"Updated notebook to fix some of the broken links"

notebooks/colab_finetuning.ipynb

@@ -188,7 +188,7 @@
       },
       "source": [
         "# Acquire a pretrained model\n",
-        "Pretrained METL models are available in the [metl-pretrained](https://github.com/gitter-lab/metl-pretrained) repository. You can use one of those, or you can pretrain your own METL model (see [pretraining.ipynb](pretraining.ipynb)).\n",
+        "Pretrained METL models are available in the [metl-pretrained](https://github.com/gitter-lab/metl-pretrained) repository. You can use one of those, or you can pretrain your own METL model (see [pretraining.ipynb](https://github.com/gitter-lab/metl/blob/main/notebooks/pretraining.ipynb)).\n",
         "\n",
         "For demonstration purposes, we include a pretrained avGFP METL-Local model from the [metl-pretrained](https://github.com/gitter-lab/metl-pretrained) repository in the [pretrained_models](https://github.com/gitter-lab/metl/tree/main/data/dms_data/avgfp/pretrained_models) directory. This model is `METL-L-2M-3D-GFP` (UUID: `Hr4GNHws`).\n",
         "It is the avGFP METL-Local source model we used for the analysis in our preprint.\n",
@@ -205,7 +205,7 @@
       "source": [
         "# Training arguments\n",
         "\n",
-        "The script for finetuning on experimental data is [train_target_model.py](train_target_model.py). This script has a number of arguments you can view by uncommenting and running the below cell. There are additional arguments related to architecture that won't show up if you run the command, but you can view them in [models.py](https://github.com/gitter-lab/metl/tree/main/data/dms_data/avgfp/code/models.py) in the `TransferModel` class."
+        "The script for finetuning on experimental data is [train_target_model.py](https://github.com/gitter-lab/metl/blob/main/code/train_target_model.py). This script has a number of arguments you can view by uncommenting and running the below cell. There are additional arguments related to architecture that won't show up if you run the command, but you can view them in [models.py](https://github.com/gitter-lab/metl/tree/main/data/dms_data/avgfp/code/models.py) in the `TransferModel` class."
       ]
     },
     {
@@ -227,7 +227,7 @@
         "id": "7ec8c31b-2da2-4ba7-9f4e-39e30dce8056"
       },
       "source": [
-        "We set up finetuning arguments for this example in [finetune_avgfp_local.txt](https://github.com/gitter-lab/metl/tree/main/data/dms_data/avgfp/args/pretrain_avgfp_local.txt) in the [args](https://github.com/gitter-lab/metl/tree/main/data/dms_data/avgfp/args) directory. This argument file can be used directly with [train_target_model.py](train_target_model.py) by calling the command `!python code/train_target_model.py @args/finetune_avgfp_local.txt` (we do this in the next section).\n",
+        "We set up finetuning arguments for this example in [finetune_avgfp_local.txt](https://github.com/gitter-lab/metl/tree/main/data/dms_data/avgfp/args/pretrain_avgfp_local.txt) in the [args](https://github.com/gitter-lab/metl/tree/main/data/dms_data/avgfp/args) directory. This argument file can be used directly with [train_target_model.py](https://github.com/gitter-lab/metl/blob/main/code/train_target_model.py) by calling the command `!python code/train_target_model.py @args/finetune_avgfp_local.txt` (we do this in the next section).\n",
         "\n",
         "Uncomment and run the cell below to view the contents of the argument file. The sections below will walk through and explain the key arguments."
       ]
@@ -477,7 +477,7 @@
         "id": "8d3d8d23-9d54-4888-842d-4fc8fd843b40"
       },
       "source": [
-        "All the arguments described above are contained in [finetune_avgfp_local.txt](https://github.com/gitter-lab/metl/tree/main/data/dms_data/avgfp/args/pretrain_avgfp_local.txt), which can be fed directly into [train_target_model.py](train_target_model.py).\n",
+        "All the arguments described above are contained in [finetune_avgfp_local.txt](https://github.com/gitter-lab/metl/tree/main/data/dms_data/avgfp/args/pretrain_avgfp_local.txt), which can be fed directly into [train_target_model.py](https://github.com/gitter-lab/metl/blob/main/code/train_target_model.py).\n",
         "\n",
         "PyTorch Lightning has a built-in progress bar that is convenient for seeing training progress, but it does not display correctly in Jupyter when calling the script with `!python`. We are going to disable the progress bar for by setting the flag `--enable_progress_bar false`. Instead, we implemented a simple print statement to track training progress, which we will enable with the flag `--enable_simple_progress_messages`."
       ]