Daniel Korat fb0c6b521d Universal Assisted Generation: Assisted generation with any assistant model (by Intel Labs) (#33383) ...

* Update candidate_generator.py

* Update utils.py

* add lookbehind params to _get_candidate_generator

* make fixup

* add unit tests

* fix failing tests

* add docstrings

* fix docstrings; remove non-optimized AnyTokenizer

* added any tokenizer generation correctness test

* make fixup

* fix assertion syntax

* PR review fixes

* address additional PR comments

* fix tests

* remove stropping criteria arg

* make fixup

* add AssistantConfig

* fix prev_tokens branching

* pass tokenizers through `generate()`kwargs

* fix lookbehind values; tokenizer params WIP

* fixup

* AssistantConfig

* remove AssistantConfig; apply PR suggestions

* restructure tests

* fixup

* fix assistant_tokenizer arg validation

* fixup

* fix tests in TestAssistedCandidateGeneratorDifferentTokenizers

* fix class docstring

* PR suggestions

* doc

* doc update and improvements to `_validate_assistant()`

---------

Co-authored-by: mosheber <moshe.berchansky@intel.com>

2024-10-10 14:41:53 +02:00

..

agents

Decorator for easier tool building (#33439)

2024-09-18 11:07:51 +02:00

benchmark

[Test refactor 1/5] Per-folder tests reorganization (#15725)

2022-02-23 15:46:28 -05:00

bettertransformer

Fixed malapropism error (#26660)

2023-10-09 11:04:57 +02:00

deepspeed

Trainer - deprecate tokenizer for processing_class (#32385)

2024-10-02 14:08:46 +01:00

extended

[tests] skip tests for xpu (#33553)

2024-09-19 19:28:04 +01:00

fixtures

Implementation of SuperPoint and AutoModelForKeypointDetection (#28966)

2024-03-19 14:43:02 +00:00

fsdp

🚨🚨🚨 Update min version of accelerate to 0.26.0 (#32627)

2024-08-20 11:42:36 +02:00

generation

Universal Assisted Generation: Assisted generation with any assistant model (by Intel Labs) (#33383)

2024-10-10 14:41:53 +02:00

models

Phi3: fix attn for sliding window (#33586)

2024-10-10 11:50:39 +02:00

optimization

fix: Fixed the 1st argument name in classmethods (#31907)

2024-07-11 12:11:50 +01:00

peft_integration

[PEFT] Support low_cpu_mem_usage option for PEFT loading adapters (#33725)

2024-10-03 16:15:36 +02:00

pipelines

Sync QuestionAnsweringPipeline (#34039)

2024-10-10 13:38:14 +01:00

quantization

Add gguf support for gpt2 (#34044)

2024-10-10 13:42:18 +02:00

repo_utils

Refactor CI: more explicit (#30674)

2024-08-30 18:17:25 +02:00

sagemaker

Trainer - deprecate tokenizer for processing_class (#32385)

2024-10-02 14:08:46 +01:00

tokenization

Fix for slow the bug tokenizer adding spaces to single id decodes (#32564)

2024-09-18 12:32:02 +02:00

trainer

Fix PIL dep for tests (#34028)

2024-10-09 10:45:06 -04:00

utils

HfArgumentParser: allow for hyhenated field names in long-options (#33990)

2024-10-10 11:58:26 +02:00

__init__.py

…

test_backbone_common.py

Align backbone stage selection with out_indices & out_features (#27606)

2023-12-20 18:33:17 +00:00

test_configuration_common.py

Refactor: Removed un-necessary object base class (#32230)

2024-07-26 10:33:02 +02:00

test_feature_extraction_common.py

Split common test from core tests (#24284)

2023-06-15 07:30:24 -04:00

test_image_processing_common.py

Update kwargs validation for preprocess with decorator (#32024)

2024-08-06 11:33:05 +01:00

test_image_transforms.py

fix: center_crop occasionally outputs off-by-one dimension matrix (#30934)

2024-05-21 13:56:52 +01:00

test_modeling_common.py

Generate: remove most decoder-only LLMs prepare_inputs_for_generation (#33870)

2024-10-09 12:15:48 +01:00

test_modeling_flax_common.py

add sdpa to ViT [follow up of #29325] (#30555)

2024-05-16 10:56:11 +01:00

test_modeling_tf_common.py

[TF] Fix Tensorflow XLA Generation on limited seq_len models (#33903)

2024-10-05 16:20:50 +02:00

test_pipeline_mixin.py

Sync QuestionAnsweringPipeline (#34039)

2024-10-10 13:38:14 +01:00

test_processing_common.py

Uniformize model processors (#31368)

2024-10-02 10:41:08 +02:00

test_sequence_feature_extraction_common.py

Fix typo (#25966)

2023-09-05 10:12:25 +02:00

test_tokenization_common.py

Trainer - deprecate tokenizer for processing_class (#32385)

2024-10-02 14:08:46 +01:00