File size: 950 Bytes
1e342b6 45ca338 1e342b6 45ca338 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 |
---
library_name: transformers
base_model:
- Qwen/Qwen2.5-32B-Instruct
license: apache-2.0
---
# Rombos-LLM-V2.5-Qwen-32b 4.5 BPW exl2
4.5 BPW quant of https://huggingface.co/rombodawg/Rombos-LLM-V2.5-Qwen-32b
Scores 63.9 on Aider benchmarks!
---
# Rombos-LLM-V2.5-Qwen-32b
Rombos-LLM-V2.5-Qwen-32b is a continues finetuned version of Qwen2.5-32B. I noticed recently that the Qwen team did not learn from my methods of continuous finetuning, the great benefits, and no downsides of it. So I took it upon myself to merge the instruct model with the base model myself using the *Ties* merge method
This version of the model shows higher performance than the original instruct and base models.
Quants: (Coming soon)
GGUF: https://huggingface.co/bartowski/Replete-LLM-V2.5-Qwen-32b-GGUF
EXL2:
Benchmarks: (Coming soon)
|