Falcon3 1B Instruct
Falcon3 1B Instruct is a compact transformer text model for English, French, Spanish and Portuguese. Its 8K context is smaller than the other three Falcon3 instruction sizes.
How this Falcon3 size was made
The 1B checkpoint was formed by pruning and retraining larger Falcon models. The card describes 18 decoder blocks and grouped-query attention. Post-training used instruction, code, safety and function-call examples.Falcon3-1B-Instruct model card
Keep long inputs out of the first comparison
Practical guidance. Start with short tasks that fit comfortably inside the context limit. If a source must be split, include the splitting method in the evaluation. Comparing a clipped prompt with a complete prompt can produce a misleading quality difference.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Launch date | December 2024 |
| Access | Downloadable |
| Model identifiers | tiiuae/Falcon3-1B-InstructFalcon3-1B-Base |
| Context window | 8,192 tokens Instruction checkpoint limit, corroborated by its max_position_embeddings configuration. Falcon3-1B-Instruct model card |
| License | TII Falcon-LLM License 2.0Falcon3-1B-Instruct model card |
| Weight access | Weights remain published.Falcon3-1B-Instruct model card |
| Variants | Base and Instruct; this page describes InstructFalcon3-1B-Instruct model card |
Related models
A 7B transformer language model for English, French, Spanish and Portuguese with a 32K context.
Launch date December 2024Downloadable