EpistemeAI2/Fireball-MathMistral-Nemo-Base-2407-v2dpo

Uploaded model

• Developed by: EpistemeAI
• License: apache-2.0
• Finetuned from model : EpistemeAI/Fireball-MathMistral-Nemo-Base-2407

This mistral model was trained 2x faster with Unsloth and Huggingface's TRL library.

---

Fireball-MathMistral-Nemo-Base-2407-v2dpo

This model is fine-tune to provide better math response than Mistral-Nemo-Base-2407, Google Gemma 2 9B, Llama 3.1 8B and others similar models.

---

Training Dataset

DPO (Direct Preference Optimization) training with math datasets.

---

This Mistral model was trained 2x faster with Unsloth and Huggingface's TRL library.

Model Card for EpistemeAI's Fireball-MathMistral-Nemo-Base-2407-v2dpo

The Fireball-MathMistral-Nemo-Base-2407 Large Language Model (LLM) is a pretrained generative text model of 12B parameters, it significantly outperforms existing models smaller or similar in size.

For more details about this model please refer to our release blog post.

Key features

• Released under the Apache 2 License
• Trained with a 128k context window
• Trained on a large proportion of multilingual and code data
• Drop-in replacement of Mistral 7B

Model Architecture

Mistral Nemo is a transformer model, with the following architecture choices:

• Layers: 40
• Dim: 5,120
• Head dim: 128
• Hidden dim: 14,436
• Activation Function: SwiGLU
• Number of heads: 32
• Number of kv-heads: 8 (GQA)
• Vocabulary size: 2**17 ~= 128k
• Rotary embeddings (theta = 1M)

Demo

After installing mistral_inference, a mistral-demo CLI command should be available in your environment.

Transformers

NOTE: Until a new release has been made, you need to install transformers from source:

pip install git+https://github.com/huggingface/transformers.git

If you want to use Hugging Face transformers to generate text, you can do something like this.

from transformers import AutoModelForCausalLM, AutoTokenizer
model_id = "EpistemeAI/Fireball-MathMistral-Nemo-Base-2407-v2dpo"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)
inputs = tokenizer("Hello my name is", return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=20)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Unlike previous Mistral models, Mistral Nemo requires smaller temperatures. We recommend to use a temperature of 0.3.

Note

EpistemeAI/Fireball-MathMistral-Nemo-Base-2407-v2dpo is a pretrained base model and therefore does not have any moderation mechanisms.