Ecosyste.ms: Timeline
Browse the timeline of events for every public repo on GitHub. Data updated hourly from GH Archive.
YangWang92 created a comment on an issue on microsoft/VPTQ
> [@YangWang92](https://github.com/YangWang92) [#112](https://github.com/microsoft/VPTQ/pull/112) :) Thanks!
YangWang92 pushed 1 commit to main microsoft/VPTQ
- Add CUDA_HOME instructions to README (#112) Co-authored-by: Yang Wang <[email protected]> 05820d6
YangWang92 pushed 1 commit to readme-update caronzh03/VPTQ
- Update README.md update readme for cuda 716fcab
YangWang92 created a comment on an issue on microsoft/VPTQ
> NVM, the above issue was fixed by: > > 1. Changing the Docker image to `12.1.0-devel-ubuntu22.04` > 2. Setting the CUDA home: `export CUDA_HOME=/usr/local/cuda-12.1` Hi @caronzh03, Thanks for yo...
YangWang92 created a comment on an issue on microsoft/VPTQ
I expect to try the BERT model in about 2-3 weeks, but this is not my main task at the moment. Do you have any recommended settings?
YangWang92 created a branch on YangWang92/VPTQ
demo - VPTQ, A Flexible and Extreme low-bit quantization algorithm
YangWang92 pushed 2 commits to main YangWang92/VPTQ
YangWang92 created a comment on an issue on microsoft/VPTQ
We currently do not have Docker images. If you are looking to develop using a Docker image, you should base it on NVIDIA Docker. Currently, you can start by creating a conda environment using the...
YangWang92 created a comment on an issue on microsoft/VPTQ
Hi @caronzh03, We are currently using the Hessian matrix collection from https://github.com/Cornell-RelaxML/quip-sharp?tab=readme-ov-file#quantization. Please refer to this link for more details....
YangWang92 closed a pull request on genmoai/models
Update README.md
add ffmpeg installation to avoid errorYangWang92 created a comment on a pull request on genmoai/models
> should be only a note to the installations instructions instead, like this > > https://ffmpeg.org/download.html > > should be good enough, not all systems use apt updated, thanks!
YangWang92 pushed 1 commit to patch-1 YangWang92/models
- Update README.md fix ffmpeg installation bf8c6f6
YangWang92 pushed 18 commits to vptq/dev VPTQ/transformers
- Fix: img size mismatch caused by incorrect unpadding in LLaVA-Next (#34522) Fix: unpadding img mismatch 1b86772
- Fix step shifting when accumulate gradient (#33673) * replace total_batched_samples with step while counting grad ac... dca93ca
- avoid calling `gc.collect` and `cuda.empty_cache` (#34514) * update * update * update * update * update ... ab98f0b
- Qwen2VL: skip base `input_ids`-`inputs_embeds` equivalence check (#34535) it has complex inputs_embeds computation 4ca004e
- fix(DPT,Depth-Anything) Address expected_slice errors inside inference tests (#34518) * fix(DPT,Depth-Anything) Addr... b5919e1
- feat: add benchmarks pg indexes (#34536) * feat: add benchmarks pg indexes * refactor: remove debug `df -h` 294c170
- make `test_eager_matches_sdpa_inference `less flaky (#34512) * try * try * try * try * try * try *... 114dd81
- Bug Fix for issue #34294 (#34295) Update SiglipVisionEmbeddings.forward to cast input to correct dtype before embedd... c443d8d
- Add image text to text pipeline (#34170) * Standardize image-text-to-text-models-output add post_process_image_te... 203e270
- [CLIPSeg] Make interpolate_pos_encoding default to True (#34419) * Remove interpolate_pos_encoding * Make fixup ... df8640c
- update doc (#34478) * update doc * Update docs/source/en/perf_train_cpu.md Co-authored-by: Steven Liu <5946235... 2801d7b
- [i18n-ar] Translated file : `docs/source/ar/multilingual.md` into Arabic (#33048) * Add docs/source/ar/multilingual.... b53e44e
- Blip: get/set input embeddings correctly (#34152) * set-get embeds * add tests * fix tests * remove * re... 6beb3f1
- BLIP: enable generation tests (#34174) * blip2 tests * instructblips * copies * fix slow tests * fix ... 4cc0813
- red_circle: :red_circle: fix `query_pre_attn_scalar` different of `num_heads` in default gemma2 config (#34540) * f... 86701f2
- Large modular logic refactoring (#34487) * rework converter * Update modular_model_converter.py * Update modul... e2ac16b
- [i18n-HI] Translated accelerate page to Hindi (#34443) * [i18n-HI] Translated accelerate page to Hindi * Update d... 33868a0
- Merge branch 'huggingface:main' into vptq/dev 1e72434
YangWang92 pushed 17 commits to main VPTQ/transformers
- Fix: img size mismatch caused by incorrect unpadding in LLaVA-Next (#34522) Fix: unpadding img mismatch 1b86772
- Fix step shifting when accumulate gradient (#33673) * replace total_batched_samples with step while counting grad ac... dca93ca
- avoid calling `gc.collect` and `cuda.empty_cache` (#34514) * update * update * update * update * update ... ab98f0b
- Qwen2VL: skip base `input_ids`-`inputs_embeds` equivalence check (#34535) it has complex inputs_embeds computation 4ca004e
- fix(DPT,Depth-Anything) Address expected_slice errors inside inference tests (#34518) * fix(DPT,Depth-Anything) Addr... b5919e1
- feat: add benchmarks pg indexes (#34536) * feat: add benchmarks pg indexes * refactor: remove debug `df -h` 294c170
- make `test_eager_matches_sdpa_inference `less flaky (#34512) * try * try * try * try * try * try *... 114dd81
- Bug Fix for issue #34294 (#34295) Update SiglipVisionEmbeddings.forward to cast input to correct dtype before embedd... c443d8d
- Add image text to text pipeline (#34170) * Standardize image-text-to-text-models-output add post_process_image_te... 203e270
- [CLIPSeg] Make interpolate_pos_encoding default to True (#34419) * Remove interpolate_pos_encoding * Make fixup ... df8640c
- update doc (#34478) * update doc * Update docs/source/en/perf_train_cpu.md Co-authored-by: Steven Liu <5946235... 2801d7b
- [i18n-ar] Translated file : `docs/source/ar/multilingual.md` into Arabic (#33048) * Add docs/source/ar/multilingual.... b53e44e
- Blip: get/set input embeddings correctly (#34152) * set-get embeds * add tests * fix tests * remove * re... 6beb3f1
- BLIP: enable generation tests (#34174) * blip2 tests * instructblips * copies * fix slow tests * fix ... 4cc0813
- red_circle: :red_circle: fix `query_pre_attn_scalar` different of `num_heads` in default gemma2 config (#34540) * f... 86701f2
- Large modular logic refactoring (#34487) * rework converter * Update modular_model_converter.py * Update modul... e2ac16b
- [i18n-HI] Translated accelerate page to Hindi (#34443) * [i18n-HI] Translated accelerate page to Hindi * Update d... 33868a0