THEBLOKE HAS QUANTS!
https://huggingface.co/TheBloke/Mythical-Destroyer-L2-13B-GPTQ
https://huggingface.co/TheBloke/Mythical-Destroyer-L2-13B-GGUF
A Merge done for @dampf
FULL FP16 Model
Base Model TheBloke/Llama-2-13B-fp16
MERGED WITH
-----Gryphe/MythoMax-L2-13b
-----totally-not-an-llm/PuddleJumper-13b
-----TheBloke/Llama-2-13B-Chat-fp16
-----rombodawg/LosslessMegaCoder-llama2-13b-mini
-----The-Face-Of-Goonery/Chronos-Beluga-v2-13bfp16
using ties-merge
Dampf's Rationale:
if you think about it, the merges kinda act as experts in my destroyer.
mythomax and chronos-beluga for creativity,
llama 2 13b chat and puddlejumper for instruct and losslessmegacoder for logic/code
if this works well...
it should be really, really good
---
mythical destroyer will be used for rp and instruct as well as coding tasks a like
and it should be good at everything
---
Script used to Merge here
Thank you for the easy to set up script, Chargoddard !
Command:
python ties_merge.py TheBloke/Llama-2-13B-fp16 ./Mythical-Destroyer-13B --merge Gryphe/MythoMax-L2-13b --merge totally-not-an-llm/PuddleJumper-13b --merge TheBloke/Llama-2-13B-Chat-fp16 --merge rombodawg/LosslessMegaCoder-llama2-13b-mini --merge The-Face-Of-Goonery/Chronos-Beluga-v2-13bfp16 --cuda
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
Metric | Value |
---|---|
Avg. | 50.13 |
ARC (25-shot) | 58.7 |
HellaSwag (10-shot) | 82.0 |
MMLU (5-shot) | 57.66 |
TruthfulQA (0-shot) | 56.35 |
Winogrande (5-shot) | 74.66 |
GSM8K (5-shot) | 8.95 |
DROP (3-shot) | 12.56 |
- Downloads last month
- 2,097