forum

1495 threads · Page 20 of 30

import error RuntimeError: Failed to import trl.trainer.alignprop_trainer because of the following e 9 messages
128GB RAM and how much VRAM do I need to comfortable run 2.51-bit version of R1? 21 messages
Llama 3.2 Vision Finetuning Unsloth - Radiography notebook not working 9 messages
Database batch size error (new GRPO notebook) 5 messages
Open
Can not import FastLanguageModel from unsloth in Kaggle notebook 5 messages
Solved
CMake Error on Mistral_Small_22B Notebook 6 messages
Lora Qwen2.5 with GRPO error 10 messages
nvidia only required for unsloth? 2 messages
Solved
How to fine tune on low amounts of data? 3 messages
Open
How to convert the following to GUFF format? 7 messages
Wondering how much GPU memory it takes to finetune, inference the 32B, 70B and 671B deepseek models 7 messages
when merging 2.51-bit model files, received this error, how to fix it? 7 messages
Helping out 50 messages
Censorship 6 messages
Json data ai integration 2 messages
unsloth_windows.ps1 2 messages
High training loss 26 messages
Tried using Deepseek R1 (1.58bit) ... model is not loading correctly/takes forever to load 12 messages
want to train a local llm using custom dataset 3 messages
Open
Looking for guide on how to create bnb-4bit 2 messages
Solved
trainer.train() issues with unsloth: /usr/bin/ld: cannot find -lcuda: No such file or directorycoll 23 messages
TypeError: must be called with a dataclass type or instance when importing unsloth.FastLanguageModel 15 messages
Open
Running out of VRAM when fine tuning for VLM 6 messages
Issue with saving model step 4 messages
Perplexity of dynamic quant R1 models 3 messages
deepseek r1 llama cpp flash_attn requires n_embd_head_k == n_embd_head_v - forcing off 2 messages
save_pretrained_gguf llama.cpp error: "quantize does not exist." 21 messages
Solved
Running Deepseek-r1:8b with langchain_ollama 8 messages
Poor results after fine-tuning 20 messages
Quantize R-1 3B 7 messages
Running the deepseek R1 via VLLM 8 messages
Trying to run Deepseek-R1 Dynamic on Ollama 5 messages
Solved
SFT Fine-tuning on custom data 2 messages
How to save model 3 messages
Training on a domain 5 messages
Unsloth webui 13 messages
Janus-Pro 6 messages
AttributeError: 'NoneType' object has no attribute 'attn_bias' 14 messages
Open Solved
unsloth training LD error 25 messages
Continuous finetuning 4 messages
Finetuning on a sustainable dataset using Llama 3.2 3B 5 messages
NotImplementedError: Unsloth: unsloth/phi-4 33 messages
Looking for Data Scientist 5 messages
Using a local dataset with WSL 3 messages
Solved
image reshape to finetune llama vision? 10 messages
Cant start unsloth training 8 messages
Open
Lora QKV 8 messages
CUDA Out of memory.... When more VRAM is available? 10 messages
H100 training 4x slower than A100 12 messages
All tokens getting masked, Finetuning Llama-3.2-1B-Instruct 11 messages
Open