IridiumMaster
IridiumMaster
Was trying this on two Google A100s in their cloud. Devices list below. Used the following parameters: neural-style -multidevice_strategy 3,7,12 -gpu 0,1 -style_image myPainting63.jpg -content_image Headcrop4.jpg -model_file vgg19-d01eb7cb.pth -image_size 3000...
> > > @IridiumMaster The `-multidevice_strategy` parameter tells the code where to slice / cut the model, and in your case 2 GPUs means you want to have the model...
Would love if the server could get look ahead decoding and contrastive search. A collection of common presets would be very helpful for fast model evaluation. The ability to edit...
> @tobi > > > Very excited about this! I think that the server should increasingly be thought of as the main deliverable of this repo. > > LocalAI serves...
This is broken for me. You can try it with: python3 unsloth-cli-v2.py \ --model_name "unsloth/Llama-3.3-70B-Instruct-bnb-4bit" \ --load_in_4bit \ --dataset "./qa_Wodehouse_unsloth_conversion.jsonl" \ --output_dir "./wodehouse_finetune_output" \ --per_device_train_batch_size 2 \ --gradient_accumulation_steps 4 \...
This is still broken for me: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████| 400/400 [1:18:58
I installed cmake and compiled Llama.cpp and was confronted with this further error after re-running: Unsloth: Saving tokenizer... Done. Done. Unsloth: Converting llama model. Can use fast conversion = False....
Running again (with CMake installed) after deleting the llama.cpp directory produces this gem: -- Adding CPU backend variant ggml-cpu: -march=native CMake Error at /usr/share/cmake-3.22/Modules/FindPackageHandleStandardArgs.cmake:230 (message): Could NOT find CURL (missing:...
Installed Curl and still got this: Unsloth: Converting llama model. Can use fast conversion = False. ==((====))== Unsloth: Conversion from QLoRA to GGUF information \\ /| [0] Installing llama.cpp might...
@AleefBilal , I tried it 3 different ways. I tried compiling it from the download that the first run of unsloth did, and then I tried deleting the directory, downloading...