license: llama3.1
base_model:
- meta-llama/Llama-3.1-8B-Instruct
SwiftKV
The Snowflake AI Research team is releasing a series of SwiftKV optimized Llama-3.1 models. SwiftKV is a series of inference optimizations that goes beyond traditional key-value (KV) cache compression. This method reduces computational overhead during prompt processing by combining model rewiring and knowledge-preserving self-distillation, allowing prefill tokens to skip up to half the model's layers. SwiftKV achieves up to 2x improvements in throughput, latency, and cost efficiency with minimal accuracy loss, making LLM deployments more performant and economically viable.
For more details about the technique
- Blog:
- arXiv paper: https://arxiv.org/abs/2410.03960
Eval metrics
| Model | Arc-Challenge | MMLU | MMLU-CoT | GSM-8k-CoT |
|---|---|---|---|---|
| meta-llama/Llama-3.1-8B-Instruct | ||||
| Snowflake/Llama-3.1-SwiftKV-8B-Instruct | ||||
| meta-llama/Llama-3.1-405B-Instruct | ||||
| Snowflake/Llama-3.1-SwiftKV-405B-Instruct-FP8 |
How to use the models
Instructions on how to use vLLM for both evaluation and performance benchmarks: https://github.com/Snowflake-Labs/vllm/tree/swiftkv/examples/swiftkv