add fastapi dependency in CI (#83)

* add fastapi dependency in CI

* update deps in README

Author: charlesfrye

.github/workflows/ci-cd.yml

@@ -35,7 +35,7 @@ jobs:
       - name: Install Modal
         run: |
           python -m pip install --upgrade pip
-          pip install modal pyyaml pandas
+          pip install modal fastapi pyyaml pandas
 
       - name: Prep config and data for CI
         run: |

README.md

@@ -34,8 +34,8 @@ Inference on the fine-tuned model displays conformity to the output structure (`
    <summary>Setting up</summary>
 
    1. Create a [Modal](https://modal.com/) account.
-   2. Install `modal` in your current Python virtual environment (`pip install modal`)
-   3. Set up a Modal token in your environment (`python3 -m modal setup`)
+   2. Install `modal` and `fastapi` in your current Python virtual environment (`pip install modal fastapi`).
+   3. Set up a Modal token in your environment (`python3 -m modal setup`).
    4. You need to have a [secret](https://modal.com/docs/guide/secrets#secrets) named `my-huggingface-secret` in your workspace. You can [create a new secret](https://modal.com/secrets) with the HuggingFace template in your Modal dashboard, using the key from HuggingFace (in settings under API tokens) to populate `HF_TOKEN`.
    5. For some LLaMA models, you need to go to the Hugging Face page (e.g. [this page for LLaMA 3 8B](https://huggingface.co/meta-llama/Meta-Llama-3-8B)\_ and agree to their Terms and Conditions for access (granted instantly).
    6. If you want to use [Weights & Biases](https://wandb.ai) for logging, you need to have a secret named `wandb` in your workspace as well. You can also create it [from a template](https://modal.com/secrets). Training is hard enough without good logs, so we recommend you try it or look into `axolotl`'s integration with [MLFlow](https://mlflow.org/)!