* fix
* nice
* where i am at
* Bro this works
* Update src/transformers/integrations/tensor_parallel.py
* cleanups
* yups that was breaking
* Update src/transformers/models/openai_moe/modeling_openai_moe.py
* gather on experts and not mlp
* add changes for latest convert branch
* adds options to get output_router_logits from config
* bring chat temlate + special tokens back into the script.
* initial commmit
* update
* working with shards
* add model.safetensors.index.json
* fix
* fix
* mxfp4 flag
* rm print
* Fix PAD/EOS/BOS (#18)
* fix pad/eos/bos
* base model maybe one day
* add some doc
* special tokens based on harmony.
* add in tokenizer config as well.
* prepare for rebase with main
* Fix for initialize_tensor_parallelism now returning 4-tuple
```
[rank0]: File "/fsx/edward/work/openai-tsm-examples/examples/generate.py", line 17, in <module>
[rank0]: model = AutoModelForCausalLM.from_pretrained(
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/fsx/edward/work/new-model-addition-openai/src/transformers/models/auto/auto_factory.py", line 600, in from_pretrained
[rank0]: return model_class.from_pretrained(
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/fsx/edward/work/new-model-addition-openai/src/transformers/modeling_utils.py", line 316, in _wrapper
[rank0]: return func(*args, **kwargs)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/fsx/edward/work/new-model-addition-openai/src/transformers/modeling_utils.py", line 4748, in from_pretrained
[rank0]: tp_plan, device_map, device_mesh = initialize_tensor_parallelism(tp_plan, tp_size=None)
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: ValueError: too many values to unpack (expected 3)
```
* mxfp4
* mxfp4 draft
* fix
* fix import
* draft
* draft impl
* finally working !
* simplify
* add import
* working version
* consider blocks and scales
* device mesh fix
* initial commit
* add working dequant + quant logic
* update
* non nan, gibberish output
* working EP + quantization finally !
* start cleaning
* remove reversing process
* style
* some cleaning
* initial commmit
* more cleaning
* more cleaning
* simplify
* more cleaning
* rm duplicated function
* changing tp_plan
* update tp plan check
* add loading attribute
* dequantizing logic
* use subfunctions
* import cleaning
* update_param_name
* adds clamped swiglu
* add clamping to training path
* simplify dequant logic
* update
* Bad merge
* more simplifications & tests
* fix !
* fix registering custom attention
* fix order
* fixes
* some test nits
* nits
* nit
* fix
* Clamp sink logits
* Clean
* Soft-max trick
* Clean up
* p
* fix deepspeed
* update both modeling and modular for cleanup
* contiguous
* update tests
* fix top_k router call
* revert renaming
* test nits
* small fixes for EP
* fix path for our local tests
* update as I should not have broken that!
* fix the loss of mixtral
* revert part of the changes related to router_scores, kernel probably no ready for that!
* deleting a small nit
* update arch
* fix post processing
* update
* running version but not expected output
* moving to cuda
* initial commit
* revert
* erroring when loading on cpu
* updates
* del blocks, scales
* fix
* style
* rm comm
* comment
* add comment
* style
* remove duplicated lines
* Fix minor issue with weight_map conversion script
* fix sampling params
* rename to final name
* upate pre-final version of template
* Update src/transformers/models/gpt_oss/convert_gpt_oss_weights_to_hf.py
* fix batched inference
* serve fixes
* swizzle !
* update final chat template by Matt.
* fix responses; pin oai
* sinplify
* Thanks Matt for his tireless efforts!
Co-authored-by: Rocketknight1 <Rocketknight1@users.noreply.github.com>
* Update src/transformers/models/gpt_oss/convert_gpt_oss_weights_to_hf.py
Co-authored-by: Matt <Rocketknight1@users.noreply.github.com>
* fix
* Use ROCm kernels from HUB
* Make kernel modes explicit
* update final chat template by Matt. x2
* Thanks Matt for his tireless efforts!
Co-authored-by: Rocketknight1 <Rocketknight1@users.noreply.github.com>
* Fix installation
* Update setup.py
Co-authored-by: Ákos Hadnagy <akos.hadnagy@gmail.com>
* allow no content
* fix: update message handling in write_tokenizer function
* Fix template logic for user message role
* last nits for CB and flash_paged!
* there was one bad merge
* fix CB (hardcode for now, its just using kv groups instead)
* fix
* better fix for device_map
* minor device fix
* Fix flash paged
* updates
* Revert "remove dtensors, not explicit (#39840)"
This reverts commit 6dfd561d9c.
* update
* Revert "remove dtensors, not explicit (#39840)"
This reverts commit 6dfd561d9c.
* fix merge
* fix
* Fix line break when custom model indentity
* nits testing
* to locals first and pass sliding window to flash paged
* register modes for MegaBlocksMoeMlp
* add integration test in fixtures -> now update the tests to use it!
* update integration tests
* initial fix
* style and update tests
* fix
* chore(gpt oss): remove mlp_bias from configuration
It was just a leftover.
* stats
* Integration tests
* whoops
* Shouldn't move model
* Ensure assistant messages without thinking always go to "final" channel
* More checks to ensure expected format
* Add pad_token_id to model configuration in write_model function (#51)
* Add oai fix fast tests (#59)
* Fix some fast tests
* Force some updates
* Remove unnecessary fixes
* Update src/transformers/models/gpt_oss/convert_gpt_oss_weights_to_hf.py
Co-authored-by: Quentin Gallouédec <45557362+qgallouedec@users.noreply.github.com>
* Update src/transformers/models/gpt_oss/convert_gpt_oss_weights_to_hf.py
Co-authored-by: Quentin Gallouédec <45557362+qgallouedec@users.noreply.github.com>
* Update src/transformers/models/gpt_oss/convert_gpt_oss_weights_to_hf.py
* reasoning -> Reasoning
* Add additional integration tests
* fixup
* Slight fixes
* align chat template with harmony
* simplify
* Add comment
* torch testing assert close
* torch testing assert close
* torch testing assert close
* torch testing assert close
* torch testing assert close
* torch testing assert close
* Revert fixup
* skip 2 test remove todo
* merge
* padding side should be left for integration tests
* fix modular wrt to changes made to modeling
* style
* isort
* fix opies for the loss
* mmmm
---------
Co-authored-by: Quentin Gallouédec <gallouedec.quentin@gmail.com>
Co-authored-by: Quentin Gallouédec <45557362+qgallouedec@users.noreply.github.com>
Co-authored-by: Marc Sun <marc@huggingface.co>
Co-authored-by: edbeeching <edbeeching@gmail.com>
Co-authored-by: Vaibhavs10 <vaibhavs10@gmail.com>
Co-authored-by: MekkCyber <mekk.cyber@gmail.com>
Co-authored-by: Marc Sun <57196510+SunMarc@users.noreply.github.com>
Co-authored-by: Edward Beeching <edbeeching@users.noreply.github.com>
Co-authored-by: Mohamed Mekkouri <93391238+MekkCyber@users.noreply.github.com>
Co-authored-by: Lewis Tunstall <lewis.c.tunstall@gmail.com>
Co-authored-by: Zhuohan Li <zhuohan@openai.com>
Co-authored-by: Pedro Cuenca <pedro@huggingface.co>
Co-authored-by: joao@huggingface.co <joao@ip-10-53-88-32.ec2.internal>
Co-authored-by: Rocketknight1 <Rocketknight1@users.noreply.github.com>
Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>
Co-authored-by: Akos Hadnagy <akos@ahadnagy.com>
Co-authored-by: Ákos Hadnagy <akos.hadnagy@gmail.com>
Co-authored-by: Alvaro Moran <alvaro.moran@huggingface.co>
Co-authored-by: Lysandre <hi@lysand.re>
Co-authored-by: Matt <rocketknight1@gmail.com>
* Update model card for DETR
* fix: applied suggested changes
* fix: simplified pipeline and modified notes and resources
* Update detr.md
---------
Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
* added code for handling video object ,as dictionary of frames and metadata, in chat template
* added new test where videos are passed as objects (dict of frames, metadata) in the chat template
* modified hardcoded video_len check that does not match with increased number of tests cases.
* Modify hardcoded video_len check that fails with increased number of tests
* update documentation of multi-modal chat templating with extra information about including video object in chat template.
* add array handling in load_video()
* temporary test video inlcuded
* skip testing smolvlm with videos that are list of frames
* update documentation & make fixup
* Address review comments
* fix: deprecate plot_keypoint_matching and make visualize_keypoint_matching for all Keypoint Matching models
* refactor: added copied from
* fix: make style
* fix: repo consistency
* fix: make style
* docs: added missing method in SuperGlue docs
* first commit
Added modular implementation for MM Grounding DINO from starting point created by add-new-model-like. Added conversion script from mmdetection to huggingface.
TODO: Some tests are failing so that needs to be fixed.
* fixed a bug with modular definition of MMGroundingDinoForObjectDetection where box and class heads were not correctly assigned to inner model
* cleaned up a hack in the conversion script
* Fixed the expected values in integration tests
Cross att masking and cpu-gpu consistency tests are still failing however.
* changes for make style and quality
* add documentation
* clean up contrastive embedding
* add mm grounding dino to loss mapping
* add model link to config docstring
* hack fix for mm grounding dino consistency tests
* add special cases for unused config attr check
* add all models and update docs
* update model doc to the new style
* Use super_kwargs for modular config
* Move init to the _init_weights function
* Add copied from for tests
* fixup
* update typehints
* Fix-copies for tests
* fix-copies
* Fix init test
* fix snippets in docs
* fix consistency
* fix consistency
* update conversion script
* fix nits in readme and remove old comments from conversion script
* add license
* remove unused config args
* remove unnecessary if/else in model init
* fix quality
* Update references
* fix test
* fixup
---------
Co-authored-by: qubvel <qubvel@gmail.com>
* Add cohere2_vision to support CohereLabs/command-a-vision-07-2025
* update and add modualr file
* update processors and check with orig impl later
* delete unused files
* image processor reduce LOC and re-use GotOCR2
* update the config to use modular
* model tests pass
* processor fixes
* check model outputs decorator
* address one more comment
* Update tokens. Temp - need to read from tokenizer'
* fix for multi-gpu
* Fix image token handling
* upadte image token expansion logic
* fix a few issues with remote code loading
* not related but modular forces us to change all files now
* Add overview and code sample to cohere vision docs
* add scripts. TMP.
* Update inference script
* Create script
* set dtype in export script
* TO revert: modular export fix
* Fix scripts
* Revert "TO revert: modular export fix"
This reverts commit bdb2f305b61027a05f0032ce70d6ca698879191c.
* Use modular weights
* Upload to hub
Removed OOD weights ad script
* Updated docs
* fix import error
Update docs
Added pipeline test
* Updated docs
* Run modular script
remove modular for config
Added patch_size
Added docstrings in modular
Fix OOM
Add docs, fixup integration tests. 8-gpu passing
* tiny updates
* address comments + fixup
* add test for chat template
* check model outputs workaround
* aya vision fix check model inputs
* Revert "add test for chat template"
This reverts commit 42c756e397f588d76b449ff1f93292d8ee0202d8.
* reveert more changes
* last revert
* skip and merge
* faulty copy from
---------
Co-authored-by: Julian Mack <julian.mack@cohere.com>
Co-authored-by: kyle-cohere <kyle@cohere.com>
* Add Fast Segformer Processor
* Modified the params according to segformer model
* modified test_image_processing_Segformer_fast args
- removed redundant params like do_center_crop,center_crop which aren't present in the original segformer class
* added segmentation_maps processing logic form the slow segformer processing module with references from beitimageprocessing fast
* fixed code_quality
* added recommended fixes and tests to make sure everything processess smoothly
* Fixed SegmentationMapsLogic
- modified the preprocessing of segmentation maps to use tensors
- added batch support
* fixed some mismatched files
* modified the tolerance for tests
* use modular
* fix ci
---------
Co-authored-by: yonigozlan <yoni.gozlan@huggingface.co>
* feat: superpoint fast image processor
* fix: reran fast cli command to generate fast config
* feat: updated test cases
* fix: removed old model add
* fix: format fix
* Update src/transformers/models/superpoint/image_processing_superpoint_fast.py
Co-authored-by: Yoni Gozlan <74535834+yonigozlan@users.noreply.github.com>
* fix: ported to torch and made requested changes
* fix: removed changes to init
* fix: init fix
* fix: init format fix
* fixed testcases and ported to torch
* fix: format fixes
* failed
test case fix
* fix superpoint fast
* fix docstring
---------
Co-authored-by: Yoni Gozlan <74535834+yonigozlan@users.noreply.github.com>
Co-authored-by: yonigozlan <yoni.gozlan@huggingface.co>
* upload initial code
* update deepseek-vl adaptor
* update hierarchy of vision model classes
* udpate aligner model
* add text model
* Added Image Processor
* Added Image Processor
* Added Image Processor
* apply masks
* remove projection; add aligner
* remove interpolate_pos_encoding
* remove unused params in config
* cleaning
* Add the __init__ file
* added processing deepseek_vl class
* modified the deepseek-vl processor
* modified the deepseek-vl processor
* update __init__
* Update the image processor class name
* Added Deepseek to src/transformers/__init__.py file
* Added Deepseek to image_processing_auto.py
* update the __init__ file
* update deepseek_vl image processor
* Update Deepseek Processor
* upload fast image processor
* Revert "upload fast image processor"
This reverts commit 68c8fd50bafbb9770ac70c9de02448e2519219b4.
* update image processor
* flatten heirarchy
* remove DeepseekVLModel
* major update (complete modeling)
* auto modeling and other files
* formatting
* fix quality
* replace torchvision in modeling
* set default do_normalize to False
* add fast image processor template using tool
* update image processors
* add fast image processor to other files
* update liscense
* Added deepseek image testcases
* update image test
* update processor
* write CHAT_TEMPLATE
* update model for processor
* fix processor
* minor fixes and formatting
* fix image processing and tests
* fix interpolation in sam
* fix output_attentions in DeepseekVLModel
* upload test_modeling
* fix tests because of vocab size
* set use_high_res_vision=False in tests
* fix all modeling tests
* fix styling
* remove explicit background_color from image processors
* added test_processor
* added test_processor
* fix processor tests
* update docs
* update docs
* update docs
* update conversion script
* Fixed typos
* minor fixes from review
- remove model_id comments in examples
- remove from pre-trained auto mapping
- move to image-text-to-text from vision-to-seq in auto mapping
- add image_token_index to __init__ for config
- remove outdated temporary config in conversion script
- update example to use chat_template in docstring example
- update liscense 2021->2025
* fix type in config docstring
Co-authored-by: Raushan Turganbay <raushan.turganbay@alumni.nu.edu.kz>
* update get_image_features
* fix config
* improve DeepseekVLImageProcessor.preprocess
* return image_hidden_states
* use AutoTokenizer and AutoImageProcessor in Processor
* fix model outputs
* make num_image_tokens configurable
* fix docstring of processor
* move system prompt to chat template
* fix repo consistency
* fix return_dict
* replace SamVisionEncoder with SamVisionModel
* update to remove deepcopy
* 🛠️ Major Architectural Changes (Adds DeepseekVLHybrid)
* fix quality checks
* add missing hybrid in auto modeling
* run make style
* update sam_hq
* update high_res_size in test
* update docs following #36979
* update code with auto_docstring
* update conversion scripts
* fix style
* fix failing test because of tuple
* set weights_only=True in conversion script
* use safetensors.torch.load_file instead of torch.load in conversion script
* make output_dir optional in conversion script
* fix code snippets in docs (now the examples work fine)
* integration tests for DeepseekVL
* update expected texts
* make style
* integration tests for DeepseekVLHybrid
* fix class name
* update expected texts for hybrid
* run "make style"
* update since changes in main
* run make-style
* nits since changes in main
* undo changes in sam
* fix tests
* fix tests; update with main
* update with main: output_attention/output_hidden_states
* fix copied part in deepseek_vl
* run fix-copies
* fix output_hidden_states
* sam: fix _init_weigths
* use modular for DeepseekVL
* make image processor more modular
* modular: use JanusPreTrainedModel
* janus: provide kwargs in loss
* update processors in conversion script
* Revert "sam: fix _init_weigths"
This reverts commit db625d0c68956c0dad45edd7a469b6a074905c27.
* run fix-copies
---------
Co-authored-by: Shakib-IO <shakib.khan17@northsouth.edu>
Co-authored-by: Raushan Turganbay <raushan.turganbay@alumni.nu.edu.kz>
* add owlv2 fast image processor
* add Owlv2ImageProcessorFast to Owlv2Processor image_processor_class
* add Owlv2ImageProcessorFast to Owlv2Processor image_processor_class
* change references to owlVit to owlv2 in docstrings for post process methods
* change type hints from List, Dict, Tuple to list, dict, tuple
* remove unused typing imports
* add disable grouping argument to group images by shape
* run make quality and repo-consistency
* use modular
* fix auto_docstring
---------
Co-authored-by: Lewis Marshall <lewism@elderda.co.uk>
Co-authored-by: yonigozlan <yoni.gozlan@huggingface.co>
* docs: Standardize OPT model card with enhanced details
* Remove incorrect link from OPT model card
* Address review feedback on OPT model card
* Update opt.md
---------
Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
- Fix Cyrillic 'Р' to Latin 'P' in Portuguese language link (README.md)
- Fix 'meanginful' to 'meaningful' in training documentation
- Fix duplicate 'Cohere' reference in modular transformers documentation
- Fix duplicate 'the the' in trainer and chat command comments
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-authored-by: Claude <claude@anthropic.com>
Co-authored-by: Claude <noreply@anthropic.com>
* First attempt
* fix
* fix
* Enhance TrackioCallback to log GPU memory usage and allocation
* Enhance Trackio integration in callbacks and training arguments documentation
* re order
* remove unused lines
* fix torch optional
* initial commit
* Apply suggestions from code review
Co-authored-by: Pavel Iakubovskii <qubvel@gmail.com>
* fix: various typos, typehints, refactors from suggestions
* fix: fine_matching method
* Added EfficientLoFTRModel and AutoModelForKeypointMatching class
* fix: got rid of compilation breaking instructions
* docs: added todo for plot
* fix: used correct hub repo
* docs: added comments
* fix: run modular
* doc: added PyTorch badge
* fix: model repo typo in config
* fix: make modular
* fix: removed mask values from outputs
* feat: added plot_keypoint_matching to EfficientLoFTRImageProcessor
* feat: added SuperGlueForKeypointMatching to AutoModelForKeypointMatching list
* fix: reformat
* refactor: renamed aggregation_sizes config parameter into q, kv aggregation kernel size and stride
* doc: added q, kv aggregation kernel size and stride doc to config
* refactor: converted efficientloftr implementation from modular to copied from mechanism
* tests: overwrote batching_equivalence for "keypoints" specific tests
* fix: changed EfficientLoFTRConfig import in test_modeling_rope_utils
* fix: make fix-copies
* fix: make style
* fix: update rope function to make meta tests pass
* fix: rename plot_keypoint_matching to visualize_output for clarity
* refactor: optimize image pair processing by removing redundant target size calculations
* feat: add EfficientLoFTRImageProcessor to image processor mapping
* refactor: removed logger and updated attention forward
* refactor: added auto_docstring and can_return_tuple decorators
* refactor: update type imports
* refactor: update type hints from List/Dict to list/dict for consistency
* refactor: update MODEL_MAPPING_NAMES and __all__ to include LightGlue and AutoModelForKeypointMatching
* fix: change type hint for size parameter in EfficientLoFTRImageProcessor to Optional[dict]
* fix typing
* fix some typing issues
* nit
* a few more typehint fixes
* Remove output_attentions and output_hidden_states from modeling code
* else -> elif to support efficientloftr
* nit
* tests: added EfficientLoFTR image processor tests
* refactor: reorder functions
* chore: update copyright year in EfficientLoFTR test file
* Use default rope
* Add docs
* Update visualization method
* fix doc order
* remove 2d rope test
* Update src/transformers/models/efficientloftr/modeling_efficientloftr.py
* fix docs
* Update src/transformers/models/efficientloftr/image_processing_efficientloftr.py
* update gradient
* refactor: removed unused codepath
* Add motivation to keep postprocessing in modeling code
* refactor: removed unnecessary variable declarations
* docs: use load_image from image_utils
* refactor: moved stage in and out channels computation to configuration
* refactor: set an intermediate_size parameter to be more explicit
* refactor: removed all mentions of attention masks as they are not used
* refactor: moved position_embeddings to be computed once in the model instead of every layer
* refactor: removed unnecessary hidden expansion parameter from config
* refactor: removed completely hidden expansions
* refactor: removed position embeddings slice function
* tests: fixed broken tests because of previous commit
* fix is_grayscale typehint
* not refactoring
* not renaming
* move h/w to embeddings class
* Precompute embeddings in init
* fix: replaced cuda device in convert script to accelerate device
* fix: replaced stevenbucaille repo to zju-community
* Remove accelerator.device from conversion script
* refactor: moved parameter computation in configuration instead of figuring it out when instantiating a Module
* fix: removed unused attributes in configuration
* fix: missing self
* fix: refactoring and tests
* fix: make style
---------
Co-authored-by: steven <steven.bucaille@buawei.com>
Co-authored-by: Pavel Iakubovskii <qubvel@gmail.com>
* init
* copied from remote
* add proper structure and llama like structure
* fixup
* revert to state that works
* get closer to llama
* slow and steady
* some removal
* masks work
* it is indeed the rope implementation, how dafuq does it mesh with the cache now hmm
* nice
* getting closer
* closer to transformers style
* let's simplify this, batching works now
* simplified
* working version with modular
* it is indeed the rotation per weights, make it complete llama style
* cleanup conversion, next to look at -> tokenizer
* remove llama artefacts
* fix modeling tests (common ones)
* style
* integration test + first look into tokenization (will need more work, focussing on modeling other models first)
* style
* working moe version, based on remote
* lets keep it simple and go step by step - transformers annotations for modular and transformers style rope (complex view)
* more cleanup
* refactor namings and remove addition forXXX classes
* our moe won't cut it it seems, correction bias seems to be missing in remote code version
* tokenization change (remote)
* our moe version works when adding normalization :D
* cleanup moe
* nits
* cleanup modeling -> let's get to modular next
* style
* modular v1
* minor things + attempt at conversion (which doesn't work)
* no conversion follow glm, fixup modular and other nits
* modular cleanup
* fixes
* tests, tests, tests + some moe dtype forcing
* simplify modular, fix fatal fa2 bug, remaining tests
* fix import issue?
* some initial docs, fix bnb faulty behavior --> needs to fix some tests because of gate needing to be float
* fix sdpa test, load on init dtype only
* fixup post merge
* style
* fix doc links
* tokenization cleanup beginnings
* simplify tokenizer by a lot as its basically llama
* tokenizer is full llama with different defaults + extra special tokens
* sync og special tokens of ernie
* fix decoding with numbers (also in remote done what a timing), begin of tok tests
* align with remote and preserve special tokens, adjust tests to ernie legacy behavior, warning for questionable behavior (also in llama)
* nits
* docs
* my daily post merge it is
* check
* tokenization update with explanations and conversion script
* review on modular (til), revert some tokenizer things i did prior, remove mtp comment (low prio)
* post merge fixes
* fixup tokenization, llama fast is the way to go
* more fixups
* check
* import fixes
* correction bias following the paddle code
* fix
* fix TP plan, fix correction bias sharding during forward
* style
* whoops
* fix tied weights
* docs and last nit
* license
* flasky tests
* move repo id, update when merged on the hub
