huggingface.co>
2025-06-27 14:40:24 +00:00
Nahieli
4336ecd1ea
add fast image processor nougat (#37661)
* add fast image processor nougat

* test fixes

* docstring white space

* last fixes

* docstring_type

* tolerance unit test

* fix tolerance

* fix rtol

* remove traling white space

* remove white space

* note for tolerance unit test

* fix tests

* remove print

---------

Co-authored-by: yonigozlan <yoni.gozlan@huggingface.co>
Co-authored-by: Yoni Gozlan <74535834+yonigozlan@users.noreply.github.com>
2025-06-27 14:39:43 +00:00
Benjamin Bossan
0c35280e58
TST PEFT integration tests with pipeline generate (#39086)
Some PEFT integration tests involving text generation pipelines were
failing since #38129 because the base model is too small to generate
longer sequences. Setting max_new_tokens fixes this.
2025-06-27 15:58:10 +02:00
JINO ROHIT
993665a5ff
fixed typo for docstring in prepare_inputs method (#39071) 2025-06-27 13:57:56 +00:00
Yih-Dar
839893c86b
fix mistral3 tests (#38989)
* fix

* fix

* fix

* fix

* fix

---------

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
2025-06-27 15:44:10 +02:00
eustlb
2b85b6ce19
[Whisper] 🚨 Fix pipeline word timestamp: timestamp token is end of token time !!! (#36632)
* timestamp token is end of token time !!!

* ensure correct alignment between tokens and timestamp tokens

* ignore input tokens for DTW computation

* use num_frames to avoid token timestamp hallucinations

* token timestamps test updates !

* num_frames: deprecate and use attention_mask instead

* avoid breaking change

* fix the pipeline usage for chunk approach

* make style

* better logging

* better logging

* make style

* update tests with correct values
2025-06-27 12:51:43 +00:00
eustlb
9c8d3a70b8
Pipeline: fix unnecessary warnings (#35753)
* return attention mask

* use correct model input name

* fix

* make
2025-06-27 14:32:03 +02:00
Yaswanth Gali
1750c518dd
Add EoMT Model || 🚨 Fix Mask2Former loss calculation (#37610)
* Initial Commit

* up

* More changes

* up

* Only mask_logits mismatch

* close enough logits debug later

* fixes

* format

* Add dummy loss

* Close enough processing for semantic seg

* nit

* Added panoptic postprocessor

* refactor

* refactor

* finally fixed panoptic postprocessor

* temp update

* Refactor ForUniversalSegmentation class

* nits and config update

* Few fixes and inference matches

* change mapping

* Added training support but loss slightly off 🥲

* Loss is matching 😀

* update

* Initial tests skelton

* changes

* tests update

* more modular

* initial tests

* updates

* better docstrings

* changes

* proc tests passing :)

* Image processor update

* tiny change

* QOL changes

* Update test w.r.t latest attn refactor

* repo-consistency fixes

* up

* Image proc fix and integration tests :)

* docs update

* integration tests

* fix

* docs update 🥰

* minor fix

* Happy CI

* fix

* obvious refactoring

* refactoring w.r.t review

* Add fask image proc skelton

* Fast Image proc and cleanups

* Use more modular

* tests update

* Add more tests

* Nit

* QOL updates

* change init_weights to torch default

* add eager func coz of make style

* up

* changes

* typo fix

* Updates

* More deterministic tests

* More modular

* go more modular 🚀

* up

* dump

* add supprot for giant ckpts

* overhaul

* modular

* refactor

* instace seg is ready

* cleanup

* forgot this

* docs cleanup

* minor changes

* EoMT - > Eomt

* Happy CI

* remove redundant comment

* Change model references

* final change

* check annealing per block

* My other PR changes 😂

---------

Co-authored-by: Cyril Vallez <cyril.vallez@huggingface.co>
2025-06-27 14:18:18 +02:00
Yao Matrix
0106a50a6b
fix a bunch of XPU UT failures on stock PyTorch 2.7 and 2.8 (#39069)
* fix a bunch of XPU UT failures on stock PyTorch 2.7 and 2.8

Signed-off-by: YAO Matrix <matrix.yao@intel.com>

* qwen3

Signed-off-by: YAO Matrix <matrix.yao@intel.com>

* quanto

Signed-off-by: YAO Matrix <matrix.yao@intel.com>

* models

Signed-off-by: YAO Matrix <matrix.yao@intel.com>

* fix style

Signed-off-by: YAO Matrix <matrix.yao@intel.com>

* idefics2

Signed-off-by: YAO Matrix <matrix.yao@intel.com>

---------

Signed-off-by: YAO Matrix <matrix.yao@intel.com>
2025-06-27 14:01:53 +02:00
Mohamed Mekkouri
cb17103bd5
Uninstallling Flash attention from quantization docker (#39078)
Some checks are pending
Self-hosted runner (benchmark) / Benchmark (aws-g5-4xlarge-cache) (push) Waiting to run
Build documentation / build (push) Waiting to run
New model PR merged notification / Notify new model (push) Waiting to run
Slow tests on important models (on Push - A10) / Get all modified files (push) Waiting to run
Slow tests on important models (on Push - A10) / Slow & FA2 tests (push) Blocked by required conditions
Self-hosted runner (push-caller) / Check if setup was changed (push) Waiting to run
Self-hosted runner (push-caller) / build-docker-containers (push) Blocked by required conditions
Self-hosted runner (push-caller) / Trigger Push CI (push) Blocked by required conditions
Secret Leaks / trufflehog (push) Waiting to run
Update Transformers metadata / build_and_package (push) Waiting to run
* update

* revert
2025-06-27 13:51:46 +02:00
BUI Van Tuan
371c471113
Fix initialization of OneFormer (#38901)
* fix initialization of OneFormer

* remove redundant initializations

* remove redundant initializations

* remove redundant initializations

* keep BC
2025-06-27 12:39:37 +02:00
Yih-Dar
540a10848c
fix Gemma3nProcessorTest (#39068)
* fix

* fix

* oups forgot style

---------

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
Co-authored-by: Cyril Vallez <cyril.vallez@gmail.com>
2025-06-27 12:28:10 +02:00
Yaswanth Gali
0d66ef7792
Cleanup Attention class for Siglip and dependent models (#39040)
* cleanup attention class

* More models

* more models

* Changes

* make style

* Should fix CI

* This should work 🙏
2025-06-27 12:14:09 +02:00
eustlb
1ccc73dee9
[Whisper] fix shape mismatch in tests (#39074)
fix shape mismatch
2025-06-27 09:27:42 +00:00
Steven Liu
a52478253b
[docs] Tensor parallelism (#38241)
Some checks are pending
Self-hosted runner (benchmark) / Benchmark (aws-g5-4xlarge-cache) (push) Waiting to run
Build documentation / build (push) Waiting to run
Slow tests on important models (on Push - A10) / Get all modified files (push) Waiting to run
Slow tests on important models (on Push - A10) / Slow & FA2 tests (push) Blocked by required conditions
Secret Leaks / trufflehog (push) Waiting to run
Update Transformers metadata / build_and_package (push) Waiting to run
* updates

* feedback

* badges

* fix?

* fix?

* fix?

* fix?
2025-06-26 14:40:45 -07:00
Steven Liu
84e8696cae
[docs] @auto_docstring (#39011)
* refactor

* feedback
2025-06-26 14:21:54 -07:00
Drew Ross
018855de63
Update PEGASUS-X model card (#38971)
* Update PEGASUS-X model card

* Add cache_implementation argument in quantization code example

* Update CLI example

* Apply suggestions from code review

Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>

* Remove TensorFlow and Flax badges

---------

Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
2025-06-26 13:54:48 -07:00
Steven Liu
757c26fb40
[docs] Model contribution (#38995)
Some checks are pending
Self-hosted runner (benchmark) / Benchmark (aws-g5-4xlarge-cache) (push) Waiting to run
Build documentation / build (push) Waiting to run
New model PR merged notification / Notify new model (push) Waiting to run
Slow tests on important models (on Push - A10) / Get all modified files (push) Waiting to run
Slow tests on important models (on Push - A10) / Slow & FA2 tests (push) Blocked by required conditions
Self-hosted runner (push-caller) / Check if setup was changed (push) Waiting to run
Self-hosted runner (push-caller) / build-docker-containers (push) Blocked by required conditions
Self-hosted runner (push-caller) / Trigger Push CI (push) Blocked by required conditions
Secret Leaks / trufflehog (push) Waiting to run
Update Transformers metadata / build_and_package (push) Waiting to run
improve
2025-06-26 12:25:14 -07:00
Yih-Dar
b372bb5ed1
fix layoutlmv3 tests (#39050)
* fix

* fix

* fix

* fix

* fix

* fix

---------

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
2025-06-26 20:07:17 +02:00
StevenBucaille
f171e7e884
Update SuperPoint model card (#38896)
* docs: first draft to more standard SuperPoint documentation

* Apply suggestions from code review

Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>

* docs: reverted changes on Auto classes

* docs: addressed the rest of the comments

* docs: remove outdated reference to keypoint detection task guide in SuperPoint documentation

* Update superpoint.md

---------

Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
2025-06-26 10:13:06 -07:00
Yih-Dar
2f50230c59
fix t5gemma tests (#39052)
* fix

* fix

* fix

* fix

* fix

---------

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
2025-06-26 18:48:14 +02:00
Yih-Dar
23b7e73f05
fix test_compare_unprocessed_logit_scores (#39053)
fix

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
2025-06-26 18:36:56 +02:00
Anton Vlasjuk
58c7689226
[Flex Attn] Fix torch 2.5.1 incompatibilities (#37406)
* remove compile on mask creation, ensure kv blocks do not explode on indices

* trigger ci

* switch dynamic compilation to false

* patch new masking functions as well

* add len check

* i was wrong

* last comment
2025-06-26 18:23:55 +02:00
Lysandre
5154497607 Dev version 2025-06-26 18:04:36 +02:00
Kyle Sayers
0a8081b03d
[Modeling] Fix encoder CPU offloading for whisper (#38994)
* fix cpu offloading for whisper

Signed-off-by: Kyle Sayers <kylesayrs@gmail.com>

* unskip offloading tests

Signed-off-by: Kyle Sayers <kylesayrs@gmail.com>

* revert small change

Signed-off-by: Kyle Sayers <kylesayrs@gmail.com>

* remove tests

Signed-off-by: Kyle Sayers <kylesayrs@gmail.com>

---------

Signed-off-by: Kyle Sayers <kylesayrs@gmail.com>
2025-06-26 15:56:33 +00:00
Ryan Mullins
c63cfd6a83
Gemma 3n (#39059)
* Gemma 3n

* initial commit of Gemma 3n scaffold

* Fixing param pass through on Gemm3p5RMSNorm

* Adds Einsum layer to Gemma 3n

* Updating EinsumLayer API

* Undoing erroneous force push

* Reverting RMSNorm to with_scale by default

* Adds LAuReL to Gemma 3n

* Adds AltUp to Gemma 3n

* Adding Gemma3p5 overall and text config with vision and audio config placeholders (#3)

* Adding gemma3p5 text configs

* Adding audio config placeholders

* Adding a placeholder for vision configs

* Updating MobileNetVisionConfig, inheriting TimmWrapperConfig

* Updating text configs

* Update src/transformers/models/gemma3p5/modular_gemma3p5.py

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Removing altup configs to accept the suggested configs

* Update src/transformers/models/gemma3p5/modular_gemma3p5.py

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Updating altup config

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Addressing review comments and updating text configs

* Adding a config for activation sparsity

* Updating configs to pass through options to super class init and adjust some name prefixes

* Updating laurel and altup with corrected config values

* Normalizing sub_config initializers

---------

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Updating MLP with activation sparsity (#2)

* Updating DecoderBlock for Gemma 3n (#3)

* Initial Gemm3nTextModel (#4)

NOTE: This implementation WILL CHANGE in the coming weeks, however, changes will be strictly additive and this will remain a suitable baseline for downstream implementations to reference.

* Adding KV Cache Sharing

* Adds Einsum layer to Gemma 3n

* Updating EinsumLayer API

* Refactored kv cache sharing in attention

* Adding KVStore for cache sharing

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Update src/transformers/cache_utils.py

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Undoing erroneous force push

* Reverting RMSNorm to with_scale by default

* Adds LAuReL to Gemma 3n

* Updating KV Cache Sharing implementation

* Updating the q and k norm definitions in the attention module

* Fixing name error for q,k,v RMS norm to use the right 3n module

* Updating MLP with activation sparsity

* Updating DecoderBlock for Gemma 3.5

* Updating kv cache sharing implementation with the use of a cache buffer and refactoring some lines of code

* Isolating KV Cache logic to relevant components

* Fixing logic error in Gemma3nAttention.forward

* Refactoring caching contributions and fixing kv_store initialization

* Simplifying Configs

* Remove errant self from super init call

* Bug fix in the Attention module - changing self.head_dim to config.head_dim

* Bug fixes in the LaurelBlock and RMS Norm super init call

* removing redundant code from a merge

* Adding per_layer_inputs to TextModel

* Adding preprocess embeddings with altup

* Adds per-layer-to-single output and a host of TODOs

* Integrating altup predict with the model workflow and other minor bug fixes

* Using nn.Embedding temporarily for text model

* It goes forward

* Minor refactor of attention sparsity and RoPE initialization

* Fixing duplicate rope_scaling param bug when loading from pretrained

---------

Co-authored-by: Sindhu Raghuram <sindhuraghuram@google.com>
Co-authored-by: SindhuRaghuram97 <114270661+SindhuRaghuram97@users.noreply.github.com>

* Normalizing on altup_num_inputs config option

* regenerating modeling file after syncing to HEAD

* Use torch.std(..., unbiased=False) for activation sparsity (#8)

* Refactoring to a single QVK Norm (#13)

* AltUp: support scale_corrected_output (#14)

* Converts einsums to nn.Linear (#7)

* Converts einsums to nn.Linear

* Removing unused variables

* Aligning SharedKVCache with HybridCache (#11)

* Alinging SharedKVStore with HybridCache

* Remove KVStore. Refactor apply_rotary_pos_emb for sharing

* Addressing review comments

* Supporting split modality embeddings in Gemma3n (#10)

* Adding the Embedder class

* Update modular

Co-authored-by: Ryan Mullins <ryan@ryanmullins.org>

* Update modular

Co-authored-by: Ryan Mullins <ryan@ryanmullins.org>

* Update modular

Co-authored-by: Ryan Mullins <ryan@ryanmullins.org>

* Update modular

Co-authored-by: Ryan Mullins <ryan@ryanmullins.org>

* Update modular

Co-authored-by: Ryan Mullins <ryan@ryanmullins.org>

* Update modular

Co-authored-by: Ryan Mullins <ryan@ryanmullins.org>

* Addressing review comments, adding audio embedding layers, integrating embedder with the remaining architecture, adding a forward method for conditional generation

* Apply suggestions from code review

Co-authored-by: Ryan Mullins <ryan@ryanmullins.org>

* Update modular

Co-authored-by: Ryan Mullins <ryan@ryanmullins.org>

* Addressing review comments, prop drilling audio and vision configs to the text config

* Removing TODO's that have been addressed

* Simplify Embedder init and add audio embeddings

* Embeddings refactor. Adds Gemma3nAudioEmbedder and Gemma3nVisionEmbedder

* Refactoring vision and audio embeddings into ConditionalGeneration model

---------

Co-authored-by: Ryan Mullins <ryan@ryanmullins.org>
Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Updating attention mask for Gemma 3.5 (#15)

* xxx_token_index to xxx_token_id

* remvoing deprecated last_cache_position

* Removing references to SigLIP

* Always init per-layer inputs

* Using torch.finfo().min for epsilon_tensor

* Gemma3nDecoderLayer inherits from Gemma3DecoderLayer. Remove gating lambdas

* fix modular GEMMA3N_INPUTS_DOCSTRING

* Gemma3nAttention inherits from Gemma3Attention

* Modular inheritance fixes

* CausalLM conversion script for 4B model (#16)

* Add Gemma3n Audio Encoder (#6)

* initial commit of Gemma 3.5 scaffold

* Fixing param pass through on Gemm3nRMSNorm

* Adds Einsum layer to Gemma 3.5

* Updating EinsumLayer API

* Undoing erroneous force push

* Reverting RMSNorm to with_scale by default

* Adds LAuReL to Gemma 3n

* Adds AltUp to Gemma 3n

* Adding Gemma3n overall and text config with vision and audio config placeholders (#3)

* Adding gemma3n text configs

* Adding audio config placeholders

* Adding a placeholder for vision configs

* Updating MobileNetVisionConfig, inheriting TimmWrapperConfig

* Updating text configs

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Removing altup configs to accept the suggested configs

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Updating altup config

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Addressing review comments and updating text configs

* Adding a config for activation sparsity

* Updating configs to pass through options to super class init and adjust some name prefixes

* Updating laurel and altup with corrected config values

* Normalizing sub_config initializers

---------

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Updating MLP with activation sparsity (#2)

* Updating DecoderBlock for Gemma 3.5 (#3)

* Initial Gemm3nTextModel (#4)

NOTE: This implementation WILL CHANGE in the coming weeks, however, changes will be strictly additive and this will remain a suitable baseline for downstream implementations to reference.

* Adding KV Cache Sharing

* Adds Einsum layer to Gemma 3.5

* Updating EinsumLayer API

* Refactored kv cache sharing in attention

* Adding KVStore for cache sharing

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Update src/transformers/cache_utils.py

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Undoing erroneous force push

* Reverting RMSNorm to with_scale by default

* Adds LAuReL to Gemma 3n

* Updating KV Cache Sharing implementation

* Updating the q and k norm definitions in the attention module

* Fixing name error for q,k,v RMS norm to use the right Gemma 3n module

* Updating MLP with activation sparsity

* Updating DecoderBlock for Gemma 3.5

* Updating kv cache sharing implementation with the use of a cache buffer and refactoring some lines of code

* Isolating KV Cache logic to relevant components

* Fixing logic error in Gemma3nAttention.forward

* Refactoring caching contributions and fixing kv_store initialization

* Simplifying Configs

* Remove errant self from super init call

* Bug fix in the Attention module - changing self.head_dim to config.head_dim

* Bug fixes in the LaurelBlock and RMS Norm super init call

* removing redundant code from a merge

* Adding per_layer_inputs to TextModel

* Adding preprocess embeddings with altup

* Adds per-layer-to-single output and a host of TODOs

* Integrating altup predict with the model workflow and other minor bug fixes

* Using nn.Embedding temporarily for text model

* It goes forward

* Minor refactor of attention sparsity and RoPE initialization

* Fixing duplicate rope_scaling param bug when loading from pretrained

---------

Co-authored-by: Sindhu Raghuram <sindhuraghuram@google.com>
Co-authored-by: SindhuRaghuram97 <114270661+SindhuRaghuram97@users.noreply.github.com>

* Normalizing on altup_num_inputs config option

* Adding audio encoder config

* Adds high-level components for Audio Encoder

* Implement uniform reducer for Audio Encoder

* Adding placeholders for Conformer components in Audio Encoder

* Adding placeholders for SubSampleConvProjection components in Audio Encoder

* Adding SequenceLayer component placeholders

* Implementing Gemma3nAudioEncoder with nn.Sequential

* Implementing Gemma3nAudioSubSampleConvProjection with nn.Sequential

* Implementing Conformer model with SequenceLayers

* Use OrderedDict in nn.Sequential initializers

* Implements sl.Residual in Torch with nn.Sequential and OrderedDict

* Adopting a base SequenceLayer class with default forward() method

* Implementing sl.GatedLinearUnit in Torch

* Implementing sl.Swish in Torch

* Implementing sl.ReLU in Torch

* Implementing sl.Scale in Torch

* Removing sl.Dropout after tree-shaking

* Implementing sl.RMSNorm in Torch with fake shape

* Implementing sl.GroupNorm in Torch

* Implementing sl.Conv2d in Torch

* Implementing sl.Dense in Torch

* Removing sl.Delay layers, which act as pass-throughs

* Connecting shapes to configs in initializers

* Removing sl.Emit

* Implementing sl.ExpandDims in Torch

* Adding sl.GradientClipping to Torch

* Implementing sl.DenseShaped in Torch

* Implementing sl.LDPA in Torch

* Removing unused sl.CombinedQKVProj class

* Fixing erroneous type hint

* Implemnenting sl.DepthwiseConv1D in Torch

* Implementing sl.MaskInvalid in Torch

* Fixes for initialization

* Fixes for saving weights

* Removing einsums per feedback from HF staff

* Removing Sequence Layers idioms from audio encoder

* Fixes for reviewer comments

* CausalLM conversion script for 4B model

* inv_timescales to non-persistent buffer

* Addressing audio encoder Attention feedback

* Addressing Gemma3nAudioSSCPConvBlock feedback

* Addressing Gemma3nAudioConformerAttention feedback

* Addressing padding feedback

* Weights conversion loads audio state dict

* Always use vision_config so saving works

* Token id updates for configs

* Stubs for interleaving audio embs

* Addressing reviewer feedback

---------

Co-authored-by: SindhuRaghuram97 <114270661+SindhuRaghuram97@users.noreply.github.com>
Co-authored-by: Sindhu Raghuram <sindhuraghuram@google.com>

* Fixing cache access error

* Removing duplicate code from a bad merge

* Gemma 3n Text + Vision Part 1 (#17)

* testing utilities for numerics comparisons

* Corrected einsum to nn.Linear weights conversion

* Inherit scaled word embs from Gemma3 not Bart

* Fixing transposes for collapsed linears

* More transpose fixes

* numpy api fix

* RMSNorm: Explicit kwargs, scale_shift=0.0 when with_scale=True

* Force AltUp  to float32

* Updating debugging script for AudioEncoder debugging

* Support divide_weight_by_sqrt_fan_in from JAX for per-layer inputs

* Correcting attention einsum conversions

* RMSNorm in type of x

* Fixing douplicate laurel norm/gating

* KV sharing using the right previous indices

* Refactor kv shared index computation. Correct frac_shared_layers

* Use num_shared_layers instead of inferring from a fraction

* fixing a bug for logging

* Fix shared data_ptrs in altup inits

* rope: adjust proj -> norm -> rope to preserve computation (#20)

* rope: adjust proj -> norm -> rope to preserve computation

* Removing some breaking language model fluff in ConditionalGeneration

* Consolidate query_states transforms

---------

Co-authored-by: Douglas Reid <21148125+douglas-reid@users.noreply.github.com>
Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Vectorize the loops in AltUp (#19)

* Vectorize the loops in AltUp

* fix typo

* Expanding to support batched inputs

* remove extra debug script

* Fix AltUp.forward

---------

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Add 'scale_shift=0.0, with_scale=True' to the final norm in TextModel

* Convert norm to 1/sqrt (#21)

* Convert norm to 1/sqrt

* Scale shift change per Phil's rec

* Adding default activation sparsity

* Fixing 2B config in weights conversion script

* Fixing RMSNorm parameters - adding scale_shift and with_scale

* Correcting query pre-attention scaling

* Adding query_rescale_scalar to text config

* Adding layer_idx to MLP

* Permafix for input_layernorm

* Use 1/sqrt instead of rsqrt in DecoderLayer

* Fix o_proj conversion

* Conversion script update for vision encoder

* Removing logging for debugging timm model

* Fixing bugs in Gemma3nForConditionalGeneration for text generation

* Generating the modeling_gemma3n.py file

* Removing the addition of an erroneous line in the modeling file

* Adding gemma3n text model to modeling_auto

* Bugfix: Updating the interleaving of inputs_embeds and vision_embeds

* Updating the modeling file with the latest bugfix changes

* Updating models/auto for Gemma 3n

* using AutoTokenizer in forward test

* Adding processing_gemma3n.py

* Gemma 3n configured for AutoModel. Conversion script updated.

* Removing errant merge artifacts

---------

Co-authored-by: Mayank Chaturvedi <imayank@google.com>
Co-authored-by: Douglas Reid <douglas-reid@users.noreply.github.com>
Co-authored-by: Douglas Reid <21148125+douglas-reid@users.noreply.github.com>
Co-authored-by: Xuan-Son Nguyen <thichthat@gmail.com>
Co-authored-by: Sindhu Raghuram <sindhuraghuram@google.com>

* Removing errant debugging statements from Gemma 3

* Gemma3n audio model (#18)

* testing utilities for numerics comparisons

* Implement CumulativeGroupNorm and add to SubSampleConvProjection and SSCPConvBlock

* Add audio version of forward script based on RyanMullins' implementation

* Updating to match encoder tests. WIP: config question needs resolving

* Updates to audio classes to enable end-to-end running

* Removing vestigial classes, cleaning up print statements

* Adding SiLU / Swish to audio conformer feed forward block

* Shifted Gemma3p5Audio naming prefix to Gemma3NanoAudio

* Adding outputs to audio test

* Fixes to padding in SSCP and 1D convolution, align RMS Norm with wider model

* Update forward test to load from local weights

* Update conversion to process / output audio layers

* Update __all__ to export audio encoder

* AutoModel registration for Gemma 3n Audio

* Use AutoModel for ConditionalGeneration.audio_tower

* Fixing input_proj_linear transpose

* Fixing Gemma3NanoAudioConformerAttention.post conversion

* Fixing Gemma3NanoAudioSSCPConvBlock.conv weights conversion

* Correcting indentation issue on Gemma3p5RMSNorm

---------

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Text + Vision Part 2 (#23)

* Updates for ConditionalGeneration.get_image_features

* Adding a WIP draft of image_processing_gemma3p5.py

* Update src/transformers/models/gemma3p5/modular_gemma3p5.py

Co-authored-by: SindhuRaghuram97 <114270661+SindhuRaghuram97@users.noreply.github.com>

* Modular conversion after github suggested change

* Text + image gives good results

* Fixing image size preset

* Updating configs for the 2B variant in the conversion script

* Using final generation config in conversion script

---------

Co-authored-by: Sindhu Raghuram <sindhuraghuram@google.com>
Co-authored-by: SindhuRaghuram97 <114270661+SindhuRaghuram97@users.noreply.github.com>

* Audio Integration (#12)

* initial commit of Gemma 3n scaffold

* Fixing param pass through on Gemm3nRMSNorm

* Adds Einsum layer to Gemma 3n

* Updating EinsumLayer API

* Undoing erroneous force push

* Reverting RMSNorm to with_scale by default

* Adds LAuReL to Gemma 3n

* Adds AltUp to Gemma 3n

* Adding Gemma 3n overall and text config with vision and audio config placeholders (#3)

* Adding Gemma 3n text configs

* Adding audio config placeholders

* Adding a placeholder for vision configs

* Updating MobileNetVisionConfig, inheriting TimmWrapperConfig

* Updating text configs

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Removing altup configs to accept the suggested configs

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Updating altup config

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Addressing review comments and updating text configs

* Adding a config for activation sparsity

* Updating configs to pass through options to super class init and adjust some name prefixes

* Updating laurel and altup with corrected config values

* Normalizing sub_config initializers

---------

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Updating MLP with activation sparsity (#2)

* Updating DecoderBlock for Gemma 3n (#3)

* Initial Gemma3nTextModel (#4)

NOTE: This implementation WILL CHANGE in the coming weeks, however, changes will be strictly additive and this will remain a suitable baseline for downstream implementations to reference.

* Adding KV Cache Sharing

* Adds Einsum layer to Gemma 3n

* Updating EinsumLayer API

* Refactored kv cache sharing in attention

* Adding KVStore for cache sharing

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Update modular

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Update src/transformers/cache_utils.py

Co-authored-by: Ryan Mullins <ryanmullins@google.com>

* Undoing erroneous force push

* Reverting RMSNorm to with_scale by default

* Adds LAuReL to Gemma 3n

* Updating KV Cache Sharing implementation

* Updating the q and k norm definitions in the attention module

* Fixing name error for q,k,v RMS norm to use the right 3n module

* Updating MLP with activation sparsity

* Updating DecoderBlock for Gemma 3n

* Updating kv cache sharing implementation with the use of a cache buffer and refactoring some lines of code

* Isolating KV Cache logic to relevant components

* Fixing logic error in Gemma3nAttention.forward

* Refactoring caching contributions and fixing kv_store initialization

* Simplifying Configs

* Remove errant self from super init call

* Bug fix in the Attention module - changing self.head_dim to config.head_dim

* Bug fixes in the LaurelBlock and RMS Norm super init call

* removing redundant code from a merge

* Adding per_layer_inputs to TextModel

* Adding preprocess embeddings with altup

* Adds per-layer-to-single output and a host of TODOs

* Integrating altup predict with the model workflow and other minor bug fixes

* Using nn.Embedding temporarily for text model

* It goes forward

* Minor refactor of attention sparsity and RoPE initialization

* Fixing duplicate rope_scaling param bug when loading from pretrained

---------

Co-authored-by: Sindhu Raghuram <sindhuraghuram@google.com>
Co-authored-by: SindhuRaghuram97 <114270661+SindhuRaghuram97@users.noreply.github.com>

* Normalizing on altup_num_inputs config option

* Adding audio encoder config

* Adds high-level components for Audio Encoder

* Implement uniform reducer for Audio Encoder

* Adding placeholders for Conformer components in Audio Encoder

* Adding placeholders for SubSampleConvProjection components in Audio Encoder

* Adding SequenceLayer component placeholders

* Implementing Gemma3nAudioEncoder with nn.Sequential

* Implementing Gemma3nAudioSubSampleConvProjection with nn.Sequential

* Implementing Conformer model with SequenceLayers

* Use OrderedDict in nn.Sequential initializers

* Implements sl.Residual in Torch with nn.Sequential and OrderedDict

* Adopting a base SequenceLayer class with default forward() method

* Implementing sl.GatedLinearUnit in Torch

* Implementing sl.Swish in Torch

* Implementing sl.ReLU in Torch

* Implementing sl.Scale in Torch

* Removing sl.Dropout after tree-shaking

* Implementing sl.RMSNorm in Torch with fake shape

* Implementing sl.GroupNorm in Torch

* Implementing sl.Conv2d in Torch

* Implementing sl.Dense in Torch

* Removing sl.Delay layers, which act as pass-throughs

* Connecting shapes to configs in initializers

* Removing sl.Emit

* Implementing sl.ExpandDims in Torch

* Adding sl.GradientClipping to Torch

* Implementing sl.DenseShaped in Torch

* Implementing sl.LDPA in Torch

* Removing unused sl.CombinedQKVProj class

* Fixing erroneous type hint

* Implemnenting sl.DepthwiseConv1D in Torch

* Implementing sl.MaskInvalid in Torch

* Fixes for initialization

* Fixes for saving weights

* Removing einsums per feedback from HF staff

* Removing Sequence Layers idioms from audio encoder

* Fixes for reviewer comments

* Converting sl.Frontend to FeatureExtractor

* Updates for ConditionalGeneration.get_image_features

* Adding a WIP draft of image_processing_gemma3n.py

* Update modular

Co-authored-by: SindhuRaghuram97 <114270661+SindhuRaghuram97@users.noreply.github.com>

* Modular conversion after github suggested change

* Text + image gives good results

* Fixing image size preset

* Draft of audio data in chat template

* Removing image processing. Using SigLIP instead.

* Audio input going end-to-end

* Fixing dtype issues in audio encoder

* x-lib formatting consistency

* Adding example data

* Save preprocessor_config.json from conversion script

* Instrumentaiton for debugging

* Additional instrumentation for preprocessing debugging

* Updates to preprocessor, padding; produces correct end-to-end results on sample

* Tackling configuraiton TODOs

* Start of feature extractor refatcor

* Adds Numpy version of USM extractor, removes Torch version and dependencies

* Fixing AltUp.correct coef permute

* Supporting batches of single audio segment inputs

* Docstrings updates for config

* In-lining audio feature extraction

* Adjustments to conversion script and smoke test script

---------

Co-authored-by: SindhuRaghuram97 <114270661+SindhuRaghuram97@users.noreply.github.com>
Co-authored-by: Sindhu Raghuram <sindhuraghuram@google.com>
Co-authored-by: pculliton <phillipculliton@gmail.com>

* Gemma 3n renaming

* Removing test data and utilities

* Renaming test files

* Gemma 3n refactor

* Fix tokenizer config in conversion script

* Address reviewer feedback

* FeatureExtractor returns float32 by default

* Adding basic tests for audio, and input name for audio encoder

* Audio integration test, updates to model_id for other integration tests

* Use scales for q and k norms (#26)

* Update audio integration test to use HF dataset

* Reviewer feedback

* Expand embedding table to full vocab size in weights conversion

* Mix-n-match MatFormers for Gemma 3n (#25)

* Remove in-place operations (#30)

* chore: removing inplace ops

* remove [tensor] * n pattern

* chore: reviewer feedback in AudioEncoder and AltUp

* More grad clipping

* Dynamo compatibility

* fix: cache slicing error

* chore: simplify shared kv cache slicing

* chore: vision encoder rename in timm

* fix: image processor do_normalize=False

* fixup: style

* chore: model_doc

* fix: docs for code quality

* chore: repo consistency

* fix: RMSNorm in float as in prior Gemmas

* fix: per_layer_inputs = None

* chore: Gemma3nForCausalLM from Gemma3nForConditionalGeneration checkpoint

* chore: repo consistency

* Add initial unit tests for Gemma3nAudioFeatureExtractor (#27)

* Add initial unit tests for Gemma3nAudioFeatureExtractor

* Add basic unit tests for Gemma3nProcessor (#28)

Co-authored-by: Douglas Reid <21148125+douglas-reid@users.noreply.github.com>

* parameterize tests

---------

Co-authored-by: Douglas Reid <21148125+douglas-reid@users.noreply.github.com>

* chore: code style

* fix: test cases

* style and consistency

* fix config in the test to be coherent with layer cache sharing

* fix hidden states in tests and code

* inits and mappings

* fix modality prefixes

* test order and prefixes

* fix test exception

* fix class order and reduce model size for faster tests

* restore _checkpoint_conversion_mapping to load Caual from Conditional

* fix config mapping!

* fix: reviewer feedback

---------

Co-authored-by: SindhuRaghuram97 <114270661+SindhuRaghuram97@users.noreply.github.com>
Co-authored-by: Sindhu Raghuram <sindhuraghuram@google.com>
Co-authored-by: raushan <raushan@huggingface.co>
Co-authored-by: Mayank Chaturvedi <imayank@google.com>
Co-authored-by: Douglas Reid <douglas-reid@users.noreply.github.com>
Co-authored-by: Douglas Reid <21148125+douglas-reid@users.noreply.github.com>
Co-authored-by: Xuan-Son Nguyen <thichthat@gmail.com>
Co-authored-by: pculliton <phillipculliton@gmail.com>
Co-authored-by: Aritra Roy Gosthipaty <aritra.born2fly@gmail.com>
Co-authored-by: Cyril Vallez <cyril.vallez@gmail.com>

* fix import test

* add model args

* auto_docstring

* replace test path

* consistency

* skip tests for now

* fix docstring for doc builder

* skip unused attr

---------

Co-authored-by: SindhuRaghuram97 <114270661+SindhuRaghuram97@users.noreply.github.com>
Co-authored-by: Sindhu Raghuram <sindhuraghuram@google.com>
Co-authored-by: raushan <raushan@huggingface.co>
Co-authored-by: Mayank Chaturvedi <imayank@google.com>
Co-authored-by: Douglas Reid <douglas-reid@users.noreply.github.com>
Co-authored-by: Douglas Reid <21148125+douglas-reid@users.noreply.github.com>
Co-authored-by: Xuan-Son Nguyen <thichthat@gmail.com>
Co-authored-by: pculliton <phillipculliton@gmail.com>
Co-authored-by: Aritra Roy Gosthipaty <aritra.born2fly@gmail.com>
Co-authored-by: Cyril Vallez <cyril.vallez@gmail.com>
Co-authored-by: Arthur <arthur.zucker@gmail.com>
2025-06-26 17:55:47 +02:00
Joao Gante
3e5cc12855
[tests] remove tests from libraries with deprecated support (flax, tensorflow_text, ...) (#39051)
* rm tf/flax tests

* more flax deletions

* revert fixture change

* reverted test that should not be deleted; rm tf/flax test

* revert

* fix a few add-model-like tests

* fix add-model-like checkpoint source

* a few more

* test_get_model_files_only_pt fix

* fix test_retrieve_info_for_model_with_xxx

* fix test_retrieve_model_classes

* relative paths are the devil

* add todo
2025-06-26 16:25:00 +01:00
eustlb
cfff7ca9a2
[Whisper] Pipeline: handle long form generation (#35750)
* handle long form generation

* add warning

* correct incorrect in place token change

* update test to catch edge case

* make style

* update warning

* add doc
2025-06-26 14:33:31 +00:00
eustlb
02ecdcfc0f
add _keep_in_fp32_modules_strict (#39058)
* add _keep_in_fp32_modules_strict

* complete test
2025-06-26 13:55:28 +00:00
vb
d973e62fdd
fix condition where torch_dtype auto collides with model_kwargs. (#39054)
* fix condition where torch_dtype auto collides with model_kwargs.

* update tests

* update comment

* fix

---------

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
2025-06-26 14:52:57 +02:00
Raushan Turganbay
44b231671d
[qwen2-vl] fix vision attention scaling (#39043)
scale lost its `-` when refactoring
2025-06-26 14:06:52 +02:00
emmmm
ae15715df1
polishing docs: error fixes for clarity (#39042)
* fix duplicate deprecate_models.py

* fix duplicate modular_model_converter.py
2025-06-26 11:56:31 +00:00
Manuel de Prada Corral
3abeaba7e5
Create test for #38916 (custom generate from local dir with imports) (#39015)
* create test for #38916 (custom generate from local dir with imports)
2025-06-26 13:54:36 +02:00