File size: 7,450 Bytes
de19672 eeafb78 de19672 eeafb78 de19672 |
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 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 |
---
license: llama3
language:
- de
library_name: transformers
---
# Llama3_DiscoLeo_Instruct_8B_v0.1
## Thanks and Accreditation
[DiscoResearch/Llama3_DiscoLeo_Instruct_8B_v0.1](https://huggingface.co/collections/DiscoResearch/discoleo-8b-llama3-for-german-6650527496c0fafefd4c9729)
is the result of a joint effort between [DiscoResearch](https://huggingface.co/DiscoResearch) and [Occiglot](https://huggingface.co/occiglot)
with support from the [DFKI](https://www.dfki.de/web/) (German Research Center for Artificial Intelligence) and [hessian.Ai](https://hessian.ai).
Occiglot kindly handled data preprocessing, filtering, and deduplication as part of their latest [dataset release](https://huggingface.co/datasets/occiglot/occiglot-fineweb-v0.5), as well as sharing their compute allocation at hessian.Ai's 42 Supercomputer.
## Model Overview
Llama3_DiscoLeo_Instruct_8B_v0 is an instruction tuned version of our [Llama3_German_8B](https://huggingface.co/DiscoResearch/Llama3_German_8B).
The base model was derived from [Meta's Llama3-8B](https://huggingface.co/meta-llama/Meta-Llama-3-8B) through continuous pretraining on 65 billion high-quality German tokens, similar to previous [LeoLM](https://huggingface.co/LeoLM) or [Occiglot](https://huggingface.co/collections/occiglot/occiglot-eu5-7b-v01-65dbed502a6348b052695e01) models.
We finetuned this checkpoint on the German Instruction dataset from DiscoResearch created by [Jan-Philipp Harries](https://huggingface.co/jphme) and [Daniel Auras](https://huggingface.co/rasdani) ([DiscoResearch](https://huggingface.co/DiscoResearch), [ellamind](https://ellamind.com)).
## How to use
Llama3_DiscoLeo_Instruct_8B_v0.1 uses the [Llama-3 chat template](https://github.com/meta-llama/llama3?tab=readme-ov-file#instruction-tuned-models), which can be easily used with [transformer's chat templating](https://huggingface.co/docs/transformers/main/en/chat_templating).
See [below](https://huggingface.co/DiscoResearch/Llama3_DiscoLeo_Instruct_8B_v0.1#usage-example) for a usage example.
## Model Training and Hyperparameters
The model was full-fintuned with axolotl on the [hessian.Ai 42](hessian.ai) with 8192 context-length, learning rate 2e-5 and batch size of 16.
## Evaluation and Results
We evaluated the model using a suite of common English Benchmarks and their German counterparts with [GermanBench](https://github.com/bjoernpl/GermanBenchmark).
In the below image and corresponding table, you can see the benchmark scores for the different instruct models compared to Metas instruct version. All checkpoints are available in this [collection](https://huggingface.co/collections/DiscoResearch/discoleo-8b-llama3-for-german-6650527496c0fafefd4c9729).
| Model | truthfulqa | truthful_qa_de | arc_challenge | arc_challenge_de | hellaswag | hellaswag_de | MMLU | MMLU_DE | mean |
|---------------------------------------------------|-------------|----------------|----------------|-------------------|-------------|--------------|----------|----------|-----------|
| DiscoResearch/Llama3_DiscoLeo_Instruct_8B_v0.1 | **0.530425** | 0.528673 | 0.595563 | **0.538396** | 0.807210| 0.664409 | 0.618989 | 0.560536 | **0.605525**|
| DiscoResearch/Llama3_DiscoLeo_Instruct_8B_32k_v0.1| 0.527493 | **0.532451** | 0.587884 | 0.537543 | **0.807708**| **0.667098** | 0.621234 | **0.562389** | 0.605475 |
| meta-llama/Meta-Llama-3-8B-Instruct | 0.516810 | 0.526288 | **0.613481** | 0.498294 | 0.785401 | 0.562537 | **0.669585** | 0.558135 | 0.591316 |
## Model Configurations
We release DiscoLeo-8B in the following configurations:
1. [Base model with continued pretraining](https://huggingface.co/DiscoResearch/Llama3_German_8B)
2. [Long-context version (32k context length)](https://huggingface.co/DiscoResearch/Llama3_German_8B_32k)
3. [Instruction-tuned version of the base model](https://huggingface.co/DiscoResearch/Llama3_DiscoLeo_Instruct_8B_v0.1) (This model)
4. [Instruction-tuned version of the long-context model](https://huggingface.co/DiscoResearch/Llama3_DiscoLeo_Instruct_8B_32k_v0.1)
5. [Experimental `DARE-TIES` Merge with Llama3-Instruct](https://huggingface.co/DiscoResearch/Llama3_DiscoLeo_8B_DARE_Experimental)
6. [Collection of Quantized versions](https://huggingface.co/collections/DiscoResearch/discoleo-8b-quants-6651bcf8f72c9a37ce485d42)
## Usage Example
Here's how to use the model with transformers:
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained(
"DiscoResearch/Llama3_DiscoLeo_Instruct_8B_v0.1",
torch_dtype="auto",
device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained("DiscoResearch/Llama3_DiscoLeo_Instruct_8B_v0.1")
prompt = "Schreibe ein Essay über die Bedeutung der Energiewende für Deutschlands Wirtschaft"
messages = [
{"role": "system", "content": "Du bist ein hilfreicher Assistent."},
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(device)
generated_ids = model.generate(
model_inputs.input_ids,
max_new_tokens=512
)
generated_ids = [
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
```
## Acknowledgements
The model was trained and evaluated by [Björn Plüster](https://huggingface.co/bjoernp) ([DiscoResearch](https://huggingface.co/DiscoResearch), [ellamind](https://ellamind.com)) with data preparation and project supervision by [Manuel Brack](http://manuel-brack.eu) ([DFKI](https://www.dfki.de/web/), [TU-Darmstadt](https://www.tu-darmstadt.de/)). Instruction tuning was done with the DiscoLM German dataset created by [Jan-Philipp Harries](https://huggingface.co/jphme) and [Daniel Auras](https://huggingface.co/rasdani) ([DiscoResearch](https://huggingface.co/DiscoResearch), [ellamind](https://ellamind.com)). We extend our gratitude to [LAION](https://laion.ai/) and friends, especially [Christoph Schuhmann](https://entwickler.de/experten/christoph-schuhmann) and [Jenia Jitsev](https://huggingface.co/JJitsev), for initiating this collaboration.
The model training was supported by a compute grant at the [42 supercomputer](https://hessian.ai/) which is a central component in the development of [hessian AI](https://hessian.ai/), the [AI Innovation Lab](https://hessian.ai/infrastructure/ai-innovationlab/) (funded by the [Hessian Ministry of Higher Education, Research and the Art (HMWK)](https://wissenschaft.hessen.de) & the [Hessian Ministry of the Interior, for Security and Homeland Security (HMinD)](https://innen.hessen.de)) and the [AI Service Centers](https://hessian.ai/infrastructure/ai-service-centre/) (funded by the [German Federal Ministry for Economic Affairs and Climate Action (BMWK)](https://www.bmwk.de/Navigation/EN/Home/home.html)).
The curation of the training data is partially funded by the [German Federal Ministry for Economic Affairs and Climate Action (BMWK)](https://www.bmwk.de/Navigation/EN/Home/home.html)
through the project [OpenGPT-X](https://opengpt-x.de/en/) (project no. 68GX21007D). |