forum
phi3
Issue in loading the dataset
PEFT model returns no logits only labels
Solved
docker
Open
darkness8i8
Solved
Merging unsloth adaptor in base model
Train on csv
phi-4
Open
Fine-tuned Model Quality
TypeError: expected string or bytes-like object
help regarding dataset.
Open
Error quantizing.
"!curl -fsSL https://ollama.com/install.sh | sh" not creating Modelfile for gemma2 9b Notebook
serverless runpod
Finetuning llama3.2 vision
export llama-3.2-3b-instruct-4bit to GUFF
Somehow, getting HybridCache error while inference:
CPO/DPO training fails spectacularly
Solved
Very slow fine-tuning, is lora_dropout to blame?
Any successful DPO with mistral models??
Train again fine-tuned model
model.save_pretrained_gguf issue
Open
Saving model to GGUF/ llama.cpp issue
Open
The following columns in the training set don't have a corresponding argument in PeftModelForCausalL
Mismatched shape when loading a modified LM head size adapter
Solved
Can't get Ollama model file
there is the word "Assistant" at the beginning of the fine tuning model response when in llama-cpp
Solved
Fine-Tuning with Custom Dataset: Formatting and Best Practices
seasonal shopping deals
Fine-tuning with private data
Fail to run Continued Pretraining Notebook
[INST]
Markdown in output
You can't pass load_in_4bitor load_in_8bit as a kwarg when passing quantization_config argument
Support for SmolLM2
Qwen 2.5 7B Instruct Fine tune. Model keeps talking won't stop, and later spits gibberish long resp.
Open
how to implement roleplay dataset?
Open
How to Continue Training using Checkpoint Artifact Wandb (kinda bug or i'am dont understand ?)
Solved
Questions about: Utilities to export Unsloth finetunes to vLLM, SGLang & Ollama - LoRA adapters only
Open
Please help getting LLaVA to work.
CPT language layers only for Pixtral
CUDA driver error: operation not supported
Open
Will there or is there any support for audio+language models?
Open
Issue with Fine-Tuned LLaMA 3.2 1B Model in Ollama Using GGUF Format
Preventing OOM while finetuning
Import Error Conda Env
Struggling to Fine-Tune LLaMA 3.2 Models: Why Does Base Model Outperform Instruct in My Use Case
Phi-3.5 generates random
Open
Unsloth on Windows (native)
Solved
Why do the example notebooks use DataCollatorForSeq2Seq?