Highlights
- Pro
Block or Report
Block or report patrickvonplaten
Report abuse
Contact GitHub support about this user’s behavior. Learn more about reporting abuse.
Report abusePopular repositories
-
Wav2Vec2_PyCTCDecode Public
Small repo describing how to use Hugging Face's Wav2Vec2 with PyCTCDecode
-
-
-
-
2,124 contributions in the last year
Less
More
Contribution activity
August 2022
Created 60 commits in 9 repositories
Created 1 repository
- patrickvonplaten/stable-diffusion Jupyter Notebook
Created a pull request in huggingface/huggingface_hub that received 16 comments
[Repository] Add better error message
This PR simply adds a better error message for the following use case: from huggingface_hub import Repository repo = Repository(local_dir="data", c…
+21
−8
•
16
comments
Opened 27 other pull requests in 5 repositories
huggingface/diffusers
16
merged
2
closed
- [Tests] Add fast pipeline tests
- Improve README
- [Stable Diffusion] Hotfix
- Check dummy file
- [Don't merge] SDE_VE integration test
- [Refactor] Remove set_seed
- [Type hint] Karras VE pipeline
- [Docs] Add some guides
- [README] Add readme for SD
- [Tests] Make sure tests are on GPU
- Refactor Pipelines / Community pipelines and add better explanations.
- [Clean up] Clean unused code
- [Loading] allow modules to be loaded in fp16
- [Do not merge] Script to compare safety checkers
- [K-LMS Scheduler] fix import
- [Naming] correct config naming of DDIM pipeline
- [PNDM] Stable diffusion
- [Half precision] Make sure half-precision is correct
huggingface/transformers
4
merged
sanchit-gandhi/seq2seq-speech
1
open
1
closed
huggingface/blog
2
merged
CompVis/stable-diffusion
1
merged
Reviewed 55 pull requests in 7 repositories
huggingface/diffusers
25 pull requests
- [Stable Diffusion] Hotfix
- [Refactor] Remove set_seed
- update expected results of slow tests
- Changed variable name from "h" to "hidden_states"
- Refactor Pipelines / Community pipelines and add better explanations.
- [WIP] Textual inversion
- Easily understandable error if inference steps not set before using scheduler (#263)
- Refactor progress bar
- Support K-LMS in img2img
- Initialize CI for code quality and testing
-
Style the
scriptsdirectory - Reproducible images by supplying latents to pipeline
- add inpainting example script
- [Clean up] Clean unused code
- split tests_modeling_utils
- Update README for 0.2.3 release
- [Do not merge] Script to compare safety checkers
- Pipeline to device
- Add safety module
- [StableDiffusionPipeline] use default params in __call__
- fix test_from_pretrained_hub_pass_model
- Add K-LMS scheduler from k-diffusion
- [Half precision] Make sure half-precision is correct
- [PNDM] Stable diffusion
- allow custom height, width in StableDiffusionPipeline
- Some pull request reviews not shown.
huggingface/transformers
19 pull requests
- Create pipeline_tutorial.mdx german docs
- [WIP] Split docs on modality
- Fixing OPT fast tokenizer option.
-
reflect max_new_tokens in
Seq2SeqTrainer - PEGASUS-X
- Add DocumentQuestionAnswering pipeline
- Fix incomplete outputs of FlaxBert
- fix missing block when there is no failure
-
Generate: validate
model_kwargson TF (and catch typos in generate arguments) - [Wav2vec2 + LM Test] Improve wav2vec2 with lm tests and make torch version dependent for now
- Determine framework automatically before ONNX export
- Mask t5 relative position bias then head pruned
- BLOOM Flax
- Add ONNX support for Longformer
- Type cast before normalize
- [Hotfix] pin detectron2 5aeb252 to avoid test fix
- Add hallucination filter
- [LayoutLMv3] Add TensorFlow implementation
-
Generate: validate
model_kwargson FLAX (and catch typos in generate arguments)
patil-suraj/stable-diffusion-jax
1 pull request
sanchit-gandhi/seq2seq-speech
1 pull request
huggingface/blog
1 pull request
CompVis/stable-diffusion
1 pull request
huggingface/huggingface_hub
1 pull request
Created an issue in huggingface/diffusers that received 9 comments
Allow non-default schedulers to be easily swapped into DiffusionPipeline classes
Is your feature request related to a problem? Please describe. By default the stable diffusion pipeline uses the PNDM scheduler, but one could easi…
9
comments
Opened 14 other issues in 4 repositories
huggingface/diffusers
8
open
3
closed
- Docs
- [Community effort] Type Hints
- Add DEIS scheduler
- Add noise interpolation and text intertpolation scripts to examples
- [Test] Add tests for unet_blocks file
- [Tests] Test attention class
- [Test] Test the resnet class
- Attention - Delete dead code
- Rename variables from single letter to better naming
- Resnet - Delete dead code
-
Split
tests_modeling_utils





