---
base_model:
- Qwen/Qwen3-1.7B
datasets:
- open-r1/OpenR1-Math-220k
license: apache-2.0
pipeline_tag: text-generation
tags:
- math
- trimkv
- KV
- Cache
- Compression
---

# TrimKV: Token Retention for Memory-Bounded Key-Value Eviction

This repository contains the weights for **TRIM-KV-Qwen3-1.7B-Math**, as presented in the paper [Cache What Lasts: Token Retention for Memory-Bounded KV Cache in LLMs](https://huggingface.co/papers/2512.03324).

The core idea behind TRIM-KV is to learn the intrinsic importance of each key–value pair at creation time, which we call *token retention*, and then decay this importance exponentially over time to mimic standard inference running with eviction. 

The retention score is query-agnostic and captures the long-term utility of tokens. This is different from attention scores, which are query-dependent: they capture the short-term utility for predicting the next token and are recomputed at every step.

- **Paper:** [Cache What Lasts: Token Retention for Memory-Bounded KV Cache in LLMs](https://huggingface.co/papers/2605.09649)
- **Code:** [GitHub - ngocbh/trimkv](https://github.com/ngocbh/trimkv)
- **Arxiv:** [2512.03324](https://arxiv.org/abs/2512.03324)

## Quick Start

### Installation

```sh
pip install trimkv
```

### Usage

```python
import torch
from trimkv.models.qwen3 import TrimKVQwen3ForCausalLM
from trimkv.cache_utils import PagedTrimKVCache
from transformers import AutoTokenizer

model_path = "ngocbh/TrimKV-Qwen3-1.7B-Math"

model = TrimKVQwen3ForCausalLM.from_pretrained(
    model_path,
    torch_dtype=torch.bfloat16,
    load_trimkv_weights=True,
    use_cache=True,
    device_map="cuda",
)
model.config._attn_implementation = "flash_attention_2"

tokenizer = AutoTokenizer.from_pretrained(
    model.config.base_model, use_fast=True, padding_side="left"
)

# PagedTrimKVCache is the inference-time cache used by TRIM-KV. 
# It allocates a global pool of blocks and reassigns them to heads on the fly.
past_key_values = PagedTrimKVCache(
    num_layers=model.config.num_hidden_layers,
    num_heads=model.config.num_key_value_heads,
    max_seq_len=32768,
    memory_size=128,
    num_blocks_ratio=1.0,
    buffer_size=32,
    strategy="fixed_budget",
    device="cuda",
)

# Use model.generate as normal — pass past_key_values to enable TrimKV eviction.
```

## Citation

```bibtex
@article{bui2025cache,
  title={Cache what lasts: Token retention for memory-bounded kv cache in llms},
  author={Bui, Ngoc and Sharma, Shubham and Lamba, Simran and Mishra, Saumitra and Ying, Rex},
  journal={arXiv preprint arXiv:2512.03324},
  year={2025}
}
@article{bui2025make,
  title={Make Each Token Count: Towards Improving Long-Context Performance with KV Cache Eviction},
  author={Bui, Ngoc and Nguyen, Hieu Trung and Cohan, Arman and Ying, Rex},
  journal={arXiv preprint arXiv:2512.03324},
  year={2025}
}
```