Upload nanoVLM using push_to_hub
Browse files- README.md +22 -0
- model.safetensors +3 -0
README.md
ADDED
|
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
|
| 2 |
+
---
|
| 3 |
+
language: en
|
| 4 |
+
license: mit
|
| 5 |
+
library_name: nanofm
|
| 6 |
+
tags:
|
| 7 |
+
- image-generation
|
| 8 |
+
- flow-matching
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
# nanoFlowMatching - ngromb/nanoFlowMatching-13M
|
| 12 |
+
|
| 13 |
+
This is a nano implementation of the DiT-LLaMA flow-matching model, trained as part of the CS-503 course.
|
| 14 |
+
|
| 15 |
+
## Model Description
|
| 16 |
+
The model consists of a transformer-based architecture that predicts the velocity field for flow matching.
|
| 17 |
+
It takes as input a noisy image, a timestep, and a class label, and outputs the predicted noise/velocity field.
|
| 18 |
+
The architecture includes adaptive layer norm modulation (AdaLN) for conditioning on both timestep and class labels.
|
| 19 |
+
|
| 20 |
+
## Usage
|
| 21 |
+
You can load this model by adapting the DiT_Llama class from the `nanofm.modeling.dit` module,
|
| 22 |
+
to allow pretrained weights to be loaded from this repository.
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e251ba4441a78e0147da33f0259e6347afe53e9dde70b4b8ef82d26a62e58fd9
|
| 3 |
+
size 53250736
|