test-llama-tiny
Chat model trained with the Chat Model Competition Bittensor subnet.
Architecture: LlamaForCausalLM (vLLM compatible)
Pipeline: Pretrain (100 steps) → SFT (50 steps) → RL/GRPO (20 steps)
| Property | Value |
|---|---|
| Parameters | 13.3M |
| Architecture | 2L / 4H / 2KV / 128D |
| Intermediate | 384 |
| Seq Length | 128 |
- Downloads last month
- 59
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support