leaderboard-pr-bot's picture
Adding Evaluation Results
e96e7cf verified
|
raw
history blame
5.16 kB
metadata
language:
  - en
license: apache-2.0
library_name: transformers
datasets:
  - yahma/alpaca-cleaned
pipeline_tag: text-generation
model-index:
  - name: speechless-mistral-moloras-7b
    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: 59.98
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=uukuguy/speechless-mistral-moloras-7b
          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: 83.29
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=uukuguy/speechless-mistral-moloras-7b
          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: 64.12
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=uukuguy/speechless-mistral-moloras-7b
          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: 42.15
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=uukuguy/speechless-mistral-moloras-7b
          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: 78.37
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=uukuguy/speechless-mistral-moloras-7b
          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: 37.68
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=uukuguy/speechless-mistral-moloras-7b
          name: Open LLM Leaderboard

speechless-mistral-moloras-7b

4-bit GGUF models for CPU+GPU inference

This model is the static version of moloras (Mixture-of-multi-LoRAs) based on the following 6 Mistral-based LoRa modules.

  • Intel/neural-chat-7b-v3-1
  • migtissera/SynthIA-7B-v1.3
  • jondurbin/airoboros-m-7b-3.1.2
  • bhenrym14/mistral-7b-platypus-fp16
  • teknium/CollectiveCognition-v1.1-Mistral-7B
  • uukuguy/speechless-mistral-dolphin-orca-platypus-samantha-7b

Totally 6 LoRA modules from speechless-mistral-7b-dare-0.85

The router of mixture-of-multi-loras enables an automatic assembling of LoRA modules, using a gradientfree approach to obtain the coefficients of LoRA modules and requiring only a handful of inference steps for unseen tasks.

Code: https://github.com/uukuguy/multi_loras?tab=readme-ov-file#mixture-of-multi-loras

LM-Evaluation-Harness

Open LLM Leaderboard

Metric Value
ARC 59.98
HellaSwag 83.29
MMLU 64.12
TruthfulQA 42.15
Winogrande 78.37
GSM8K 37.68
Average 60.93

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric Value
Avg. 60.93
AI2 Reasoning Challenge (25-Shot) 59.98
HellaSwag (10-Shot) 83.29
MMLU (5-Shot) 64.12
TruthfulQA (0-shot) 42.15
Winogrande (5-shot) 78.37
GSM8k (5-shot) 37.68