Update README.md
Browse files
README.md
CHANGED
|
@@ -9,35 +9,4 @@ tags:
|
|
| 9 |
- gemma2
|
| 10 |
- gemma
|
| 11 |
---
|
| 12 |
-
|
| 13 |
-
## Reminder to use the dev version Transformers:
|
| 14 |
-
`pip install git+https://github.com/huggingface/transformers.git`
|
| 15 |
-
|
| 16 |
-
# Finetune Gemma, Llama 3, Mistral 2-5x faster with 70% less memory via Unsloth!
|
| 17 |
-
|
| 18 |
-
Directly quantized 4bit model with `bitsandbytes`.
|
| 19 |
-
|
| 20 |
-
We have a Google Colab Tesla T4 notebook for **Gemma 2 (9B)** here: https://colab.research.google.com/drive/1vIrqH5uYDQwsJ4-OO3DErvuv4pBgVwk4?usp=sharing
|
| 21 |
-
|
| 22 |
-
[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/Discord%20button.png" width="200"/>](https://discord.gg/u54VK8m8tk)
|
| 23 |
-
[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/buy%20me%20a%20coffee%20button.png" width="200"/>](https://ko-fi.com/unsloth)
|
| 24 |
-
[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
|
| 25 |
-
|
| 26 |
-
## ✨ Finetune for Free
|
| 27 |
-
|
| 28 |
-
All notebooks are **beginner friendly**! Add your dataset, click "Run All", and you'll get a 2x faster finetuned model which can be exported to GGUF, vLLM or uploaded to Hugging Face.
|
| 29 |
-
|
| 30 |
-
| Unsloth supports | Free Notebooks | Performance | Memory use |
|
| 31 |
-
|-----------------|--------------------------------------------------------------------------------------------------------------------------|-------------|----------|
|
| 32 |
-
| **Llama 3 (8B)** | [▶️ Start on Colab](https://colab.research.google.com/drive/135ced7oHytdxu3N2DNe1Z0kqjyYIkDXp?usp=sharing) | 2.4x faster | 58% less |
|
| 33 |
-
| **Gemma 2 (9B)** | [▶️ Start on Colab](https://colab.research.google.com/drive/1vIrqH5uYDQwsJ4-OO3DErvuv4pBgVwk4?usp=sharing) | 2x faster | 63% less |
|
| 34 |
-
| **Mistral (9B)** | [▶️ Start on Colab](https://colab.research.google.com/drive/1Dyauq4kTZoLewQ1cApceUQVNcnnNTzg_?usp=sharing) | 2.2x faster | 62% less |
|
| 35 |
-
| **Phi 3 (mini)** | [▶️ Start on Colab](https://colab.research.google.com/drive/1lN6hPQveB_mHSnTOYifygFcrO8C1bxq4?usp=sharing) | 2x faster | 63% less |
|
| 36 |
-
| **TinyLlama** | [▶️ Start on Colab](https://colab.research.google.com/drive/1AZghoNBQaMDgWJpi4RbffGM1h6raLUj9?usp=sharing) | 3.9x faster | 74% less |
|
| 37 |
-
| **CodeLlama (34B)** A100 | [▶️ Start on Colab](https://colab.research.google.com/drive/1y7A0AxE3y8gdj4AVkl2aZX47Xu3P1wJT?usp=sharing) | 1.9x faster | 27% less |
|
| 38 |
-
| **Mistral (7B)** 1xT4 | [▶️ Start on Kaggle](https://www.kaggle.com/code/danielhanchen/kaggle-mistral-7b-unsloth-notebook) | 5x faster\* | 62% less |
|
| 39 |
-
| **DPO - Zephyr** | [▶️ Start on Colab](https://colab.research.google.com/drive/15vttTpzzVXv_tJwEk-hIcQ0S9FcEWvwP?usp=sharing) | 1.9x faster | 19% less |
|
| 40 |
-
|
| 41 |
-
- This [conversational notebook](https://colab.research.google.com/drive/1Aau3lgPzeZKQ-98h69CCu1UJcvIBLmy2?usp=sharing) is useful for ShareGPT ChatML / Vicuna templates.
|
| 42 |
-
- This [text completion notebook](https://colab.research.google.com/drive/1ef-tab5bhkvWmBOObepl1WgJvfvSzn5Q?usp=sharing) is for raw text. This [DPO notebook](https://colab.research.google.com/drive/15vttTpzzVXv_tJwEk-hIcQ0S9FcEWvwP?usp=sharing) replicates Zephyr.
|
| 43 |
-
- \* Kaggle has 2x T4s, but we use 1. Due to overhead, 1x T4 is 5x faster.
|
|
|
|
| 9 |
- gemma2
|
| 10 |
- gemma
|
| 11 |
---
|
| 12 |
+
This is just a copy of unsloth/gemma-2-27b for experimental purposes.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|