alkinun's picture

alkinun

AtAndDev

·

AI & ML interests

LLMs, Alignment, Merging, Unsloth, DPO, SFT, ORPO, SPIN..

Recent Activity

reacted to davanstrien's post with 🤝 about 4 hours ago

The https://huggingface.co/datasets/data-is-better-together/fineweb-c dataset is growing! This week a few more languages have got 1,000 annotations for the educational quality of data from https://huggingface.co/datasets/HuggingFaceFW/fineweb-2. Why should you care? The quality of pre-training data can have a big impact on the performance of downstream language models trained on that data (https://huggingface.co/spaces/HuggingFaceFW/blogpost-fineweb-v1). Being able to filter by educational quality is on way of improving the quality of the data you use for training an LLM. Very importantly this approach can also reduce the amount of data needed for pertaining. Why not use an LLM? LLMs can be used to annotate educational quality for a subset of data. This data can then be used to train a smaller encoder only model to label the full dataset. However, this may not work well for languages outside of english. This is where fineweb-c (community) comes in. The community is annotating the educational quality of fineweb2 data. Currently 114 languages have some annotations. These annotations will enable a number of things: - Evaluate whether an LLM can label the educational quality for texts in that language well - Directly be used for training quality classifiers - Help discover other rules and huerisitcs for refining fineweb2 further for different languages. This week the following languages where done: Swedish thanks to: @Lauler @AntonVic @ohallstrom @bjarlestam @menbom @Ekgren @apsod Ukrainian thanks to: @hannayukhymenko @robinhad @realPivo @RabotiahovDmytro @reciprocate Assamese thanks to: @moyoor97 @Arpanjyoti @nawaf-helmi123 @pahigogoi1 @aelhence @kishorekashyap Want to learn more: https://huggingface.co/blog/davanstrien/fineweb2-community Contribute yourself here: https://huggingface.co/spaces/data-is-better-together/fineweb-c

reacted to davanstrien's post with 🔥 about 4 hours ago

The https://huggingface.co/datasets/data-is-better-together/fineweb-c dataset is growing! This week a few more languages have got 1,000 annotations for the educational quality of data from https://huggingface.co/datasets/HuggingFaceFW/fineweb-2. Why should you care? The quality of pre-training data can have a big impact on the performance of downstream language models trained on that data (https://huggingface.co/spaces/HuggingFaceFW/blogpost-fineweb-v1). Being able to filter by educational quality is on way of improving the quality of the data you use for training an LLM. Very importantly this approach can also reduce the amount of data needed for pertaining. Why not use an LLM? LLMs can be used to annotate educational quality for a subset of data. This data can then be used to train a smaller encoder only model to label the full dataset. However, this may not work well for languages outside of english. This is where fineweb-c (community) comes in. The community is annotating the educational quality of fineweb2 data. Currently 114 languages have some annotations. These annotations will enable a number of things: - Evaluate whether an LLM can label the educational quality for texts in that language well - Directly be used for training quality classifiers - Help discover other rules and huerisitcs for refining fineweb2 further for different languages. This week the following languages where done: Swedish thanks to: @Lauler @AntonVic @ohallstrom @bjarlestam @menbom @Ekgren @apsod Ukrainian thanks to: @hannayukhymenko @robinhad @realPivo @RabotiahovDmytro @reciprocate Assamese thanks to: @moyoor97 @Arpanjyoti @nawaf-helmi123 @pahigogoi1 @aelhence @kishorekashyap Want to learn more: https://huggingface.co/blog/davanstrien/fineweb2-community Contribute yourself here: https://huggingface.co/spaces/data-is-better-together/fineweb-c

reacted to davanstrien's post with ❤️ about 4 hours ago

The https://huggingface.co/datasets/data-is-better-together/fineweb-c dataset is growing! This week a few more languages have got 1,000 annotations for the educational quality of data from https://huggingface.co/datasets/HuggingFaceFW/fineweb-2. Why should you care? The quality of pre-training data can have a big impact on the performance of downstream language models trained on that data (https://huggingface.co/spaces/HuggingFaceFW/blogpost-fineweb-v1). Being able to filter by educational quality is on way of improving the quality of the data you use for training an LLM. Very importantly this approach can also reduce the amount of data needed for pertaining. Why not use an LLM? LLMs can be used to annotate educational quality for a subset of data. This data can then be used to train a smaller encoder only model to label the full dataset. However, this may not work well for languages outside of english. This is where fineweb-c (community) comes in. The community is annotating the educational quality of fineweb2 data. Currently 114 languages have some annotations. These annotations will enable a number of things: - Evaluate whether an LLM can label the educational quality for texts in that language well - Directly be used for training quality classifiers - Help discover other rules and huerisitcs for refining fineweb2 further for different languages. This week the following languages where done: Swedish thanks to: @Lauler @AntonVic @ohallstrom @bjarlestam @menbom @Ekgren @apsod Ukrainian thanks to: @hannayukhymenko @robinhad @realPivo @RabotiahovDmytro @reciprocate Assamese thanks to: @moyoor97 @Arpanjyoti @nawaf-helmi123 @pahigogoi1 @aelhence @kishorekashyap Want to learn more: https://huggingface.co/blog/davanstrien/fineweb2-community Contribute yourself here: https://huggingface.co/spaces/data-is-better-together/fineweb-c

View all activity

Organizations

AtAndDev's activity

New activity in reciperesearch/dolphin-sft-v0.1-preference about 2 months ago

aaa

#2 opened 3 months ago by

New activity in mattshumer/Reflection-Llama-3.1-70B 4 months ago

Bruh why isnt this model down yet

#259 opened 4 months ago by

New activity in unsloth/gemma-2-27b-it-bnb-4bit 4 months ago

Aphrodite/VLLM/SGLang all refuse to load this model

#5 opened 4 months ago by

commented a paper 6 months ago

AgentInstruct: Toward Generative Teaching with Agentic Flows

Paper • 2407.03502 • Published Jul 3, 2024 • 51 •

New activity in AtAndDev/ShortKing-3b-v0.2 9 months ago

Dataset link on line 1 renders 404

#4 opened 9 months ago by

New activity in open-llm-leaderboard/open_llm_leaderboard over 1 year ago

Benchmarks full names

#303 opened over 1 year ago by

New activity in AtAndDev/ShortKing-3b-v0.2 over 1 year ago

Adding `safetensors` variant of this model

#1 opened over 1 year ago by

New activity in cognitivecomputations/dolphin over 1 year ago

Number of samples?

#7 opened over 1 year ago by

New activity in Chat-Error/wizard_alpaca_dolly_orca over 1 year ago

License / Creadits

#2 opened over 1 year ago by

New activity in AtAndDev/ShortKing-1.4b-v0.1 over 1 year ago

Adding `safetensors` variant of this model

#1 opened over 1 year ago by

New activity in vicgalle/alpaca-gpt4 over 1 year ago

License

#3 opened over 1 year ago by

New activity in s3nh/Photolens-llama-2-13b-langchain-chat-GGUF over 1 year ago

Convert to GGUF format for other languages

#1 opened over 1 year ago by

New activity in bigscience/bloom almost 2 years ago

Code generation

#147 opened about 2 years ago by

New activity in bigscience/test-bloomd-6b3 about 2 years ago

Create README.md

#1 opened about 2 years ago by

Create README.md

#1 opened about 2 years ago by

Create README.md

#1 opened about 2 years ago by