|
|
--- |
|
|
license: apache-2.0 |
|
|
language: |
|
|
- en |
|
|
tags: |
|
|
- checkpoint |
|
|
- training-artifact |
|
|
--- |
|
|
# DeepBrainz-R1-0.6B-16K-ckpt |
|
|
|
|
|
⚠️ **Checkpoint Repository — Not an Inference-Ready Model** |
|
|
|
|
|
This repository contains **raw training checkpoints** produced during the |
|
|
DeepBrainz-R1 model development process. |
|
|
|
|
|
These files are intended for: |
|
|
- Training lineage and reproducibility |
|
|
- Research inspection |
|
|
- Internal experimentation |
|
|
- Conversion into finalized inference artifacts |
|
|
|
|
|
--- |
|
|
|
|
|
## Important Notes |
|
|
|
|
|
- ❌ This repository is **NOT** intended for direct use with |
|
|
`AutoModelForCausalLM.from_pretrained` |
|
|
- ❌ Weights may be **sharded** and require merging |
|
|
- ❌ Inference configuration may be incomplete |
|
|
- ❌ No guarantees are provided regarding inference behavior |
|
|
|
|
|
--- |
|
|
|
|
|
## For Inference / Usage |
|
|
|
|
|
Please use a **finalized DeepBrainz-R1 model**. |
|
|
Checkpoint repositories are not intended for direct consumption. |
|
|
|
|
|
--- |
|
|
|
|
|
## License |
|
|
|
|
|
Apache 2.0 |
|
|
|
|
|
--- |
|
|
|
|
|
## About DeepBrainz |
|
|
|
|
|
DeepBrainz builds reasoning-first AI systems focused on efficiency, structure, |
|
|
and real-world problem solving. |
