Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
---
|
| 4 |
+
```
|
| 5 |
+
===================================================================================================================
|
| 6 |
+
Layer (type:depth-idx) Output Shape Param #
|
| 7 |
+
===================================================================================================================
|
| 8 |
+
MegaForMaskedLM [4, 2048, 50265] --
|
| 9 |
+
├─MegaModel: 1-1 [4, 2048, 768] --
|
| 10 |
+
│ └─MegaEmbeddings: 2-1 [4, 2048, 768] --
|
| 11 |
+
│ │ └─Embedding: 3-1 [4, 2048, 768] 38,603,520
|
| 12 |
+
│ └─ModuleList: 2-2 -- --
|
| 13 |
+
│ │ └─MegaBlock: 3-2 [2048, 4, 768] 6,202,626
|
| 14 |
+
│ │ └─MegaBlock: 3-3 [2048, 4, 768] 6,202,626
|
| 15 |
+
│ │ └─MegaBlock: 3-4 [2048, 4, 768] 6,202,626
|
| 16 |
+
│ │ └─MegaBlock: 3-5 [2048, 4, 768] 6,202,626
|
| 17 |
+
│ │ └─MegaBlock: 3-6 [2048, 4, 768] 6,202,626
|
| 18 |
+
│ │ └─MegaBlock: 3-7 [2048, 4, 768] 6,202,626
|
| 19 |
+
│ │ └─MegaBlock: 3-8 [2048, 4, 768] 6,202,626
|
| 20 |
+
│ │ └─MegaBlock: 3-9 [2048, 4, 768] 6,202,626
|
| 21 |
+
│ │ └─MegaBlock: 3-10 [2048, 4, 768] 6,202,626
|
| 22 |
+
│ │ └─MegaBlock: 3-11 [2048, 4, 768] 6,202,626
|
| 23 |
+
│ │ └─MegaBlock: 3-12 [2048, 4, 768] 6,202,626
|
| 24 |
+
│ │ └─MegaBlock: 3-13 [2048, 4, 768] 6,202,626
|
| 25 |
+
├─Linear: 1-2 [4, 2048, 50265] 38,653,785
|
| 26 |
+
===================================================================================================================
|
| 27 |
+
Total params: 151,688,817
|
| 28 |
+
Trainable params: 151,688,817
|
| 29 |
+
Non-trainable params: 0
|
| 30 |
+
Total mult-adds (G): 150.35
|
| 31 |
+
===================================================================================================================
|
| 32 |
+
Input size (MB): 0.07
|
| 33 |
+
Forward/backward pass size (MB): 10818.75
|
| 34 |
+
Params size (MB): 606.71
|
| 35 |
+
Estimated Total Size (MB): 11425.52
|
| 36 |
+
===================================================================================================================
|
| 37 |
+
```
|