forum

1495 threads · Page 5 of 30

ValueError: bf16 mixed precision requires PyTorch >= 1.10 and a supported device.bf16 2 messages
windows 11 ,rtx 3090, qwen3 moe 3 messages
Unsloth's gradient checkpointing crashing during GRPO training on evaluation 12 messages
Gemma 3 regression? 5 messages
help 7 messages
Unable to convert model from HF to GGUF format 44 messages
name 'psutil' is not defined 6 messages
Solved
text classification 2 messages
Convert finetuned ministral-3b to gguf failed 28 messages
Extremely High VRAM Usage with Ministral 3 3B VLM QLoRA 7 messages
Open
Model works with lora, but if lora is merged, the model starts making gibberish output 3 messages
Open
Ministral model can't load - no Config file 2 messages
Solved
Packing Compatibility with Ministral 6 messages
Open
Building docker image in DGX Spark 10 messages
Open
Pytorch + xformers compatibility issues on DGX Spark docker image build 69 messages
Open
Can QAT aware trained phi-4 model using unsloth and torchao collab packages be converted to gguf? 4 messages
Open
Confused on model differences 2 messages
Nemotron-3-Nano-30B-A3B-FP8 3 messages
train_on_responses_only does not work when using "packing" 4 messages
Open
Does packing not work with multigpu? 30 messages
Open
Error when upgrading to the latest unsloth version 12 messages
Solved
does unsloth support transformers v5? 4 messages
Qwen3VL (or any VLM) OOM from large dataset 23 messages
qwen3 vl 235b crash 45 messages
Fine-tune on multi GPUs 9 messages
Unable to export Ministral model to GGUF 6 messages
Open
Documentation Improvement: Common Mistakes 2 messages
Open
Trouble pushing unsloth/Ministral-3-3B-Instruct-2512 to my HF account 11 messages
VyvoTTS: ImportError: libnvshmem_host.so.3 5 messages
Open
Wondering what to consider if training a binary classifier 2 messages
Can not finetune whisper model with error about AssertionError: expected size 128==128, stride 1500 2 messages
How to serve unsloth/gpt-oss-120b-unsloth-bnb-4bit on vllm? 2 messages
Open
Synth Reasoning Gen 9 messages
Open
Finetuning Ministral-3-3B-Instruct-2512 9 messages
Open
Loading ft version (Gelato) of Qwen3VL into unsloth 3 messages
Fast Inference vllm GRPO tuning on Qwen3VL 2 messages
Open
What is the purchase price of Unsloth Enterprise 3 messages
Open
Hosting 4bit SFT/RFT using Vllm on DGX Sparks 2 messages
PaddleOCR-VL sft error: `Unsloth: Failed to make input require gradients!` 35 messages
Solved
Is this the correct way load a trained model? 5 messages
Image Size for finetuning qwen3-VL-2B 4 messages
Open
eval loss and train loss always identical 9 messages
answer cropped after finetune 3 messages
Batch size increase causes training time to increase (inverse scaling) on H100 3 messages
Open
How to use Fsdp via unsloth 7 messages
Open
Problem with GGUF conversion 195 messages
flashinfer issues 14 messages
Solved
Use qwen3vl thinking model to get final output after thinking 7 messages
Can I fine-tune Gemma-3-12B on a single RTX 3060 12 GB with Unsloth? 2 messages
Subject: Request for Guidance on Fine-Tuning GPT-OSS-20B Using COT + Structured DNA Reasoning Prompt 16 messages