FusionNet_34Bx2_MoE / README.md
TomGrc's picture
Adding Evaluation Results (#2)
a077e8d verified
|
raw
history blame
3.93 kB
metadata
language:
  - en
license: mit
tags:
  - moe
pipeline_tag: text-generation
model-index:
  - name: FusionNet_34Bx2_MoE
    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: 72.95
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TomGrc/FusionNet_34Bx2_MoE
          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: 86.22
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TomGrc/FusionNet_34Bx2_MoE
          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: 77.05
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TomGrc/FusionNet_34Bx2_MoE
          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: 71.31
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TomGrc/FusionNet_34Bx2_MoE
          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: 83.98
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TomGrc/FusionNet_34Bx2_MoE
          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: 70.89
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=TomGrc/FusionNet_34Bx2_MoE
          name: Open LLM Leaderboard

FusionNet_34Bx2_MoE

Fine-tuned model on English language using MoE method.

Model description

The FusionNet_34Bx2_MoE is a model to experiment with the MoE method, which could significantly increase the performance of the original model. The FusionNet_34Bx2_MoE has 60.8B parameters, and this model is fine-tuned. Enjoy!

Usage

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("TomGrc/FusionNet_34Bx2_MoE")
model = AutoModelForCausalLM.from_pretrained("TomGrc/FusionNet_34Bx2_MoE")

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric Value
Avg. 77.07
AI2 Reasoning Challenge (25-Shot) 72.95
HellaSwag (10-Shot) 86.22
MMLU (5-Shot) 77.05
TruthfulQA (0-shot) 71.31
Winogrande (5-shot) 83.98
GSM8k (5-shot) 70.89