mention the useful tiny model
Browse files
README.md
CHANGED
|
@@ -107,6 +107,8 @@ More information needed
|
|
| 107 |
</details>
|
| 108 |
|
| 109 |
|
|
|
|
|
|
|
| 110 |
# Training Details
|
| 111 |
|
| 112 |
We closel follow the training procedure layed out in [Flamingo](https://huggingface.co/papers/2204.14198). We combine two open-source pre-trained models ([laion/CLIP-ViT-H-14-laion2B-s32B-b79K](https://huggingface.co/laion/CLIP-ViT-H-14-laion2B-s32B-b79K) and [huggyllama/llama-65b](https://huggingface.co/huggyllama/llama-65b)) by initializing new Transformer blocks. The pre-trained backbones are frozen while we train the newly initialized parameters.
|
|
|
|
| 107 |
</details>
|
| 108 |
|
| 109 |
|
| 110 |
+
To quickly test your software without waiting for the huge model to download/load you can use `HuggingFaceM4/tiny-random-idefics` - it hasn't been trained and has random weights but it is very useful for quick testing.
|
| 111 |
+
|
| 112 |
# Training Details
|
| 113 |
|
| 114 |
We closel follow the training procedure layed out in [Flamingo](https://huggingface.co/papers/2204.14198). We combine two open-source pre-trained models ([laion/CLIP-ViT-H-14-laion2B-s32B-b79K](https://huggingface.co/laion/CLIP-ViT-H-14-laion2B-s32B-b79K) and [huggyllama/llama-65b](https://huggingface.co/huggyllama/llama-65b)) by initializing new Transformer blocks. The pre-trained backbones are frozen while we train the newly initialized parameters.
|