Ecosyste.ms: Timeline
Browse the timeline of events for every public repo on GitHub. Data updated hourly from GH Archive.
zRzRzRzRzRzRzR created a comment on an issue on THUDM/CogVideo
We are doing fine-tuning work on the cogvideox-factory, please use the diffusers version for fine-tuning, as the sat version will run out of memory (OOM) even if it succeeds.
zRzRzRzRzRzRzR created a comment on an issue on THUDM/CogVideo
we have not conducted experiments on multi-machine multi-card inference, all are single-machine. I'm sorry we couldn't solve the problem.
zRzRzRzRzRzRzR created a comment on an issue on THUDM/CogVideo
But did you use the I2V model? The T2V model doesn't have this ofs, are you sure you installed the latest diffusers source code?
zRzRzRzRzRzRzR created a comment on an issue on THUDM/CogVideo
No, it is the clone of the main branch, https://github.com/huggingface/diffusers and then pip install -e .
zRzRzRzRzRzRzR created a comment on an issue on THUDM/CogVideo
We have released the version of diffusers, now you can give it a try, please download the diffusers library from the source code and run the diffusers model we open-sourced on Hugging Face.
zRzRzRzRzRzRzR created a comment on an issue on THUDM/CogVideo
Please update the latest diffusers library; I believe this issue can be resolved.
zRzRzRzRzRzRzR created a comment on an issue on THUDM/CogVideo
Please use the latest diffusers main branch to install, I believe this issue can be resolved.
zRzRzRzRzRzRzR created a comment on an issue on THUDM/CogVideo
The first and second frames (ID 0 and 1) are the same, so the first frame should be discarded.
zRzRzRzRzRzRzR closed an issue on THUDM/CogVideo
Forward error for CogVidX-5B-I2V model
### System Info / 系統信息 Hi, when I run the code on CogVidX-5B-I2V model, I got the error as below in cogvideox_transformer_3d.py file: ``` output = hidden_states.reshape(batch_size, num_frames, h...
zRzRzRzRzRzRzR created a comment on an issue on THUDM/CogVideo
I believe there are a few places that need to be revised 1 Update huggingface's CogVideoX1.5-5B latest commit regarding the T2V model 2 Update the source code and install the diffusers library ...
zRzRzRzRzRzRzR created a comment on an issue on THUDM/CogVideo
你是不是sat跑的，diffusers版本由于提前encode了，所以占用是固定，sat没办法因为是边编码边训练的，速度很慢，推荐cogvideox-factory
zRzRzRzRzRzRzR pushed 40 commits to main zRzRzRzRzRzRzR/transformers
- Add OLMo November 2024 (#34551) * Add model skeletion with transformers-cli add-new-model-like * Convert config t... 3ee24e2
- VLMs: `patch_size` -> `num_image_tokens` in processing (#33424) * use num additional tokens * fix copies + docs ... 1646ffb
- Fix broken link (#34618) eb0ab3e
- fix a typo bug where 'id2label' was incorrectly written as 'i2label' when reading config (#34637) fix a bug where 'i... c772d4d
- Fix skip of test_training_gradient_checkpointing (#34723) 19d58d31f has introduced a context manager to manage subte... 1c471fc
- make sure to disable gradients for integer tensor (#32943) 36759f3
- [docs] make `empty_cache` device-agnostic (#34774) make device-agnostic 8568bf1
- [docs] add XPU besides CUDA, MPS etc. (#34777) add XPU 9568a9d
- [tests] add XPU part to testing (#34778) add XPU part to testing Signed-off-by: Lin, Fanli <[email protected]> e80a65b
- fix: Update pixel_values parameter in hf_model input (#34782) 1ef6c5f
- Fix callback key name (#34762) Fixes typo. 7693b62
- fix: Wrong task mentioned in docs (#34757) 7df93d6
- Simplify Tensor Parallel implementation with PyTorch TP (#34184) * Simplify Tensor Parallel implementation with PyTo... 20142ab
- Allow handling files as args for a tool created with Tool.from_space (#34687) * Allow handling files as args for a t... 759a378
- Fix Whisper CI (#34617) * Revert "Revert "Fix Whisper CI" (#34605)" This reverts commit 74d3824cc0725829e7d92e1d4... eed11f3
- protect tensor parallel usage (#34800) protect dadb286
- Trainer hyperparameter search kwargs docs update (#34459) * doc: Trainer.hyperparameter_search docstring discrepancy... d8a5d31
- feat: allow to use hf-hub models for timm backbone (#34729) Currently a backbone name like 'hf-hub:bioptimus/H-optim... 1a0cd69
- Support gradient checkpointing in Qwen2VL ViT (#34724) * Support gradient checkpointing in Qwen2VL ViT * Enable g... 0db91c3
- Fix: siglip image processor rgb_convert is not being applied correctly. (#34301) Fix: do_convert_rgb 3cd78be
- and 20 more ...
zRzRzRzRzRzRzR created a comment on an issue on THUDM/CogVideo
已经支持了，请源代码安装最新的diffusers库和我在HF的最新一个commit
zRzRzRzRzRzRzR pushed 11 commits to main zRzRzRzRzRzRzR/transformers
- 🌐 [i18n-KO] Translated marian.md to Korean (#34698) * initial translation * removed english * Fixed Trivial Ty... be8748a
- [docs] Broken link in generation_strategies (#34717) [docs] Broken link e7c36a9
- Fix example in EsmConfig docstring (#34653) 68f8186
- [docs] add xpu device check (#34684) * add XPU path * use accelerate API * Update docs/source/en/tasks/semant... a3d69a8
- Retain newlines in chat template when `continue_final_message=True` (#34253) * Retain newlines in chat template when... 8ba3e15
- Update llava.md (#34749) LLava -> Llava f5dbfab
- fix(wandb): pass fake dataset to avoid exception in trainer (see #34455) (#34720) 7b3d615
- add xpu path for awq (#34712) * add xpu path for awq * update readme 52ea4aa
- FSDP grad accum fix (#34645) * add gradient accumulation steps tests for fsdp * invert no_sync context to fix tra... b0c0ba7
- Remove FSDP wrapping from sub-models. (#34452) * Remove FSDP wrapping from sub-models. * solve conflict trainer.p... 8d50fda
- 🧼 remove v4.44 deprecations (#34245) * remove v4.44 deprecations * PR comments * deprecations scheduled for v4... 1349321
zRzRzRzRzRzRzR pushed 11 commits to main THUDM/CogVideo
- update diffusers code c8c7b62
- update table fb806ee
- friendly link 2475902
- update 155456b
- fix 68d93ce
- update with diffusers 2151a3b
- add width and height bb2cb13
- update friendly link e7a35ea
- Merge pull request #500 from THUDM/main Merge 5e3e3aa
- update 17996f1
- Merge pull request #507 from THUDM/CogVideoX_dev diffusers version 2fdc59c
zRzRzRzRzRzRzR created a comment on an issue on THUDM/CogVideo
https://github.com/huggingface/diffusers/pull/9877 https://github.com/zRzRzRzRzRzRzR/diffusers/tree/cogvideox1.1-5b Try installing diffusers from here, as the pull request has not been merged y...
zRzRzRzRzRzRzR created a comment on an issue on THUDM/CogVideo
https://github.com/huggingface/diffusers/pull/9877 https://github.com/zRzRzRzRzRzRzR/diffusers/tree/cogvideox1.1-5b Try installing diffusers from here, as the PR has not been merged yet.
zRzRzRzRzRzRzR created a comment on an issue on THUDM/CogVideo
We are aware of this issue, but I have been working on the diffusers version recently. You can check the latest PR; we are in the final stage, and the diffusers version only requires a minimum of 9...
zRzRzRzRzRzRzR created a comment on a pull request on THUDM/CogVideo
Please note that this content can only be used normally after our [PR](https://github.com/huggingface/diffusers/pull/9877) is merged. Before that, please install diffusers from my branch correspond...
zRzRzRzRzRzRzR created a comment on an issue on THUDM/CogVideo
The structure of CogVideoX1.5 is almost the same as that of CogVideoX1.0, therefore, we did not produce a technical report for CogVideoX1.5, you can refer to the CogVideoX1.0's documents.
zRzRzRzRzRzRzR created a comment on a pull request on huggingface/diffusers
အားလုပ်ဆောင်မှု အကြောင်းအရာ အား ၇၆၈ * ၁၃၆၀ အတွက် မမှန်ကန်ပါဘူး၊ သို့ရာတွင် @kijai နဲ့ @a-r-r-o-w တို့နှစ်ဦးစလုံး အောင်မြင်ခဲ့ကြသလား? ကျွန်ုပ်၏ အလုပ်လုပ်ခြင်းတွင် @kijai က ယခင်က ဖော်ပြခဲ့သော မှန်း...
zRzRzRzRzRzRzR created a comment on a pull request on huggingface/diffusers
I have replied to some of the issues mentioned by @a-r-r-o-w on Slack. I am currently reproducing the issue mentioned in the latest issue report.