---
base_model:
- cstr/llama3.1-8b-spaetzle-v59
- cstr/llama3.1-8b-spaetzle-v63
- cstr/llama3.1-8b-spaetzle-v66
- cstr/llama3.1-8b-spaetzle-v73
tags:
- merge
- mergekit
license: llama3.1
language:
- en
- de
library_name: transformers
---

# llama3.1-8b-spaetzle-v74

llama3.1-8b-spaetzle-v74 is a merge of the following models:
* [cstr/llama3.1-8b-spaetzle-v59](https://huggingface.co/cstr/llama3.1-8b-spaetzle-v59)
* [cstr/llama3.1-8b-spaetzle-v63](https://huggingface.co/cstr/llama3.1-8b-spaetzle-v63)
* [cstr/llama3.1-8b-spaetzle-v66](https://huggingface.co/cstr/llama3.1-8b-spaetzle-v66)
* [cstr/llama3.1-8b-spaetzle-v73](https://huggingface.co/cstr/llama3.1-8b-spaetzle-v73)

EQ-Bench v2_de: 68.05 169/171, en: 75.27 - which is not the best, but it produces decent answers for some trick questions, and i have a sweet spot for that ;)

## 🧩 Configuration

```yamlmodels:
models:
  - model: cstr/llama3.1-8b-spaetzle-v59
    parameters:
      weight: 0.3
      density: 0.5
  - model: cstr/llama3.1-8b-spaetzle-v63
    parameters:
      weight: 0.15
      density: 0.5
  - model: cstr/llama3.1-8b-spaetzle-v66
    parameters:
      weight: 0.15
      density: 0.5
  - model: cstr/llama3.1-8b-spaetzle-v73
    parameters:
      weight: 0.4
      density: 0.5
base_model: cstr/llama3.1-8b-spaetzle-v59
merge_method: della_linear
parameters:
  int8_mask: true
  normalize: true
  epsilon: 0.1  
  lambda: 1.0   
  density: 0.7
dtype: bfloat16
```

## 💻 Usage

```python
!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "cstr/llama3.1-8b-spaetzle-v74"
messages = [{"role": "user", "content": "What is a large language model?"}]

tokenizer = AutoTokenizer.from_pretrained(model)
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    torch_dtype=torch.float16,
    device_map="auto",
)

outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])
```

## EU AI Act Art. 53 — provider obligations

Added 2026-08-02 during an account-wide provenance review.

**This is a model merge, not a format conversion.** Most `cstr/*` repositories
are GGUF conversions, where the upstream research team remains the provider of
the model and the conversion changes only the numeric representation of the
weights. A merge produces a model that did not previously exist, so under
Regulation (EU) 2024/1689 the maintainer of this repository is plausibly the
**provider** of it, and the duties that survive the Art. 53(2)
free-and-open-source exemption — Art. 53(1)(c) and 53(1)(d) — attach here rather
than upstream.

**Art. 53(1)(c) — copyright policy.** No training corpus was assembled by this
repository. Merging combines weights that other providers already published; it
performs no text or data mining, so no rights reservation under Art. 4(3) of
Directive (EU) 2019/790 was engaged by this step. Copyright questions arising
from how the constituent models were themselves trained attach to their
respective providers. Any credible claim that this repository redistributes
material it has no right to redistribute will be acted on — contact via the
Community tab.

**Art. 53(1)(d) — training content.** No data was used to train this model: it
is a weight-space combination of models trained by others, and its training
content is theirs. Of the 4 constituent models this card names, **1 are still published and 3 are not**: `cstr/llama3.1-8b-spaetzle-v63`, `cstr/llama3.1-8b-spaetzle-v66`, `cstr/llama3.1-8b-spaetzle-v73`. For those, the training-content chain cannot be followed from this card, and no summary is reconstructed here in their place — an untraceable summary presented as a traceable one would be worse than the gap.