README.md

metadata

language:
  - en
license: mit
library_name: transformers
tags:
  - peft
datasets:
  - HuggingFaceH4/ultrachat_200k
model-index:
  - name: Maixtchup-4x7b-QLoRA-SFT-UltraChat
    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: 60.92
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=kaitchup/Maixtchup-4x7b-QLoRA-SFT-UltraChat
          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: 83.23
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=kaitchup/Maixtchup-4x7b-QLoRA-SFT-UltraChat
          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: 60.78
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=kaitchup/Maixtchup-4x7b-QLoRA-SFT-UltraChat
          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: 53.33
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=kaitchup/Maixtchup-4x7b-QLoRA-SFT-UltraChat
          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: 77.19
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=kaitchup/Maixtchup-4x7b-QLoRA-SFT-UltraChat
          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: 43.21
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=kaitchup/Maixtchup-4x7b-QLoRA-SFT-UltraChat
          name: Open LLM Leaderboard

LoRA adapter for kaitchup/Maixtchup-4x7b briefly fine-tuned on UltraChat.

To load and use this adapter:

model_name = "kaitchup/Maixtchup-4x7b"
#Tokenizer
tokenizer = AutoTokenizer.from_pretrained(model_name, use_fast=True)
compute_dtype = getattr(torch, "float16")
bnb_config = BitsAndBytesConfig(
        load_in_4bit=True,
        bnb_4bit_quant_type="nf4",
        bnb_4bit_compute_dtype=compute_dtype,
        bnb_4bit_use_double_quant=True,
)
model = AutoModelForCausalLM.from_pretrained(
          model_name, quantization_config=bnb_config, device_map="auto", attn_implementation="flash_attention_2",
)

model.config.use_cache = True

model = PeftModel.from_pretrained(model, "kaitchup/Maixtchup-4x7b-QLoRA-SFT-UltraChat")

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	63.11
AI2 Reasoning Challenge (25-Shot)	60.92
HellaSwag (10-Shot)	83.23
MMLU (5-Shot)	60.78
TruthfulQA (0-shot)	53.33
Winogrande (5-shot)	77.19
GSM8k (5-shot)	43.21