🧹 Remove deprecated RotaryEmbedding parts in the Attention layers #34858
Conversation
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
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
Cyrilvallez
changed the title
Cyrilvallez
changed the title
ArthurZucker
reviewed
Dec 4, 2024
Comment on lines
-317
to
-319
| kv_seq_len = key_states.shape[-2] | ||
| if past_key_value is not None: | ||
| kv_seq_len += cache_position[0] |
There was a problem hiding this comment.
Unwanted change I think
There was a problem hiding this comment.
Nop, this is actually doing nothing (never used, no purpose here) so I cleaned it up!
There was a problem hiding this comment.
Indeed not linked to the rotary though, but as I'm cleaning stuff, let's remove it as the same time
|
@ArthurZucker should be good to go! |
ArthurZucker
reviewed
Dec 5, 2024
ArthurZucker
approved these changes
Dec 11, 2024
|
Looks like there is a regression? |
Merged
|
True! Aria relies on Llama and was merged in the meantime - #35208 correctly reapplies new Llama change to Aria |
BenjaminBossan
added a commit
to BenjaminBossan/peft
that referenced
this pull request
Dec 12, 2024
This PR resolves the failing adoption prompt tests in the CI using transformers installed from source. In this transformers PR: huggingface/transformers#34858 the module.rotary_emb attribute has been removed, which adoption prompt so far assumed was present. Instead, the position_embeddings are now already computed and can be taken directly from the kwargs.
BenjaminBossan
added a commit
to BenjaminBossan/peft
that referenced
this pull request
Dec 12, 2024
This PR resolves the failing adoption prompt tests in the CI using transformers installed from source. In this transformers PR: huggingface/transformers#34858 the module.rotary_emb attribute has been removed, which adoption prompt so far assumed was present. Instead, the position_embeddings are now already computed and can be taken directly from the kwargs.
BenjaminBossan
added a commit
to huggingface/peft
that referenced
this pull request
Dec 13, 2024
This PR resolves the failing adoption prompt tests in the CI using transformers installed from source. In this transformers PR: huggingface/transformers#34858 the module.rotary_emb attribute has been removed, which adoption prompt so far assumed was present. Instead, the position_embeddings are now already computed and can be taken directly from the kwargs.
3 tasks
dsikka
added a commit
to vllm-project/llm-compressor
that referenced
this pull request
Feb 11, 2025
## Purpose ##
* SparseGPT
* Fix behavior where `targets` specifies which modules to sparsity, not
which layers to target
* Fix broken behavior with `_infer_owl_layer_sparsity` and add test
* Fix owl argument validation
* Add type hints and abstract methods for clarity
* Pipelines
* Fix bug revealed by decorators added to the llama model definition in
the latest transformers release
* huggingface/transformers#35757
* For the sequential pipeline, this revealed a bug in
torch.fx._symbolic_trace where wrapped functions were not being handled
properly
* Future work could involve upstreaming a bug fix
* Fix issue caused by changes to llama model definition
* huggingface/transformers#34858
* For the layer sequential pipeline, this challenges the assumption that
each layer input is the previous layer's output (which was known to be a
fragile assumption)
* Fix issue related to basic pipeline slowdowns and inaccuracy
## Changes ##
* SparseGPT
* Fully separate `targets` and `sequential_targets`
* Modify hooks adding logic to reflect this change
* Fix behavior of `_infer_owl_layer_sparsity` and add test
* Code clarity
* Add additional type hints
* Designate `calibrate_module` as an abstract method on the sgpt mixin
* Pipelines
* Sequential pipeline: unwrap model forward function to avoid issues
with pytorch function patching
* Layer Sequential Pipeline: Add `maybe_inject_pos_embeddings` to
sequential pipeline to hackily support models with `position_embeddings`
* Basic Pipeline: Fix `on_sequential_batch_end` to call on the end of
epoch, rather than every batch
* Calling every batch was likely causing slowdowns
## Followups ##
* Remove deprecated `sequential_update` option from examples and tests
## Testing ##
* Added `tests/llmcompressor/transformers/obcq/test_obcq_owl.py`
* Tested OBCQ+llama with sequential, layer sequential, and basic
pipelines independently
## Regression Evaluations ##
Models were compressed using
`examples/sparse_2of4_quantization_fp8/llama3_8b_2of4.py` without fp8
option
<details><summary>sparsegpt</summary>
Main
```
vllm (pretrained=/home/kyle/llm-compressor/Meta-Llama-3-8B-InstructSparseGPTModifierMAIN,dtype=bfloat16,add_bos_token=True), gen_kwargs: (None), limit: None, num_fewshot: 5, batch_size: 1
| Tasks |Version|Filter|n-shot|Metric| |Value | |Stderr|
|----------|------:|------|-----:|------|---|-----:|---|-----:|
|winogrande| 1|none | 5|acc |↑ |0.6243|± |0.0136|
```
This branch
```
vllm (pretrained=/home/kyle/llm-compressor/Meta-Llama-3-8B-InstructSparseGPTModifierFEATURE,dtype=bfloat16,add_bos_token=True), gen_kwargs: (None), limit: None, num_fewshot: 5, batch_size: 1
| Tasks |Version|Filter|n-shot|Metric| |Value | |Stderr|
|----------|------:|------|-----:|------|---|-----:|---|-----:|
|winogrande| 1|none | 5|acc |↑ |0.6306|± |0.0136|
```
</details>
To test wanda, the `SparseGPTModifier` was replaced with the
`WandaPruningModifier`
<details><summary>wanda</summary>
Main
```
vllm (pretrained=/home/kyle/llm-compressor/Meta-Llama-3-8B-InstructWandaPruningModifierMAIN,dtype=bfloat16,add_bos_token=True), gen_kwargs: (None), limit: None, num_fewshot: 5, batch_size: 1
| Tasks |Version|Filter|n-shot|Metric| |Value | |Stderr|
|----------|------:|------|-----:|------|---|-----:|---|-----:|
|winogrande| 1|none | 5|acc |↑ |0.5912|± |0.0138|
```
This branch
```
vllm (pretrained=/home/kyle/llm-compressor/Meta-Llama-3-8B-InstructWandaPruningModifierFEATURE,dtype=bfloat16,add_bos_token=True), gen_kwargs: (None), limit: None, num_fewshot: 5, batch_size: 1
| Tasks |Version|Filter|n-shot|Metric| |Value | |Stderr|
|----------|------:|------|-----:|------|---|-----:|---|-----:|
|winogrande| 1|none | 5|acc |↑ |0.5817|± |0.0139|
```
</details>
---------
Signed-off-by: Kyle Sayers <[email protected]>
Co-authored-by: Dipika Sikka <[email protected]>
Guy-Bilitski
pushed a commit
to Guy-Bilitski/peft
that referenced
this pull request
May 13, 2025
This PR resolves the failing adoption prompt tests in the CI using transformers installed from source. In this transformers PR: huggingface/transformers#34858 the module.rotary_emb attribute has been removed, which adoption prompt so far assumed was present. Instead, the position_embeddings are now already computed and can be taken directly from the kwargs.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
What does this PR do?
This cleans-up the (expired) deprecated cycle for the rotary embeddings and fully move them to the Model instead of the Attention. Also removes deprecated EmbeddingClasses, and fix the tests accordingly.