/aicg/ - A general dedicated to the discussion and development of AI chatbotsMiku Ryona Edition>NewsGemini 2.0 Flash released, "more model sizes" in January https://developers.googleblog.com/en/the-next-chapter-of-the-gemini-era-for-developers/Gemini-exp-1206 is now available via the API, use <option value="gemini-exp-1206">Gemini Exp 1206</option> or update ST staging https://x.com/OfficialLoganK/status/1865081419015352689Non-preview OpenAI o1 out for web, no API yet https://cdn.openai.com/o1-system-card-20241205.pdfNew GPT-4o snapshot gpt-4o-2024-11-20 released https://x.com/OpenAI/status/1859296125947347164additional info: https://rentry.org/aicg_extra_information>FrontendsSillyTavern: https://docs.sillytavern.appRisuAI: https://risuai.netAgnai: https://agnai.chat | https://rentry.org/agnai_guides>Botshttps://characterhub.org | https://chub.aihttps://realm.risuai.nethttps://char-archive.evulid.cchttps://partyintheanchorhold.neocities.orghttps://rentry.org/meta_bot_list>Modelsjailbreaks: https://rentry.org/jb-listinggpt: https://platform.openai.com/playground | https://platform.openai.com/docsclaude: https://console.anthropic.com | https://docs.anthropic.com | https://rentry.org/how2claudegemini: https://gemini.google.com | https://ai.google.dev/docs | https://rentry.org/gemini-qrlocal: >>>/g/lmg | https://rentry.org/meta_golocal_list | https://openrouter.ai>Botmakinghttps://rentry.org/meta_botmaking_listhttps://desune.moe/aicharedhttps://agnai.chat/editor>MetaOP templates: https://rentry.org/aicgOPaicg botmaking events: https://rentry.org/meta_event_listlore: https://rentry.org/aicg_chroniclescard v3 status: https://rentry.org/zdwfrgqqservices assessment: https://rentry.org/aicg_metalogs: https://sprites.neocities.org/logs/reader | https://chatlogs.neocities.orglatest proxies: https://rentry.org/desuproxyrebornPrevious: >>103505518
ANCHOR
>>103510356>>103510356>>103510356
>>103510371I hate that shiny yet discolored Miku model so much.
>>103511459That's because you have shit taste.
>>103510374fuck botniggers
Buongiorno Pebble, thank you for making finals week easier. Love.
hi
>>103513814>making finals week easier>waiting 5+minutes for a response for some basic ass "le sexy time with anime girl" card when you should be cramming
Thank you very much 'oishicord for deleting the OP pic, by the way.
>>103515925true
>>103510371Does there exist a transformer model or something on huggingface that is around a lightweight 3B param model and can do pure C code well? Or that could at least be trained to do C code well?
>>103517427Here are some recents: LG EXAONE 3.5B, Qwen 2.5 3B, Llama 3.2 3B, IBM Granite 3.0 3B. But that's too small to be really helpful even if you finetune the shit out of it making it purely a C coding retard.Anyway, >>>/g/lmg
>>103517491Thank you I will look into those.So anything sub-7b at this point is still retarded basically?
>>103517580Relatively. The smallest usable model might be one of the Phi3.5 MoEs. But yeah, llama 3.1 8B, Ministral 8B, Qwen 2.5 with the base instruct model or their coder finetune.But as far as I know actually usable as coding assistants local models start at even higher beaks. Codestral, Qwen2.5-Coder-32B, QwQ, Deepseek's smaller Coder MoE. Usable when you have a 3090/4090 or cheap if API and good performance.
>>103517427https://huggingface.co/Qwen/Qwen2.5-Coder-3B
