Fatih C. Akyon's picture

Fatih C. Akyon

fcakyon

·

https://www.linkedin.com/in/fcakyon/

AI & ML interests

multi-modal learning, video understanding

Recent Activity

new activity 1 day ago

WalidBouss/LeGrad:🚩 Report: Not working

liked a Space 1 day ago

WalidBouss/LeGrad

reacted to joaogante's post with 🤗 1 day ago

New sampling strategy dropped in 🤗 transformers -- Min P sampling 🔥 Are you tired of having `top_k` arbitrarily discarding high-quality continuations? Or `top_p` forgetting to exclude low-probability tokens, derailing your generation? Try out the new `min_p` flag in `generate`, fresh from a PR merged today! 🥬 Min P consists of a dynamic token filter -- as opposed to Top K, which keeps the K most likely tokens, and Top P, which keeps the most likely tokens up to a fixed cumulative probability, both static filters. Min P takes a base probability (defined in the `min_p` flag) and multiplies it by the probability of the most likely token in the distribution for the next token. All tokens less likely than the resulting value are filtered. What happens with this strategy? 👉 High probability token present -> aggressive filter (we don't want to miss on that high-probability case and risk derailing generation) 👉 No high probability token present -> relaxed filter (there are many continuation possibilities that the model finds plausible) You should set `min_p` to a low value, between 0.05 and 0.1. It behaves particularly well for creative text generation when paired up with temperature > 1. Kudos to @kalomaze and @menhguin for creating this technique 🔥 Read their discussion in the original issue for benchmarks (https://github.com/huggingface/transformers/issues/27670) Copy-pasteable version of the example in the image below here: https://pastebin.com/VqXNtuxd Have fun experimenting! 😎

View all activity

Organizations

fcakyon's activity

New activity in WalidBouss/LeGrad 1 day ago

🚩 Report: Not working

#1 opened 1 day ago by

New activity in microsoft/Florence-2-large 1 day ago

add_confidence_score

#56 opened 6 months ago by

New activity in microsoft/Florence-2-base 1 day ago

add confidence scores for `od` and `description_with_bboxes` tasks

#25 opened 1 day ago by

confidence score

#24 opened 1 day ago by

New activity in Ultralytics/YOLOv5 1 day ago

Enable download stats

#2 opened 3 days ago by

New activity in Ultralytics/YOLOv8 1 day ago

Enable download stats

#1 opened 3 days ago by

New activity in Ultralytics/YOLO11 1 day ago

Update library name

#1 opened 3 days ago by

New activity in thwri/CogFlorence-2 9 days ago

🚩 Report: Not working

#1 opened 12 days ago by

New activity in BAAI/bge-m3 12 days ago

broken link in the model card

#99 opened 12 days ago by

New activity in microsoft/Florence-2-base 2 months ago

Inherit from GenerationMixin

#22 opened 2 months ago by

New activity in microsoft/Florence-2-large 2 months ago

Inherit from GenerationMixin

#80 opened 3 months ago by

New activity in hf-accelerate/model-memory-usage 4 months ago

microsoft/Florence-2-base raises error without any detail

#36 opened 4 months ago by

New activity in Helsinki-NLP/opus-mt-tc-big-en-tr 4 months ago

how many parameters are there in the model?

#9 opened 4 months ago by

New activity in merve/idefics3llama-vqav2 5 months ago

parameter number of the model

#2 opened 5 months ago by

New activity in gsarti/gradio_highlightedtextbox 5 months ago

🚩 Report: Not working

#3 opened 5 months ago by

New activity in dawood/gradio_videogallery 5 months ago

🚩 Report: Not working

#3 opened 5 months ago by

New activity in evaluate-metric/bertscore 5 months ago

🚩 Report: Not working

#10 opened 5 months ago by

New activity in probing-vits/attention-heat-maps 6 months ago

🚩 Report: Not working

#4 opened 6 months ago by

fix ModuleNotFoundError: No module named 'altair.vegalite.v4'

#3 opened 6 months ago by

New activity in probing-vits/attention-rollout 6 months ago

🚩 Report: Not working

#1 opened 6 months ago by