Skip to navigation

Qwen2.5-1.5B

View as MarkdownOpen in Claude

Qwen2.5-1.5B has a checked-in NeMo AutoModel recipe for text generation. The Hugging Face configuration declares the Qwen2ForCausalLM architecture.

Fine-Tune Qwen2.5-1.5B

Follow the installation instructions, then run the recipe from the repository root:

automodel examples/llm_finetune/qwen/qwen2_5_1p5b_tulu3_gradient_sketch.yaml --nproc-per-node 8

Choose a Workflow

GoalStart Here
Run the primary recipe for this modelUse the recipe configuration.

Configuration

SettingConfiguration
Hardware-
StrategyFSDP2; tp_size=1, cp_size=1
Nodes1
FeaturesGradient-sketch optimizer (rank 256), chat SFT on Tulu-3
Advancedfp32 parameters, sequence length 2048

Model Reference

Model Architecture

PropertyValue
TaskText generation
ArchitectureQwen2ForCausalLM

Available Models