New Fine-Tuning Job

1. Model Selection

Search and select a model or enter a custom HuggingFace path.

Model Family	Version	Specialized Task	VRAM
Qwen2.5	7B Instruct	Text Generation	~8 GB
Qwen2.5	14B Instruct	Text Generation	~14 GB
Llama	3.1 8B	Text Generation	~8 GB
Mistral	7B Instruct v0.3	Text Generation	~8 GB
Qwen2.5	0.5B Instruct	Text Generation	~4 GB

8 models · page 1 of 2

Pick an existing dataset or upload a new one.

No datasets found. Upload one now.

Epochs

Batch Size

Learning Rate

Train/Val Split

Warmup Steps

Gradient Accumulation Steps

Max Steps

Select a model to continue.