README.md

metadata

license: llama3
library_name: transformers
base_model:
  - nbeerbower/llama-3-Stheno-Mahou-8B
datasets:
  - flammenai/FlameMix-DPO-v1
  - flammenai/Grill-preprod-v1_chatML
  - flammenai/Grill-preprod-v2_chatML
model-index:
  - name: Mahou-1.2a-llama3-8B
    results:
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: IFEval (0-Shot)
          type: HuggingFaceH4/ifeval
          args:
            num_few_shot: 0
        metrics:
          - type: inst_level_strict_acc and prompt_level_strict_acc
            value: 50.93
            name: strict accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=flammenai/Mahou-1.2a-llama3-8B
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: BBH (3-Shot)
          type: BBH
          args:
            num_few_shot: 3
        metrics:
          - type: acc_norm
            value: 28.97
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=flammenai/Mahou-1.2a-llama3-8B
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MATH Lvl 5 (4-Shot)
          type: hendrycks/competition_math
          args:
            num_few_shot: 4
        metrics:
          - type: exact_match
            value: 7.55
            name: exact match
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=flammenai/Mahou-1.2a-llama3-8B
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: GPQA (0-shot)
          type: Idavidrein/gpqa
          args:
            num_few_shot: 0
        metrics:
          - type: acc_norm
            value: 5.15
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=flammenai/Mahou-1.2a-llama3-8B
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MuSR (0-shot)
          type: TAUR-Lab/MuSR
          args:
            num_few_shot: 0
        metrics:
          - type: acc_norm
            value: 6.02
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=flammenai/Mahou-1.2a-llama3-8B
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MMLU-PRO (5-shot)
          type: TIGER-Lab/MMLU-Pro
          config: main
          split: test
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 31.3
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=flammenai/Mahou-1.2a-llama3-8B
          name: Open LLM Leaderboard

Mahou-1.2a-llama3-8B

Mahou is our attempt to build a production-ready conversational/roleplay LLM.

Future versions will be released iteratively and finetuned from flammen.ai conversational data.

Chat Format

This model has been trained to use ChatML format.

<|im_start|>system
{{system}}<|im_end|>
<|im_start|>{{char}}
{{message}}<|im_end|>
<|im_start|>{{user}}
{{message}}<|im_end|>

Roleplay Format

• Speech without quotes.
• Actions in *asterisks*

*leans against wall cooly* so like, i just casted a super strong spell at magician academy today, not gonna lie, felt badass.

ST Settings

1. Use ChatML for the Context Template.
2. Turn on Instruct Mode for ChatML.
3. Use the following stopping strings: ["<", "|", "<|", "\n"]

Method

Finetuned using an A100 on Google Colab.

Fine-tune a Mistral-7b model with Direct Preference Optimization - Maxime Labonne

Configuration

LoRA, model, and training settings:

# LoRA configuration
peft_config = LoraConfig(
    r=16,
    lora_alpha=16,
    lora_dropout=0.05,
    bias="none",
    task_type="CAUSAL_LM",
    target_modules=['k_proj', 'gate_proj', 'v_proj', 'up_proj', 'q_proj', 'o_proj', 'down_proj']
)

# Model to fine-tune
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype=torch.bfloat16,
    load_in_4bit=True
)
model.config.use_cache = False

# Reference model
ref_model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype=torch.bfloat16,
    load_in_4bit=True
)

# Training arguments
training_args = TrainingArguments(
    per_device_train_batch_size=4,
    gradient_accumulation_steps=4,
    gradient_checkpointing=True,
    learning_rate=5e-5,
    lr_scheduler_type="cosine",
    max_steps=2000,
    save_strategy="no",
    logging_steps=1,
    output_dir=new_model,
    optim="paged_adamw_32bit",
    warmup_steps=100,
    bf16=True,
    report_to="wandb",
)

# Create DPO trainer
dpo_trainer = DPOTrainer(
    model,
    ref_model,
    args=training_args,
    train_dataset=dataset,
    tokenizer=tokenizer,
    peft_config=peft_config,
    beta=0.1,
    force_use_ref_model=True
)

# Fine-tune model with DPO
dpo_trainer.train()

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	21.65
IFEval (0-Shot)	50.93
BBH (3-Shot)	28.97
MATH Lvl 5 (4-Shot)	7.55
GPQA (0-shot)	5.15
MuSR (0-shot)	6.02
MMLU-PRO (5-shot)	31.30