Reubencf (Reuben fernandes)

reacted to their post with 🔥 12 days ago

Post

138

Open Video Craft v1.0.2 is live 🎬

I built an open-source, local-first screen recorder and timeline editor for macOS and Windows. Record your screen, camera and audio; trim clips, add subtitles and export without an account.

GitHub + downloads: https://github.com/Reubencfernandes/Open-Video-Craft/releases/tag/v1.0.2

I’d love feedback from the Hugging Face community—especially ideas for useful captioning and AI-assisted editing workflows.

posted an update 12 days ago

Post

138

Open Video Craft v1.0.2 is live 🎬

I built an open-source, local-first screen recorder and timeline editor for macOS and Windows. Record your screen, camera and audio; trim clips, add subtitles and export without an account.

GitHub + downloads: https://github.com/Reubencfernandes/Open-Video-Craft/releases/tag/v1.0.2

I’d love feedback from the Hugging Face community—especially ideas for useful captioning and AI-assisted editing workflows.

reacted to their post with 🤗 21 days ago

Post

954

GPT 5.6 SOL builds hugging city a multiplayer online game hosted at Reubencf/Hugging-city

2 replies

·

replied to their post 22 days ago

I would have improved it but the usage gets over fast

posted an update 23 days ago

Post

954

GPT 5.6 SOL builds hugging city a multiplayer online game hosted at Reubencf/Hugging-city

2 replies

·

reacted to their post with 🔥 27 days ago

Post

3807

Reubencf/HF_QR_Gen
has been improved create custom Qr-codes

posted an update 27 days ago

Post

3807

Reubencf/HF_QR_Gen
has been improved create custom Qr-codes

reacted to RDTvlokip's post with 👍 29 days ago

Post

2002

I finally changed the architecture of my 15M French LLM. It worked. Then I almost fooled myself about how much and catching that was the real win.

After proving last time that architecture is a threshold, not a lever, I got stubborn: could I change how the model learns? Four honest attempts, Lion, a sharper AdamW β2, multi-token prediction, LayerScale. Four failures. The bottleneck wasn't the learning rule either.

So I changed the shape of the computation instead: loop the same transformer blocks 4×, deeper reasoning, zero added parameters. It beat the baseline on perplexity, the first thing in the whole project to move that number. Then I added my own twist: let each token decide how deep to think, halting on its own entropy.

My first evaluation was spectacular. Coherence up 65%. Hallucinated names down 62%.

It was noise.

Eight prompts, one seed. I re-ran on 50 prompts × 200 tokens and watched the gains shrink to "modest" and on out-of-domain prompts, recurrence actually made things worse. No universal winner. And none of it is new: it's Adaptive Computation Time (2016), the Universal Transformer (2018), and LoopViT (2026), recombined and measured honestly.

The real lesson:

A number from 8 prompts is a rumor. The eval harness that kills your own best result is worth more than the result it kills. Cite your lineage. Stay preliminary until multiple seeds say otherwise.

The three models are live. The write-up is honest about every caveat 👇

🔗 https://huggingface.co/blog/RDTvlokip/teaching-a-15m-french-llm-to-think-deeper

6 replies

·

reacted to their post with 🧠🤯👀 29 days ago

Post

150

Introducing Reubencf/Document_Query
powered by CohereLabs/command-a-plus-05-2026-w4a4 by cohere
just open admin page -> drop in any document -> and ask questions

1 reply

·

reacted to pankajpandey-dev's post with ❤️🚀🔥 29 days ago

Post

4120

🇮🇳 Qwen3.5-9B Hindi Instruct — it stops thinking in English
Ask base Qwen3.5-9B a question in Hindi and it burns hundreds of tokens thinking in English inside its think block before a single Devanagari word appears — then code-switches in the answer. I fine-tuned it to close the think block instantly and reply in pure, native Hindi.
✅ Model (16-bit): pankajpandey-dev/qwen3.5-9b-hindi-instruct
✅ GGUF (Q4/Q5/Q8): pankajpandey-dev/qwen3.5-9b-hindi-instruct-GGUF
✅ Try it in the browser: pankajpandey-dev/qwen3.5-9b-hindi-demo
Recipe: Unsloth + LoRA (r=16, response-only loss) on 12.9k Hindi pairs — AI4Bharat anudesh + dolly-hi + wikiHow-hi + Aya Hindi (human-written). The Q4_K_M is 5.4 GB and runs on a plain laptop CPU.
New in this run vs my earlier models: mixed in long-form native sources (wikiHow) after my last eval showed the fine-tune traded detail for conciseness — this one keeps answers detailed and native.
Part of my weekly 🇮🇳 Hindi LLM Series. Feedback welcome 🙏
#Hindi #IndicNLP #Qwen #GGUF #LocalLLM #Unsloth

4 replies

·

replied to their post about 1 month ago

@CohereLabs

reacted to their post with 🔥 about 1 month ago

Post

150

Introducing Reubencf/Document_Query
powered by CohereLabs/command-a-plus-05-2026-w4a4 by cohere
just open admin page -> drop in any document -> and ask questions

1 reply

·

posted an update about 1 month ago

Post

150

Introducing Reubencf/Document_Query
powered by CohereLabs/command-a-plus-05-2026-w4a4 by cohere
just open admin page -> drop in any document -> and ask questions

1 reply

·

replied to their post about 1 month ago

k

reacted to Passpass119's post with 🧠 about 1 month ago

Post

162

I am excited to announce that I have nothing to announce

reacted to Bc-AI's post with 👀 about 1 month ago

Post

227

# 🔥 Nova-1 Beta: Test Our New LLMs!

**Smilyai Labs** is building **Nova-1** — open-source LLMs with novel architectures. Join our beta program!

## 🎯 Available Now:

**Nova-1-Standard (1.2B)** — Phase 2 of pretraining in progress
- PPL 13.5 (beats GPT-2 Large!)
- 48K tok/s on consumer GPUs
- Great for code, reasoning, edge deployment

**Nova-1-Large (3.5B)** — Training live RIGHT NOW
- Current: 30.9 PPL, improving fast, loss at 3.5 right now
- Will finish with ~1.7B tokens today
- Better reasoning & longer context

**Nova-1-XL (10B MoE)** — Coming soon (We dont know yet! haha)
- Final Specs not decided yet

## What Makes Nova Special?

✨ **Mixture of Depths (MoD)** — Routes tokens dynamically, 30% faster
✨ **Grouped Query Attention** — Efficient like LLaMA 2/3
✨ **Phased Training** — Fresh 1B tokens each phase (no overfitting!)
✨ **RoPE** — Context extendable to 8K+

## 🤝 Join Beta Testing:

👉 **[Smilyai-labs-beta-testers](

Smilyai-labs-beta-testers

Get early access, shape the roadmap, and help build transparent open-source AI!

1 reply

·

Reuben fernandes PRO

AI & ML interests

Recent Activity

Organizations

Reuben fernandes PRO

AI & ML interests

Recent Activity

Organizations

Reubencf's activity