Skip to content

Pull requests: vllm-project/llm-compressor

New pull request New
39 Open 399 Closed
39 Open 399 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
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

wandb/tensorboard loggers set default init to False ready When a PR is ready for review
#1235 opened Mar 7, 2025 by brian-dellabetta Loading…
2
1
[Quantization] Channel-wise Output Activation Quantization for Attention QKV Modules + KV-cache channel quantization ready When a PR is ready for review
#1233 opened Mar 7, 2025 by horheynm • Draft
3
fixing reproducibility of lmeval tests ready When a PR is ready for review
#1220 opened Mar 4, 2025 by brian-dellabetta Loading…
14
[Train] Training Pipeline ready When a PR is ready for review
#1214 opened Feb 28, 2025 by horheynm Loading…
1
[StageRunner] Stage Runner entrypoint and pipeline
#1202 opened Feb 26, 2025 by horheynm • Draft
[Callbacks][Docs] Add docstrings to saving functions
#1201 opened Feb 26, 2025 by kylesayrs Loading…
1
Use KV cache constant names provided by compressed tensors
#1200 opened Feb 26, 2025 by kylesayrs Loading…
@kylesayrs
2
[Callbacks] Remove on_update
#1199 opened Feb 26, 2025 by kylesayrs Loading…
@kylesayrs
[Utils] Add skip_weights_download and skip_weights_initialize utils
#1188 opened Feb 24, 2025 by kylesayrs • Draft
[Utils] Replace preserve_attr with patch_attr
#1187 opened Feb 24, 2025 by kylesayrs Loading…
@kylesayrs
1
[Bugfix] Raise error for processor remote code
#1184 opened Feb 23, 2025 by kylesayrs Loading…
@kylesayrs
2
Bdellabe/awq modifier v3
#1177 opened Feb 19, 2025 by brian-dellabetta • Draft
14
Remove unused/duplicated/non-applicable utils from pytorch/utils/helpers
#1174 opened Feb 19, 2025 by kylesayrs Loading…
@kylesayrs
5
Deprecate oneshot_device arg in favor of device_map
#1173 opened Feb 19, 2025 by kylesayrs Loading…
@kylesayrs
3
[Callbacks] Remove EventLifecycle and on_start event
#1170 opened Feb 19, 2025 by kylesayrs Loading…
4
[Callbacks] Remove double initialization, replace with updating the state directly ready When a PR is ready for review
#1169 opened Feb 19, 2025 by kylesayrs Loading…
@kylesayrs
13
Skip lm_eval tests if no config file is present
#1149 opened Feb 13, 2025 by kylesayrs • Draft
4
Deprecate obcq path in favor of sparsegpt path
#1148 opened Feb 13, 2025 by kylesayrs Loading…
2
Offload Cache Support torch.dtype ready When a PR is ready for review
#1141 opened Feb 12, 2025 by kylesayrs Loading…
@kylesayrs
4
[Entrypoint] Update oneshot to post_train
#1136 opened Feb 11, 2025 by horheynm • Draft
4
[Docs] Sphinx documentation
#1133 opened Feb 9, 2025 by kylesayrs • Draft
2
Update examples to only load required number of samples from dataset
#1118 opened Jan 31, 2025 by kylesayrs • Draft
1
Support user-defined batch size for one shot
#1117 opened Jan 30, 2025 by kylesayrs Loading…
@kylesayrs
16
Implement lazy loading for traceable models ready When a PR is ready for review
#1105 opened Jan 28, 2025 by kylesayrs Loading…
@kylesayrs
10
Add compressed-tensors appropriate code from llm-compressor
#1098 opened Jan 24, 2025 by horheynm • Draft
2
ProTip! Adding no:label will show everything without a label.