netdur

10 ms·

91.
by netdur 2y ago
I have made similar android app for semantic image search, works offline too, still gathering feedback and polishing UI, but it works, if you are brave enough here is it https://drive.google.com/file/d/1tE0cY6umj
92.
by netdur 2y ago
The problem with Flutter is that it's a UI framework that relies on a single main thread. If you put something heavy there, you'll get lag. In contrast, other languages like SwiftUI let you easily spawn threads. Flutter dismisses
93.
by netdur 2y ago
Yes, Egypt has faced significant financial challenges. Recently, there have been substantial investments from the UAE, purchasing a city for $35 billion, with potential similar investments from Saudi Arabia. Despite these inflows, concerns
94.
by netdur 2y ago
got it, I will not build micro services, I will build just regular one.
95.
by netdur 3y ago
I understand the methods to address the fine-tuning and RAG issues but lack the time and possibly the technical skills to implement the solution. Fine-tuning can potentially dumb down a perfect model, and RAG has context limitations and may
96.
by netdur 3y ago
Use Gemini 1.5 Pro, which has 1.5 million tokens. Prompt it with a logical question and observe it struggling to answer. Then, upload a book on logical thinking in PDF format and ask the same question again. Notice how it can now answer the
97.
by netdur 3y ago
Could you share more about your strategy or the approach you took for chunking the articles? I'm curious about the criteria or methods you used to decide the boundaries of each chunk and how you ensured the chunks remained meaningful f
98.
by netdur 3y ago
Yes, The team behind Ximian, before being acquired by SUSE, was involved in early efforts to improve the usability of desktop Linux for end users. They conducted usability studies and published videos of these sessions to highlight where us
99.
by netdur 3y ago
I believe Google stands apart; although their current offerings sucks, they possess all the necessary components, including software, data, and hardware.
100.
by netdur 3y ago
Yes Quantization compresses float32 values to int8 by mapping the large range of floats to a smaller integer range using a scale factor. This scale factor is key for converting back to floats (dequantization), aiming to preserve as much inf
101.
by netdur 3y ago
> - Need moar GPUs.. Why is there not a greater focus on quantization to optimize model performance, given the evident need for more GPU resources?
102.
by netdur 3y ago
original Mistral or GGUF one?
103.
by netdur 3y ago
oh thanks! I am getting down votted for a typo? reminds me grammar nazies during slashdot days... okay
104.
by netdur 3y ago
yes RAM is shared between CPU and GPU, I am able to run Mixtral-8x7B in this machine at decent speed, this is pretty good for testing LLM, yes
105.
by netdur 3y ago
I have MBP M1Max 32GB - it is $20000 now very good for LLM.
106.
by netdur 3y ago
I had toshiba satellite with some whooping 96MB RAM... as main computer even... happily ran Windows 98... them I got the book "Linux from scratch" the rest is history... now I am happy mac user.
107.
by netdur 3y ago
I hepled jQuery UI which landed me first real job with contract and everything, first side project was react like for canvas where only diff is drawen not all canvas, I have full family eating from this.
108.
by netdur 3y ago
maybe he is CEO or someone rich.
109.
by netdur 3y ago
12,800 CUDA - $4,000.00
110.
by netdur 3y ago
Sorry Microsoft, you had a small window to win on search with ChatGPT, but instead, you opted to immediately use this opportunity to force users to install software they don't care about. You are small because of you, not because of Go
111.
by netdur 3y ago
I am not sure how can a binary have open source license at first place
112.
by netdur 3y ago
later he said HF's tensor has almost zero copy, so 720GB is not needed using this format
113.
by netdur 3y ago
Generally worst than GPT4 but have some killer features, today I asked it for Mortal Kombat 1 release time in my time zone, I can also upload photo and have conversation about it But if you really wonder what they are building, get access t
114.
by netdur 3y ago
no not llama-cpp-python, it uses llama.cpp's built in server.
115.
by netdur 3y ago
I have helped making behind sense cases, one was to classify emails and redirect them to intended sides, second was quality monitoring of call center.
116.
by netdur 3y ago
GCP complies with various industry standards, regulations, and certifications that attest to its security and privacy controls. These certifications can give you added assurance that your data is being handled according to recognized standa
117.
by netdur 3y ago
I am trying to do the same but using C++ instead of python... wonder if someone else also trying it!
118.
by netdur 3y ago
no, the open source apply on code not binaries, it means right to demand source code if you "purchase" binaries, means companies can still restrict or sell software binaries... in this case the model can be considered a binary.
119.
by netdur 3y ago
what do you mean? they said this exactly from first day, the license allowed commercial use unless you have more than 700m users
120.
by netdur 3y ago
isn't this good? banks can raise security while we can still use any browser to check hackers news! or this 3 cows story?

More ›