README.md

---
base_model:
- sometimesanotion/Lamarck-14B-v0.6
- deepseek-ai/DeepSeek-R1-Distill-Qwen-14B
- sometimesanotion/Lamarck-14B-v0.3
- sometimesanotion/Qwenvergence-14B-v9
- sometimesanotion/Qwenvergence-14B-v3-Prose
- arcee-ai/Virtuoso-Small
library_name: transformers
tags:
- mergekit
- merge
license: apache-2.0
language:
- en
pipeline_tag: text-generation
metrics:
- accuracy
---
![Lamarck.webp](https://huggingface.co/sometimesanotion/Lamarck-14B-v0.7/resolve/main/LamarckShades.webp)
---

> [!TIP] With no regressions, mostly gains over the previous release, this version of Lamarck has [broken the 41.0 average](https://shorturl.at/jUqEk) maximum for 14B parameter models.  As of this writing, Lamarck v0.7 ranks #8 among models under 70B parameters on the Open LLM Leaderboard.  Given the quality models in the 32B range, I think Lamarck deserves his shades.  A little layer analysis of a model in the 14B range goes a long, long way.

> [!TIP] The first DPO finetune of Lamarck has appeared!  Check out [jpacifico/Chocolatine-2-14B-Instruct-v2.0b3](http://huggingface.co/jpacifico/Chocolatine-2-14B-Instruct-v2.0b3), whose notes say, "The Chocolatine model series is a quick demonstration that a base model can be easily fine-tuned to achieve compelling performance."  Lamarck's painstaking merge process was intended to make finetuning to a desired polish as easy and energy-efficient as possible.  Thank you, @jpacifico!

Lamarck 14B v0.7:  A generalist merge with emphasis on multi-step reasoning, prose, and multi-language ability.  The 14B parameter model class has a lot of strong performers, and Lamarck strives to be well-rounded and solid: ![14b.png](https://huggingface.co/sometimesanotion/Lamarck-14B-v0.7/resolve/main/14b.png)

Lamarck is produced by a custom toolchain to automate a complex sequences of LoRAs and various layer-targeting merges:

- **Extracted LoRA adapters from special-purpose merges**
- **Custom base models and model_stocks, with LoRAs from from [huihui-ai/Qwen2.5-14B-Instruct-abliterated-v2](https://huggingface.co/huihui-ai/Qwen2.5-14B-Instruct-abliterated-v2) to minimize IFEVAL loss often seen in model_stock merges**
- **Separate branches for aggressive breadcrumbs and conservative DELLA merges**
- **Highly targeted weight/density gradients for every 2-4 layers, at each stage**
- **Finalization through SLERP+TIES merges recombining the breadcrumbs and DELLA branches to taste**

Lamarck's performance comes from an ancestry that goes back through careful merges to select finetuning work, upcycled and combined.  Through intermediate merges, [arcee-ai/Virtuoso-Small](https://huggingface.co/arcee-ai/Virtuoso-Small) [sthenno-com/miscii-14b-1225](https://huggingface.co/sthenno-com/miscii-14b-1225) and [VAGOsolutions/SauerkrautLM-v2-14b-DPO](https://huggingface.co/VAGOsolutions/SauerkrautLM-v2-14b-DPO) are emphasized in early layers for extra BBH; later layers add synergistic influence from [deepseek-ai/DeepSeek-R1-Distill-Qwen-14B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B), [Krystalan/DRT-o1-14B](https://huggingface.co/Krystalan/DRT-o1-14B), [EVA-UNIT-01/EVA-Qwen2.5-14B-v0.2](https://huggingface.co/EVA-UNIT-01/EVA-Qwen2.5-14B-v0.2), and [CultriX/Qwen2.5-14B-Wernicke](https://huggingface.co/CultriX/Qwen2.5-14B-Wernicke).

More subjectively, its prose and translation abilities are boosted by repeated re-emphasis of [Krystalan/DRT-o1-14B](https://huggingface.co/Krystalan/DRT-o1-14B) and [underwoods/medius-erebus-magnum-14b](https://huggingface.co/underwoods/medius-erebus-magnum-14b).  Other models found in [sometimesanotion/Qwenvergence-14B-v3-Prose](https://huggingface/sometimesanotion/Qwenvergence-14B-v3-Prose) have their impact on prose quality - and surprising synergy of reasoning.

Kudos to @arcee-ai, @deepseek-ai, @Krystalan, @underwoods, @VAGOSolutions, @CultriX, @sthenno-com, and @rombodawg whose models had the most influence.  [Vimarckoso v3](https://huggingface.co/sometimesanotion/Qwen2.5-14B-Vimarckoso-v3) has the model card which documents its extended lineage.