forum
Is it possible to fine tune glm-4.5V for vision?
Training Qwen3 8B
QAT fine-tuning
error in code
Gemma3-12B-finetuning
Bootstrap Relative Policy Optimization
Open
Does Unsloth w/ ART work with GPT-OSS:20b?
What changes are to be made to switch to Dr. GRPO from an existing GRPO file?
Solved
Help Training version of models and merging
Open
Gemma3-27B-finetuning (image and text)
ValueError: Invalid input type. Must be a single image, a list of images, or a list of batches of im
Open
Qwen fails to extract Names along with their bounding boxes from images
failing to save model after training
Want to help work on support for diffusion models for college project.
Prefered image resolution for qwen3 vl 8b
Qwen 3 32B Base
GRPO OOM Error partway through training
Solved
auto find batch size
help with changing so i get accuracy while training
accuracy while training
Poor quality at inference using Ollama or llama.cpp
Llama3.1 fine tuned model finetuning
Export to Ollama on VM
Summarization task
official ollama models vs unsloth GGUF
Open
About finetune with video on Qwen3-VL with unsloth
Open
gemma 3n failed training
How do I mask data properly
Fine tuning on non-reasoning data but inference with reasoning on
Vision Fine Tune with mixed dataset
Open
data preparation for instruct sft
Fine-Tune Whisper_large_v3_turbo.SST
Open
train model.
How do I create a custom datasset that can be used with Unsloth?
Multi-GPU Support for GRPO Training with Vision-Language Models
Open
GPT-OSS 20B - train_on_responses_only
Open
Qwen3 - Train on Responses Only
Query About Flash Attention
How to downgrade Unsloth
Open math reasoning by unsloth
Are you looking for developer?
i cant run torch (& gpu) on jupyter notebook for unsloth
DDP with Unsloth
Qwen3-VL 8B Instruct — Text Training Workflow for Domain Adaptation using Unsloth
UserWarning: Model qwen3-32b-bnb-4bit not found locally or on HuggingFace
Can't figure out why my finetuned model never gets better than base...
Revision argument in push_to_hub_merged not being used
4-bit Quantization after save_pretrained_merge for Qwen3-VL-8B
Error when attempting to load Granite 4.0 - h 350M model
trying to finetune on a quite complex dataset including alot of images of text