LuckyMind
LuckyMind
Hi @xuebinqin, how would you make a model smaller than your full one? So the full u2net is ~170mb, u2netp ~ 4mb. How can I do ~ 80/40/20 mb model...
Debug version works fine. But the same release version crashed on **module = LiteModuleLoader.load(absPath);** in **MainActivity.java** with errors: ``` E/AndroidRuntime(15751): FATAL EXCEPTION: main E/AndroidRuntime(15751): Process: com.example.pyml, PID: 15751 E/AndroidRuntime(15751): java.lang.ExceptionInInitializerError...
First of all - thank you for the code and model. Even for 7B it works awesome! But I stumbled upon strange problem. It doesn't matter how I set generator...
Sorry for my probably newbie questions. As I understand with current finetuning we teach how model should answer to our questions based on already present knowledge from basic training. But...
@tloen It would be great to have another small but useful script from you: - converting any trained checkpoint from finetuning (by giving it path) to lora weight (adapter_model and...
I have motion graphics engine in rust (Skia based) and trying to create GUI for it in Flutter. The problem is while I renders desired image very fast on the...
Working on my gui style inside freya. Noticed, that some labels rendered differently (anti aliased?). Please see screenshot: dropdown options and inside input text field labels have the same parameters...
Could you please show me an example where LLama-3 model used (better GGUF quantized) and initial prompt is more then 4096 tokens long? Or better 16-64K long (for RAG). Currently...
I find it not very convinient to export workflow to API version and not be able to load it later in GUI again. What I want to do is to...
Hi! Is it a simple way to play external midi file instead of waiting for inputs from midi keyboard?