add architecture info

architectures/.ipynb_checkpoints/codeparrot-checkpoint.txt

@@ -0,0 +1,5 @@
+[CodeParrot](https://huggingface.co/lvwerra/codeparrot) uses GPT-2 architecture with BPE tokenizer trained on Python code.
+
+|Model | # parameters | 
+|   -   |   -  | 
+| GPT2 | 1.5B |

architectures/.ipynb_checkpoints/incoder-checkpoint.txt

@@ -0,0 +1,5 @@
+[InCoder](https://huggingface.co/facebook/incoder-6B) uses a decoder-only Transformer with [Causal Masking objective](https://arxiv.org/abs/2201.07520), to train a left-to-right language model to fill in masked token segments. 
+
+|Model | # parameters |
+|   -   |   -  | 
+| Decoder |6.7B |

architectures/.ipynb_checkpoints/opt-checkpoint.txt

@@ -0,0 +1,5 @@
+[OPT](https://huggingface.co/facebook/opt-30b) uses decoder-only models like GPT-3. It was trained on datasets with a small portion of code. In this demo we use the 30B parameters model. The largest model has 176B parameters.
+
+|Model | # parameters |
+|   -   |   -  | 
+| Decoder |30B |

architectures/codeparrot.txt

@@ -0,0 +1,5 @@
+[CodeParrot](https://huggingface.co/lvwerra/codeparrot) uses GPT-2 architecture with BPE tokenizer trained on Python code.
+
+|Model | # parameters | 
+|   -   |   -  | 
+| GPT2 | 1.5B |

architectures/incoder.txt

@@ -0,0 +1,5 @@
+[InCoder](https://huggingface.co/facebook/incoder-6B) uses a decoder-only Transformer with [Causal Masking objective](https://arxiv.org/abs/2201.07520), to train a left-to-right language model to fill in masked token segments. 
+
+|Model | # parameters |
+|   -   |   -  | 
+| Decoder |6.7B |

architectures/opt.txt

@@ -0,0 +1,5 @@
+[OPT](https://huggingface.co/facebook/opt-30b) uses decoder-only models like GPT-3. It was trained on datasets with a small portion of code. In this demo we use the 30B parameters model. The largest model has 176B parameters.
+
+|Model | # parameters |
+|   -   |   -  | 
+| Decoder |30B |