Ecosyste.ms: Timeline
Browse the timeline of events for every public repo on GitHub. Data updated hourly from GH Archive.
zRzRzRzRzRzRzR closed an issue on THUDM/CogAgent
有谁在24G的显存上，把CogAgent跑起来么？
下面是我测试的过程： hf版本： python cli_demo_hf.py --fp16 --quant 4 （推理出错） python cli_demo_hf.py --fp16 （加载62%崩溃） python cli_demo_hf.py --fp16 --quant 8（不支持参数值8，看源码只能支持4） sat版本： # In SAT version，--quan...
zRzRzRzRzRzRzR pushed 2 commits to main THUDM/CogVideo
zRzRzRzRzRzRzR closed a pull request on THUDM/CogVideo
[fix]fix typo in train_cogvideox_image_to_video_lora.py
zRzRzRzRzRzRzR created a comment on an issue on THUDM/CogVideo
"Necessary, otherwise the default is 49, which is 8 * 6 + 1 frame, please adjust and run each parameter according to cli_demo. Thank you."
zRzRzRzRzRzRzR created a comment on an issue on THUDM/CogVideo
The model cannot run with BF16 precision in a VRAM of 24G. You should run it on a GPU with more than 28G of GPU Memory
zRzRzRzRzRzRzR pushed 6 commits to main THUDM/CogVideo
- Merge pull request #567 from THUDM/main New Finetune 87ccd38
- docs: change "read this in English" to "中文阅读" Update README.md to use Chinese text for language switch link 3ff9d30
- Fix for Disney video dataset 4d1b9fd
- Fix for deepspeed training 36f1333
- Fix for CogVideoX-{2B,5B} When loading CogVideX-{2B,5B}, `patch_size_t` is None, which results in the `prepare_rotar... 7b4c9db
- Merge pull request #599 from THUDM/CogVideoX_dev Cog video x dev 1605e95
zRzRzRzRzRzRzR pushed 43 commits to main zRzRzRzRzRzRzR/transformers
- Typo in warning switching to optimum-quanto (#35028) fix typos 7b5f76e
- Add type hints for forward functions in Gemma2 (#35034) * feat: add gemma2 type hints * fix: mask is optional f41d5d8
- Fix `test_eager_matches_sdpa_inference` for `XPU` backend (#34889) * Use torch.nn.attention.sdpa_kernel instead of d... 3183047
- Multiple typo fixes in Tutorials docs (#35035) * Fixed typo in multi gpu docs and OLMoE version * Fixed typos in ... 3129967
- add docstring example for compute_loss_func (#35020) f0dec87
- [i18n-ar] Translated file : `docs/source/ar/notebooks.md` into Arabic (#33049) * Add docs/source/ar/notebooks.md to ... 4955e4e
- [docs] add the missing import for Image and bug fix (#34776) * add the missing import for Image lib * add more de... 527dc04
- Translate bertlogy.md into Chinese (#34908) * bertology translation * Update docs/source/zh/_toctree.yml Co-au... f9c7e60
- Automatic compilation in generate: do not rely on inner function (#34923) * compiled forward in PreTrainedModel *... ee37bf0
- Add token cost + runtime monitoring to Agent and HfEngine children (#34548) * Add monitoring to Agent and HfEngine c... 901f504
- Fix `BertGeneration` (#35043) fix Co-authored-by: ydshieh <[email protected]> 7a7f276
- fix speecht5 failure issue in test_peft_gradient_checkpointing_enable… (#34454) * fix speecht5 failure issue in test... 125de41
- [docs] fix example code bug (#35054) fix code bug 3deaa81
- Translate community.md into Chinese (#35013) * community translation * Update docs/source/zh/community.md Co-a... 346597b
- [docs] use device-agnostic instead of `cuda` (#35047) * fix on xpu * [run_all] * add the missing import for Im... b8cdc26
- [docs] use device-agnostic API instead of hard-coded cuda (#35048) replace cuda 329f5db
- Fix `pad_token_tensor` is None in warning (#34005) Fix pad_token_tensor is None in warning c7a109e
- Add Pytorch Tensor Parallel support for Qwen2, Qwen2Moe, Starcoder2 (#35007) * add base tp plan for qwen2 and qwen2m... accb720
- [`GPTNeoX`] Flex Attention + Refactor (#34896) * gpt neox flex attention + refactor * some formatting * small fix ... 46df859
- Support for easier multimodal use of modular (#35056) * update modular and add examples * style * improve exam... 1da1e0d
- and 23 more ...