README.md

metadata

language:
  - en
license: apache-2.0
library_name: transformers
datasets:
  - NeuralNovel/Neural-Story-v1
base_model: mistralai/Mistral-7B-Instruct-v0.2
inference: false
model-index:
  - name: Mistral-7B-Instruct-v0.2-Neural-Story
    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: 64.08
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=NeuralNovel/Mistral-7B-Instruct-v0.2-Neural-Story
          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: 83.97
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=NeuralNovel/Mistral-7B-Instruct-v0.2-Neural-Story
          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: 60.67
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=NeuralNovel/Mistral-7B-Instruct-v0.2-Neural-Story
          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: 66.89
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=NeuralNovel/Mistral-7B-Instruct-v0.2-Neural-Story
          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: 75.85
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=NeuralNovel/Mistral-7B-Instruct-v0.2-Neural-Story
          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: 38.29
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=NeuralNovel/Mistral-7B-Instruct-v0.2-Neural-Story
          name: Open LLM Leaderboard

NeuralNovel/Mistral-7B-Instruct-v0.2-Neural-Story

GGUF FILES HERE

The Mistral-7B-Instruct-v0.2-Neural-Story model, developed by NeuralNovel and funded by Techmind, is a language model finetuned from Mistral-7B-Instruct-v0.2.

Designed to generate instructive and narrative text, with a specific focus on storytelling. This fine-tune has been tailored to provide detailed and creative responses in the context of narrative and optimised for short story telling.

Based on mistralAI, with apache-2.0 license, suitable for commercial or non-commercial use.

Join our Discord!

Data-set

The model was finetuned using the Neural-Story-v1 dataset.

Benchmark

Metric	Value
Avg.	64.96
ARC	64.08
HellaSwag	66.89
MMLU	60.67
TruthfulQA	66.89
Winogrande	75.85
GSM8K	38.29

Evaluated on HuggingFaceH4/open_llm_leaderboard

Summary

Fine-tuned with the intention of generating creative and narrative text, making it more suitable for creative writing prompts and storytelling.

Out-of-Scope Use

The model may not perform well in scenarios unrelated to instructive and narrative text generation. Misuse or applications outside its designed scope may result in suboptimal outcomes.

Bias, Risks, and Limitations

The model may exhibit biases or limitations inherent in the training data. It is essential to consider these factors when deploying the model to avoid unintended consequences.

While the Neural-Story-v0.1 dataset serves as an excellent starting point for testing language models, users are advised to exercise caution, as there might be some inherent genre or writing bias.

Hardware and Training

  n_epochs = 3,
  n_checkpoints = 3,
  batch_size = 12,
  learning_rate = 1e-5,

Sincere appreciation to Techmind for their generous sponsorship.

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	64.96
AI2 Reasoning Challenge (25-Shot)	64.08
HellaSwag (10-Shot)	83.97
MMLU (5-Shot)	60.67
TruthfulQA (0-shot)	66.89
Winogrande (5-shot)	75.85
GSM8k (5-shot)	38.29