README.md

metadata

license: apache-2.0
model-index:
  - name: open-llama-3b-v2-chat
    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: 40.61
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mediocredev/open-llama-3b-v2-chat
          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: 70.3
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mediocredev/open-llama-3b-v2-chat
          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: 28.73
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mediocredev/open-llama-3b-v2-chat
          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: 37.84
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mediocredev/open-llama-3b-v2-chat
          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: 65.51
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mediocredev/open-llama-3b-v2-chat
          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: 2.58
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=mediocredev/open-llama-3b-v2-chat
          name: Open LLM Leaderboard

Prerequisites

In addition to pytorch and transformers, install required packages:

pip install sentencepiece

Usage

To use, copy the following script:

ffrom transformers import AutoModelForCausalLM, AutoTokenizer

model_id = 'mediocredev/open-llama-3b-v2-chat'
tokenizer_id = 'mediocredev/open-llama-3b-v2-chat'
tokenizer = AutoTokenizer.from_pretrained(tokenizer_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

chat_history = [
    {"role": "user", "content": "Hello!"},
    {"role": "assistant", "content": "I am here."},
    {"role": "user", "content": "How many days are there in a leap year?"},
]

input_ids = tokenizer.apply_chat_template(
    chat_history, tokenize=True, add_generation_prompt=True, return_tensors="pt"
).to(model.device)
output_tokens = model.generate(
    input_ids,
    repetition_penalty=1.05,
    max_new_tokens=1000,
)
output_text = tokenizer.decode(
    output_tokens[0][len(input_ids[0]) :], skip_special_tokens=True
)

print(output_text)
# Assistant: There are 366 days in a leap year, which is one more day than the standard year.

Limitations

mediocredev/open-llama-3b-v2-chat is based on LLaMA 3B v2. It can struggle with factual accuracy, particularly when presented with conflicting information or nuanced topics. Its outputs are not deterministic and require critical evaluation to avoid relying solely on its assertions. Additionally, its generative capabilities, while promising, can sometimes produce factually incorrect or offensive content, necessitating careful curation and human oversight. As an evolving model, LLaMA is still under development, and its limitations in areas like bias mitigation and interpretability are being actively addressed. By using this model responsibly and being aware of its shortcomings, we can unlock its potential while mitigating its risks.

Contact

Welcome any feedback, questions, and discussions. Feel free to reach out: [email protected]

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	40.93
AI2 Reasoning Challenge (25-Shot)	40.61
HellaSwag (10-Shot)	70.30
MMLU (5-Shot)	28.73
TruthfulQA (0-shot)	37.84
Winogrande (5-shot)	65.51
GSM8k (5-shot)	2.58