Skip to content

Pull requests: neuralmagic/compressed-tensors

New pull request New
26 Open 334 Closed
26 Open 334 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Deprecate iter_named_leaf_modules and iter_named_quantizable_modules
#381 opened Jul 7, 2025 by kylesayrs Loading…
1
Add torch.float64 as a viable dtype for scales
#379 opened Jul 3, 2025 by eldarkurtic Loading…
4
[Transforms] apply transforms to torch.nn.Embedding modules
#377 opened Jul 2, 2025 by brian-dellabetta • Draft
7
Added support for compression on meta device
#376 opened Jul 2, 2025 by shanjiaz Loading…
31
[WIP] [Research] Attention quantization and transformation
#374 opened Jul 1, 2025 by kylesayrs • Draft
Support DeepSeekV3-style block FP8 quantization
#372 opened Jun 30, 2025 by mgoin Loading…
@kylesayrs
15
[Transform] Serialize with tied weights
#370 opened Jun 28, 2025 by kylesayrs Loading…
1
[WIP] Initialize qparams with valid values
#368 opened Jun 23, 2025 by kylesayrs • Draft
[Accelerate] Remove is_module_offloaded and update_prefix_dict
#366 opened Jun 20, 2025 by kylesayrs • Draft
[Accelerate] Support get_offloaded_device for models
#364 opened Jun 20, 2025 by kylesayrs • Draft
[Bugfix] Fix updating parameters under align_modules
#361 opened Jun 18, 2025 by kylesayrs Loading…
Optimize sparse 2:4 compression performance
#358 opened Jun 16, 2025 by rahul-tuli • Draft
8 tasks done
[Transform] apply_transform_config
#348 opened Jun 10, 2025 by kylesayrs • Draft
16
[Bugfix] Safeguard against submodule parameter deletion in decompress_model
#347 opened Jun 10, 2025 by kylesayrs Loading…
2
[Bugfix] Fix find_name_or_class_matches return type
#346 opened Jun 9, 2025 by kylesayrs Loading…
relax setuptools_scm version requirement
#343 opened Jun 6, 2025 by envolution Loading…
[WIP] [Transform] q_attn and k_cache locations
#334 opened May 31, 2025 by kylesayrs • Draft
1
[WIP] [Transform] Compress, decompress
#333 opened May 31, 2025 by kylesayrs • Draft
[Misc] Remove unused config name definitions
#332 opened May 31, 2025 by kylesayrs Loading…
2
Make MSE the default
#324 opened May 22, 2025 by dsikka Loading…
4
[Transforms][WIP] Update wrapped_forward to use hooks; apply transforms to activations
#286 opened Mar 31, 2025 by dsikka • Draft
1
[Transforms] Enable shared memory and introduce permutations
#284 opened Mar 26, 2025 by dsikka Loading…
11
[Transforms] Initial Implementation
#277 opened Mar 11, 2025 by dsikka Loading…
[Transforms] Apply, serialize, deserialize
#276 opened Mar 11, 2025 by dsikka Loading…
6
[Quantization] Channel wise output activation quantization for QKV Attention layers
#270 opened Mar 7, 2025 by horheynm • Draft
ProTip! Add no:assignee to see everything that’s not assigned.