DiscoLM-70b / README.md
leaderboard-pr-bot's picture
Adding Evaluation Results
f607016 verified
|
raw
history blame
11.6 kB
metadata
language:
  - en
  - de
license: llama2
library_name: transformers
tags:
  - goliath
  - deutsch
  - llama2
  - discoresearch
datasets:
  - Open-Orca/SlimOrca-Dedup
  - teknium/openhermes
  - meta-math/MetaMathQA
  - migtissera/Synthia-v1.3
  - THUDM/AgentInstruct
  - LeoLM/German_Songs
  - LeoLM/German_Poems
  - LeoLM/OpenSchnabeltier
  - bjoernp/ultrachat_de
  - LDJnr/Capybara
pipeline_tag: text-generation
model_creator: DiscoResearch
model_type: llama
model-index:
  - name: DiscoLM-70b
    results:
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: AI2 Reasoning Challenge (25-Shot)
          type: ai2_arc
          config: ARC-Challenge
          split: test
          args:
            num_few_shot: 25
        metrics:
          - type: acc_norm
            value: 68.77
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=DiscoResearch/DiscoLM-70b
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: HellaSwag (10-Shot)
          type: hellaswag
          split: validation
          args:
            num_few_shot: 10
        metrics:
          - type: acc_norm
            value: 86.1
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=DiscoResearch/DiscoLM-70b
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MMLU (5-Shot)
          type: cais/mmlu
          config: all
          split: test
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 68.58
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=DiscoResearch/DiscoLM-70b
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: TruthfulQA (0-shot)
          type: truthful_qa
          config: multiple_choice
          split: validation
          args:
            num_few_shot: 0
        metrics:
          - type: mc2
            value: 57.64
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=DiscoResearch/DiscoLM-70b
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: Winogrande (5-shot)
          type: winogrande
          config: winogrande_xl
          split: validation
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 83.58
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=DiscoResearch/DiscoLM-70b
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: GSM8k (5-shot)
          type: gsm8k
          config: main
          split: test
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 63.53
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=DiscoResearch/DiscoLM-70b
          name: Open LLM Leaderboard

EM Logo

DiscoLM 70b

DiscoLM 70b is a 70b model based on Laion's LeoLM 70b which underwent additional continued pretraining for 65b tokens of German text, strengthening it's multilingual capabilities while retaining (and partially improving) English capabilities. This was then further finetuned on a combination of some the most popular open-source instruction sets. DiscoLM 70b is a DiscoResearch project and was trained by Björn Plüster.

Many thanks to LAION and HessianAI for scientific supervision, coordination and compute resources provided for this project on supercomputer 42 by HessianAI!

Table of Contents

  1. Download
  2. Benchmarks
  3. Prompt Format
  4. Dataset
  5. Acknowledgements
  6. Contact
  7. About DiscoResearch
  8. Disclaimer

Download

Huggingface GPTQ GGUF AWQ Base Model
Link @TheBloke @TheBloke @TheBloke LeoLM 70b

Benchmarks

Hugginface Leaderboard

This models is still an early Alpha and we can't guarantee that there isn't any contamination. The following are the scores from our own evaluation.

Metric Value
ARC (25-shot) 68.77
HellaSwag (10-shot) 85.41
MMLU (5-shot) 68.64
TruthfulQA (0-shot) 57.69
Winogrande (5-shot) 83.27
GSM8k (5-shot) 63.68
Avg. 71.24

The model is now also officially ranked on the Open LLM Leaderboard as #6 overall and as the second strongest Llama-2-70b based model (ranking only begind TigerBot 70b):

image/png (Screenshot from the 05. of December 2023)

We use Language Model Evaluation Harness to run the benchmark tests above, using the same version as the HuggingFace LLM Leaderboard.

FastEval

Metric Value
GSM8K 70.6
Math 17.8
BBH 63.4
MMLU 64.7
Avg. 48.87

Screenshot of the current (sadly no longer maintained) FastEval CoT leaderboard: FastEval Leaderboard

MTBench

{
    "first_turn": 7.9,
    "second_turn": 7.0625,
    "categories": {
        "writing": 9.55,
        "roleplay": 8.35,
        "reasoning": 6.15,
        "math": 4.7,
        "coding": 4.8,
        "extraction": 7.35,
        "stem": 9.1,
        "humanities": 9.85
    },
    "average": 7.48125
}

Screenshot of the current FastEval MT Bench leaderboard: FastEval Leaderboard

Prompt Format

This model follows the ChatML format:

<|im_start|>system
You are DiscoLM, a helpful assistant.
<|im_end|>
<|im_start|>user
Please tell me possible reasons to call a research collective "Disco Research"<|im_end|>
<|im_start|>assistant

This formatting is also available via a pre-defined Transformers chat template, which means that lists of messages can be formatted for you with the apply_chat_template() method:

chat = [
  {"role": "system", "content": "You are DiscoLM, a helpful assistant."},
  {"role": "user", "content": "Please tell me possible reasons to call a research collective Disco Research"}
]
tokenizer.apply_chat_template(chat, tokenize=False, add_generation_prompt=True)

If you use tokenize=True and return_tensors="pt" instead, then you will get a tokenized and formatted conversation ready to pass to model.generate().

Dataset

The dataset curation for DiscoLM 70b followed a "brute force"/"PoC" approach.

The following datasets were used for training DiscoLM 70b:

Many thanks for all dataset providers/curators!

Contact

Best way to reach us is on our Discord.

About DiscoResearch

DiscoResearch is an aspiring open research community. Disco should be a place where researchers from many communities can come together to combine their expertise and create innovative and groundbreaking LLMs. Come join our Discord, share your opinions and ideas, and advance open LLM research with us!

Acknowledgements

Disco 70b is a DiscoResearch project and was trained by Björn Plüster. Jan Harries helped with technical adivce, logistics and the Model Card. AutoMeta also provided helpful technical advice and rounded up his connections to select a set of high-quality datasets. The model was trained with compute provided by HessianAI in collaboration with LAION - many thanks in particular to Patrick Schramowski for his support.

We are standing on the shoulders of giants; many thanks in no particular order to Laion for LeoLM 70b (especially to Christoph Schuhmann who got us all connected), TheBloke for providing quantized versions, winglian for Axolotl which was used to train the model and the SlimOrca dataset, garage-bAInd, Teknium, Migel Tissera, MetaMath, and LDJnr for their great datasets (please contact us if we forgot to mention you here!).

Built with Axolotl

Disclaimer

The license on this model does not constitute legal advice. We are not responsible for the actions of third parties who use this model. This model should only be used for research purposes. The original Llama2 license and all restrictions of datasets used to train this model apply.

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric Value
Avg. 71.37
AI2 Reasoning Challenge (25-Shot) 68.77
HellaSwag (10-Shot) 86.10
MMLU (5-Shot) 68.58
TruthfulQA (0-shot) 57.64
Winogrande (5-shot) 83.58
GSM8k (5-shot) 63.53