Update README.md
Browse files
README.md
CHANGED
|
@@ -16,6 +16,7 @@ Welcome to join us!
|
|
| 16 |
Organization page: https://m-a-p.ai
|
| 17 |
|
| 18 |
The development log of our Multimodal Art Projection (m-a-p) model family:
|
|
|
|
| 19 |
- 13/01/2024: 🔥 we release a series of **Music Pretrained Transformer (MuPT)** checkpoints, with [**size up to 1.3B and 8192 context length**](https://huggingface.co/m-a-p/MuPT_v0_8192_1.3B). Our models are **LLAMA2**-based, pre-trained on **world's largest 10B tokens symbolic music dataset** ([ABC notation format](https://en.wikipedia.org/wiki/ABC_notation)). We currently support Megatron-LM format and will release huggingface checkpoints soon.
|
| 20 |
- 02/06/2023: officially release the [MERT pre-print paper](https://arxiv.org/abs/2306.00107) and training [codes](https://github.com/yizhilll/MERT).
|
| 21 |
- 17/03/2023: we release two advanced music understanding models, [MERT-v1-95M](https://huggingface.co/m-a-p/MERT-v1-95M) and [MERT-v1-330M](https://huggingface.co/m-a-p/MERT-v1-330M) , trained with new paradigm and dataset. They outperform the previous models and can better generalize to more tasks.
|
|
|
|
| 16 |
Organization page: https://m-a-p.ai
|
| 17 |
|
| 18 |
The development log of our Multimodal Art Projection (m-a-p) model family:
|
| 19 |
+
- 23/01/2024: we release [**CMMMU**](https://huggingface.co/datasets/m-a-p/CMMMU) for better Chinese LMMs' Evaluation.
|
| 20 |
- 13/01/2024: 🔥 we release a series of **Music Pretrained Transformer (MuPT)** checkpoints, with [**size up to 1.3B and 8192 context length**](https://huggingface.co/m-a-p/MuPT_v0_8192_1.3B). Our models are **LLAMA2**-based, pre-trained on **world's largest 10B tokens symbolic music dataset** ([ABC notation format](https://en.wikipedia.org/wiki/ABC_notation)). We currently support Megatron-LM format and will release huggingface checkpoints soon.
|
| 21 |
- 02/06/2023: officially release the [MERT pre-print paper](https://arxiv.org/abs/2306.00107) and training [codes](https://github.com/yizhilll/MERT).
|
| 22 |
- 17/03/2023: we release two advanced music understanding models, [MERT-v1-95M](https://huggingface.co/m-a-p/MERT-v1-95M) and [MERT-v1-330M](https://huggingface.co/m-a-p/MERT-v1-330M) , trained with new paradigm and dataset. They outperform the previous models and can better generalize to more tasks.
|