Not Every Token Is Worth Distilling: Selective Supervision for Direct-OPD
Paper • 2609.29142 • Published • 6
AI & ML interests
None defined yet.
Recent Activity
Papers
Not Every Token Is Worth Distilling: Selective Supervision for Direct-OPD
ReNIO: Reweighting Negative Trajectory Importance for LLM On-Policy Distillation
Luli3220
submitted a
paper to Daily Papers 17 days ago
Post
2,494
Anthropic gave me six months of Claude Max 20x through the Claude for Open Source program, granted based on my Hugging Face work. Thank you 
Anthropic for supporting open source.
So far I've been pointing it at Markdown Minimap, an Obsidian plugin that adds a scrollable IDE-style minimap to your notes. This week I've been clearing a backlog of user-reported issues on it, with Claude often handling them end to end.
https://github.com/Nymbo/Markdown-Minimap — issues and PRs welcome.
So far I've been pointing it at Markdown Minimap, an Obsidian plugin that adds a scrollable IDE-style minimap to your notes. This week I've been clearing a backlog of user-reported issues on it, with Claude often handling them end to end.
https://github.com/Nymbo/Markdown-Minimap — issues and PRs welcome.
Post
6,181
Introducing Inflect-v2, two exceptionally small, open-weight English TTS models at just 3.9M and 9.3M parameters. Both generate speech multiple times faster than real-time on CPU. Despite their size, Inflect-v2 delivers quality that is competitive with much larger lightweight TTS systems, including KittenTTS, Piper, and Supertonic-3.
CPU, CUDA, PyTorch, and ONNX are supported. Apache 2.0.
See it for yourselves:
owensong/Inflect-Micro-v2
owensong/Inflect-Nano-v2
Try the Demos:
Nymbo/Inflect-TTS (unlimited CPU usage)
owensong/Inflect-v2 (ultra-fast ZeroGPU usage)
CPU, CUDA, PyTorch, and ONNX are supported. Apache 2.0.
See it for yourselves:
owensong/Inflect-Micro-v2
owensong/Inflect-Nano-v2
Try the Demos:
Nymbo/Inflect-TTS (unlimited CPU usage)
owensong/Inflect-v2 (ultra-fast ZeroGPU usage)
- 6 replies
Alephia
submitted a
paper to Daily Papers 4 months ago
heroding77
submitted a
paper to Daily Papers 5 months ago
Ujjwal-Tyagi
posted an update 6 months ago
Post
657
6 Open-Source Libraries to FineTune LLMs
1. Unsloth
GitHub: https://github.com/unslothai/unsloth
→ Fastest way to fine-tune LLMs locally
→ Optimized for low VRAM (even laptops)
→ Plug-and-play with Hugging Face models
2. Axolotl
GitHub: https://github.com/OpenAccess-AI-Collective/axolotl
→ Flexible LLM fine-tuning configs
→ Supports LoRA, QLoRA, multi-GPU
→ Great for custom training pipelines
3. TRL (Transformer Reinforcement Learning)
GitHub: https://github.com/huggingface/trl
→ RLHF, DPO, PPO for LLM alignment
→ Built on Hugging Face ecosystem
→ Essential for post-training optimization
4. DeepSpeed
GitHub: https://github.com/microsoft/DeepSpeed
→ Train massive models efficiently
→ Memory + speed optimization
→ Industry standard for scaling
5. LLaMA-Factory
GitHub: https://github.com/hiyouga/LLaMA-Factory
→ All-in-one fine-tuning UI + CLI
→ Supports multiple models (LLaMA, Qwen, etc.)
→ Beginner-friendly + powerful
6. PEFT
GitHub: https://github.com/huggingface/peft
→ Fine-tune with minimal compute
→ LoRA, adapters, prefix tuning
→ Best for cost-efficient training
1. Unsloth
GitHub: https://github.com/unslothai/unsloth
→ Fastest way to fine-tune LLMs locally
→ Optimized for low VRAM (even laptops)
→ Plug-and-play with Hugging Face models
2. Axolotl
GitHub: https://github.com/OpenAccess-AI-Collective/axolotl
→ Flexible LLM fine-tuning configs
→ Supports LoRA, QLoRA, multi-GPU
→ Great for custom training pipelines
3. TRL (Transformer Reinforcement Learning)
GitHub: https://github.com/huggingface/trl
→ RLHF, DPO, PPO for LLM alignment
→ Built on Hugging Face ecosystem
→ Essential for post-training optimization
4. DeepSpeed
GitHub: https://github.com/microsoft/DeepSpeed
→ Train massive models efficiently
→ Memory + speed optimization
→ Industry standard for scaling
5. LLaMA-Factory
GitHub: https://github.com/hiyouga/LLaMA-Factory
→ All-in-one fine-tuning UI + CLI
→ Supports multiple models (LLaMA, Qwen, etc.)
→ Beginner-friendly + powerful
6. PEFT
GitHub: https://github.com/huggingface/peft
→ Fine-tune with minimal compute
→ LoRA, adapters, prefix tuning
→ Best for cost-efficient training
- 1 reply
Ujjwal-Tyagi
posted an update 6 months ago
Post
582
This is the best set of AI and ML books and a full guide to learning machine learning from the ground up. This is my study material that I used, so I thought it would be helpful to share it with others. Like, share, and add it to your collection at Ujjwal-Tyagi/ai-ml-foundations-book-collection.
Ujjwal-Tyagi
posted an update 6 months ago
Post
3,988
We are hiring at Shirova AI. We need AI researchers and engineers to work in our research lab. Shirova AI is a research lab in India, so we can help our researchers move to nearby workspaces or let them work from home without ever coming to the lab. We're building our founding team, so the pay will be good. You can learn, so don't hesitate to mail us at: careers@shirova.com
FuNz
submitted a
paper to Daily Papers 6 months ago
chenjiawei1018
submitted a
paper to Daily Papers 7 months ago
Ujjwal-Tyagi
posted an update 7 months ago
Post
2,848
I am sharing my study material for AI & ML, these books are really a "bible" and gives very strong foundation, I also have given guidance, introduction and my master notes in the dataset repo card! I hope you will find them helpful, if you have any queries, just start a discussion and I am always there to help you out!
Ujjwal-Tyagi/ai-ml-foundations-book-collection
Ujjwal-Tyagi/ai-ml-foundations-book-collection
- 4 replies
Post
8,004
We should really have a release date range slider on the /models page. Tired of "trending/most downloaded" being the best way to sort and still seeing models from 2023 on the first page just because they're embedded in enterprise pipelines and get downloaded repeatedly. "Recently Created/Recently Updated" don't solve the discovery problem considering the amount of noise to sift through.
Slight caveat: Trending actually does have some recency bias, but it's not strong/precise enough.
Slight caveat: Trending actually does have some recency bias, but it's not strong/precise enough.
- 3 replies
Ujjwal-Tyagi
posted an update 7 months ago
Post
438
We have now LTX 2.3 with more better visual quality and richer sound, check it out! Lightricks/LTX-2.3
Ujjwal-Tyagi
posted an update 8 months ago
Post
3,045
Public reports allege that Anthropic gobbled up trillions of tokens of copyrighted material and public data to build their castle. 🏰📄 Now that they're sitting on top, they're begging for special laws to protect their profits while pulling the ladder up behind them. 🪜🚫
But the hypocrisy meter just broke! 📉 They are accusing Chinese labs like DeepSeek, Minimax, and Kimi of "huge distillation attacks. The Reality is that You can't just loot the entire internet's library, lock the door, and then sue everyone else for reading through the window. Stop trying to gatekeep the tech you didn't own in the first place. Read the complete article on it: https://huggingface.co/blog/Ujjwal-Tyagi/the-dark-underbelly-of-anthropic
But the hypocrisy meter just broke! 📉 They are accusing Chinese labs like DeepSeek, Minimax, and Kimi of "huge distillation attacks. The Reality is that You can't just loot the entire internet's library, lock the door, and then sue everyone else for reading through the window. Stop trying to gatekeep the tech you didn't own in the first place. Read the complete article on it: https://huggingface.co/blog/Ujjwal-Tyagi/the-dark-underbelly-of-anthropic
- 3 replies
Ujjwal-Tyagi
posted an update 8 months ago
Post
242
Qwen 3.5 Model is here! Supporting 1m context length by default, It is giving much good performance and competitive to Claude Opus 4.6, Qwen/Qwen3.5-397B-A17B, here it's GGUF: unsloth/Qwen3.5-397B-A17B-GGUF, Follow me and turn on the notification for the latest news!
Ujjwal-Tyagi
posted an update 8 months ago
Post
3,057
GLM 5 is insane, it ranks #4 Globally!
- 4 replies
Ujjwal-Tyagi
posted an update 9 months ago
Post
1,384
Finally we got a benchmark and research paper on ai safety, I am very excited to see what comes next on protecting AGI AgentDoG: A Diagnostic Guardrail Framework for AI Agent Safety and Security (2601.18491)
AI45Research/ATBench
AI45Research/ATBench
Ccloud0525
published a
model 9 months ago
Ujjwal-Tyagi
posted an update 9 months ago
Post
1,827
There is a new open-source music generation model called HeartMuLa. It offers strong, competitive performance compared to Suno and supports English, Chinese, Japanese, Korean, and Spanish. It is optimized to run easily on RTX GPUs and other consumer-grade hardware. HeartMuLa/HeartMuLa-oss-3B
https://github.com/HeartMuLa/heartlib
https://github.com/HeartMuLa/heartlib
- 1 reply
Ujjwal-Tyagi
posted an update 9 months ago
Post
2,803
So, Koreans are also doing great progress behind Chinese,
Their two open source ai models that are actually good in coding. upstage/Solar-Open-100B skt/A.X-K1
Their two open source ai models that are actually good in coding. upstage/Solar-Open-100B skt/A.X-K1
- 1 reply