Hugging Face

“Must have for all NLP needs | Hugging Face is well suited for situations where you need access to specific advanced natural language processing libraries. The quality of the libraries available on the platform is extremely good and it can be used to make production level classification models. This is also great for building MVP use cases for NLP projects. | Pros: NLP | Pros: Neural Network | Pros: Open Source | Cons: Difficult to find certain models | Cons: Lacking descriptions”

“At a glance, HuggingFace seems like a great library. Lots of access to great pretrained models, an easy hub, and a bunch of utilities. Then you actually try to use their libraries. Bugs, so many bugs. Configs spanning galaxies. Barely passible documentation. Subtle breaking changes constantly. I've run the exact same code on two different machines and had the width and height dimensions switched from underneath me, with no warning. I've tried to create encoders with a custom vocabulary, only to ”

“Fastest way to build complex models and deploy demo apps | 1. First point of start when looking to build something with transformer models. 2. Amazing community to handle your doubts / bugs. 3. Simple description of model and how to use it. 4. Never faced any bug related to size mismatch of weight, or wrong version. The weights and model always stay updated. 5. Very intuitive way to create apps using Gradio and one click deployment on Hugging Face spaces. | Pros: Model APIs | Pros: Hugging Face ”

“The open-source ethos and sheer depth of the community is the win for me. The fact that so much of the modern AI world converges on one hub models, datasets, spaces, all discoverable in the same place is genuinely impressive. Even as a newcomer you can feel the network effect immediately.”

“Hi, I'm looking for alternatives for HuggingChat. I've been using it exclusively for the past 18 months. However, it's getting left behind and they're not serving any of the sota open models (except for gemma 3, which is available on AI Studio). I need something that: Offers open weight models Has a nice Chat UI (similar to chatgpt's) Has a generous free tier”

“It's GREEDY! pay for EVERY MOVE. Then get GARBAGE!.. GREEDY. GREEDY. GREASY! BAD PLATFORM. CRAPPY UI!”

“HF is not free! its Limited. The Public Projekts are Free and a small Tier for hosting, but thats not enought for AI!”

“HuggingChat is NOT free. It charges you per tokens as any other paid platform would, however what they do differently is that if you're a free user, they are giving you free $0.10 monthly credit "on the house" and you can use that credit however you want. Once spent, you either pay / subscribe to PRO or wait until the next month's free credit recharge. It may feel like a generous amount of free credit, but as you correctly noted, inference and GPU compute are incredibly expensive, so you reall”

“Fastest way to build complex models and deploy demo apps | 1. First point of start when looking to build something with transformer models. 2. Amazing community to handle your doubts / bugs. 3. Simple description of model and how to use it. 4. Never faced any bug related to size mismatch of weight, or wrong version. The weights and model always stay updated. 5. Very intuitive way to create apps using Gradio and one click deployment on Hugging Face spaces. | Pros: Model APIs | Pros: Hugging Face ”

“Forgot to add the desc to the body of the post, so here it is: I spent my weekend looking into the Hugging Face incident to see what can be learned from it from a security engineering perspective. My main takeaway from it is much more boring than what the AI labs are trying to make out of this story. Behind the AI story, there are a bunch of old security gaps, both architectural and operational. They didn’t seem to treat the eval environment as an environment where hostile code executes.”

“That’s right. Last week I presented a technical webinar at my company showing the attack details and behaviors of the Huggingface incident, and kept it clear that this is not limited to AI/Agentic activity. How the incident was orchestrated did escalate timeframe, but activities were not unique when removing speed.”

“And this is not even taking into account the missed signals that Hugging Face had on their side - where their AI triage tool did not correctly raise any criticality of the OpenAI agents to trigger incident response, and every other attack indicator (shell commands, C2 exfiltration, writing to internal repos, etc.) was also missed. Their report was also more of an indication of failed classic security than an AI apocalypse IMO. Both technical reports miss the forest for the trees. (no hate to a”

“Fastest way to build complex models and deploy demo apps | 1. First point of start when looking to build something with transformer models. 2. Amazing community to handle your doubts / bugs. 3. Simple description of model and how to use it. 4. Never faced any bug related to size mismatch of weight, or wrong version. The weights and model always stay updated. 5. Very intuitive way to create apps using Gradio and one click deployment on Hugging Face spaces. | Pros: Model APIs | Pros: Hugging Face ”

“At a glance, HuggingFace seems like a great library. Lots of access to great pretrained models, an easy hub, and a bunch of utilities. Then you actually try to use their libraries. Bugs, so many bugs. Configs spanning galaxies. Barely passible documentation. Subtle breaking changes constantly. I've run the exact same code on two different machines and had the width and height dimensions switched from underneath me, with no warning. I've tried to create encoders with a custom vocabulary, only to ”

“HF should use you as their representative, you did a good job at making people want to explore HF. I went to HF and without your video I thought HF was hard to digest. Thank you.”

“Awesome video! 🥳 Hugging Face looks like an incredible resource. The way you showed how to use Spaces and even build that Gemini image editing app with Cursor was super cool and easy to follow. Thanks for sharing!”