11 41

Oliver Hawthorne

Astris

AI & ML interests

None yet

Recent Activity

reacted to lewtun's post with 🔥 about 15 hours ago

I was initially pretty sceptical about Meta's Coconut paper [1] because the largest perf gains were reported on toy linguistic problems. However, these results on machine translation are pretty impressive! https://x.com/casper_hansen_/status/1875872309996855343 Together with the recent PRIME method [2] for scaling RL, reasoning for open models is looking pretty exciting for 2025! [1] https://huggingface.co/papers/2412.06769 [2] https://huggingface.co/blog/ganqu/prime

liked a model 5 days ago

deepseek-ai/DeepSeek-V3-Base

liked a model 9 days ago

Qwen/QVQ-72B-Preview

View all activity

Organizations

Astris's activity

New activity in Astris/Furry-AO3-LoRA 9 days ago

What do?

#2 opened 10 days ago by

AnonFurry

New activity in Astris/Ensemble_of_models_VQA_v0.5 6 months ago

[bot] Conversion to Parquet

#1 opened 6 months ago by

parquet-converter

Librarian Bot: Add language metadata for dataset

#2 opened 6 months ago by

librarian-bot

New activity in Astris/Nectar-Ranked-DPO 6 months ago

Librarian Bot: Add dpo tag

#3 opened 6 months ago by

librarian-bot

New activity in Astris/LA-Times 8 months ago

Question

#2 opened 9 months ago by

mrfakename

New activity in Astris/LA-Times 9 months ago

[bot] Conversion to Parquet

#1 opened 9 months ago by

parquet-converter

New activity in Astris/Nectar-Ranked-DPO 11 months ago

[bot] Conversion to Parquet

#1 opened 11 months ago by

parquet-converter

Librarian Bot: Add language metadata for dataset

#2 opened 11 months ago by

librarian-bot

New activity in Astris/Furry-AO3-LoRA 12 months ago

Base Mistral?

#1 opened 12 months ago by

KatyTheCutie

New activity in HuggingFaceH4/zephyr-7b-beta about 1 year ago

It seems the model doesnt work well on lower temperatures

#2 opened about 1 year ago by

fernandofernandes

New activity in Gryphe/MythoMax-L2-13b over 1 year ago

'use_cache: false' reduces tokens/sec significantly

#10 opened over 1 year ago by

Astris

'use_cache: false' reduces tokens/sec significantly

#10 opened over 1 year ago by

Astris