--- language: - en license: cc-by-sa-4.0 library_name: transformers base_model: - elinas/Llama-3-15B-Instruct-zeroed datasets: - TheSkullery/Aether-Lite-v1.8.1 model-index: - name: L3-Aethora-15B-V2 results: - task: type: text-generation name: Text Generation dataset: name: IFEval (0-Shot) type: HuggingFaceH4/ifeval args: num_few_shot: 0 metrics: - type: inst_level_strict_acc and prompt_level_strict_acc value: 72.08 name: strict accuracy source: url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=ZeusLabs/L3-Aethora-15B-V2 name: Open LLM Leaderboard - task: type: text-generation name: Text Generation dataset: name: BBH (3-Shot) type: BBH args: num_few_shot: 3 metrics: - type: acc_norm value: 28.97 name: normalized accuracy source: url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=ZeusLabs/L3-Aethora-15B-V2 name: Open LLM Leaderboard - task: type: text-generation name: Text Generation dataset: name: MATH Lvl 5 (4-Shot) type: hendrycks/competition_math args: num_few_shot: 4 metrics: - type: exact_match value: 7.33 name: exact match source: url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=ZeusLabs/L3-Aethora-15B-V2 name: Open LLM Leaderboard - task: type: text-generation name: Text Generation dataset: name: GPQA (0-shot) type: Idavidrein/gpqa args: num_few_shot: 0 metrics: - type: acc_norm value: 5.03 name: acc_norm source: url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=ZeusLabs/L3-Aethora-15B-V2 name: Open LLM Leaderboard - task: type: text-generation name: Text Generation dataset: name: MuSR (0-shot) type: TAUR-Lab/MuSR args: num_few_shot: 0 metrics: - type: acc_norm value: 6.25 name: acc_norm source: url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=ZeusLabs/L3-Aethora-15B-V2 name: Open LLM Leaderboard - task: type: text-generation name: Text Generation dataset: name: MMLU-PRO (5-shot) type: TIGER-Lab/MMLU-Pro config: main split: test args: num_few_shot: 5 metrics: - type: acc value: 27.78 name: accuracy source: url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=ZeusLabs/L3-Aethora-15B-V2 name: Open LLM Leaderboard --- L3-Aethora-15B v2 Data Card

L3-Aethora-15B v2

Presented by:

Creators: ZeusLabs

Dataset: Theskullery/Aether-Lite-V1.8.1

Trained: 4 x A100 for 17.5 hours on 125k samples

Sponsored by: Garg (@g4rg)

About L3-Aethora-15B v2:

 L3 = Llama3

L3-Aethora-15B v2 is an advanced language model built upon the Llama 3 architecture. It employs state-of-the-art training techniques and a curated dataset to deliver enhanced performance across a wide range of tasks.

(Thank you all for the interest! the model has surpassed 260k downloads on all formats!)

Quants:

Training Process:

Model Capabilities:

The goal of L3-Aethora-15B v2 is to have an expanded proficiency across a wide spectrum of tasks with a focus in creative writing:

Dataset Creation Process:

The Aether-Lite-V1.8.1 dataset used for training L3-Aethora-15B v2 underwent a rigorous creation and curation process:

  1. Data Collection: Aggregated from 12 diverse high-quality datasets, including:
    • jondurbin/airoboros-3.2
    • jtatman/medical-sci-instruct-100k-sharegpt
    • Doctor-Shotgun/no-robots-sharegpt
    • QuietImpostor/Sao10K-Claude-3-Opus-Instruct-15K-ShareGPT
    • TheSkullery/WizardLM_evol_instruct_v2_Filtered_Fuzzy_Dedup_ShareGPT
    • TheSkullery/Gryphe-Opus-WritingPrompts-merged
    • Alignment-Lab-AI/RPGuild-sharegpt-filtered
    • And others, providing a rich mix of instruction, creative writing, and specialized knowledge
  2. Data Preprocessing:
    • Language Detection: Utilized a FastText language model to ensure English-language content
    • Text Sanitization: Cleaned and normalized text, removing or replacing problematic characters
    • Phrase Filtering: Removed specific unwanted phrases and content types
  3. Deduplication:
    • Implemented advanced fuzzy deduplication with a 95% similarity threshold
    • Utilized text embeddings and cosine similarity calculations for efficient comparison
    • Removed 16,250 duplicate entries, ensuring dataset uniqueness
  4. Data Balancing:
    • Carefully sampled from each source dataset to maintain diversity
    • Implemented data shuffling to ensure random distribution of samples

The final dataset comprises 125,119 high-quality, diverse samples, striking a balance between creativity, practical knowledge, and intellectual depth.

The full dataset used has been released to the public and is avalible for all (see presented section), any ideas or recomendations are always welcome to expand on the dataset further

# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard) Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_ZeusLabs__L3-Aethora-15B-V2) | Metric |Value| |-------------------|----:| |Avg. |24.57| |IFEval (0-Shot) |72.08| |BBH (3-Shot) |28.97| |MATH Lvl 5 (4-Shot)| 7.33| |GPQA (0-shot) | 5.03| |MuSR (0-shot) | 6.25| |MMLU-PRO (5-shot) |27.78|