📄 Sync docs with latest code changes

CONTRIBUTING.md

@@ -65,4 +65,4 @@ By contributing, you agree that your contributions will be licensed under the Ap
 
 ## Questions?
 
-Feel free to contact us at [support@slashml.com](mailto:support@slashml.com) if you have any questions about contributing!
+Feel free to contact us at [support@slashml.com](mailto:support@slashml.com) if you have any questions about contributing!

about.mdx

@@ -22,9 +22,9 @@ Do submit your feature requests at https://magemaker.featurebase.app/
 - Querying within Magemaker currently only works with text-based models
 - Deleting a model is not instant, it may show up briefly after deletion
 - Deploying the same model within the same minute will break
-- Hugging-face models on Azure have different Ids than their Hugging-face counterparts. Follow the steps specified in the quick-start guide to find the relevant models
-- For Azure deploying models other than Hugging-face is not supported yet. 
-- Python3.13 is not supported because of an open-issue by Azure. https://github.com/Azure/azure-sdk-for-python/issues/37600
+- Hugging Face models on Azure have different IDs than their Hugging Face counterparts. Follow the steps specified in the quick-start guide to find the relevant models.
+- For Azure, deploying models other than Hugging Face is not supported yet.
+- Python 3.12 is **not** supported because of an open issue in the Azure SDK (see https://github.com/Azure/azure-sdk-for-python/issues/37600).
 
 
 If there is anything we missed, do point them out at https://magemaker.featurebase.app/

concepts/contributing.mdx

@@ -165,4 +165,4 @@ We are committed to providing a welcoming and inclusive experience for everyone.
 
 ## License
 
-By contributing to Magemaker, you agree that your contributions will be licensed under the Apache 2.0 License.
+By contributing to Magemaker, you agree that your contributions will be licensed under the Apache 2.0 License.

concepts/deployment.mdx

@@ -62,8 +62,8 @@ deployment: !Deployment
   destination: gcp
   endpoint_name: opt-125m-gcp
   instance_count: 1
-  machine_type: n1-standard-4
-  accelerator_type: NVIDIA_TESLA_T4
+  instance_type: n1-standard-4
+  accelerator_type: NVIDIA_TESLA_T4   # or NVIDIA_L4
   accelerator_count: 1
 
 models:
@@ -113,6 +113,7 @@ deployment: !Deployment
   instance_count: 1
   instance_type: ml.g5.12xlarge
   num_gpus: 4
+  # quantization: bitsandbytes   # Optional
 
 models:
   - !Model
@@ -202,10 +203,9 @@ Choose your instance type based on your model's requirements:
 4. Set up monitoring and alerting for your endpoints
 
 <Warning>
-Make sure you setup budget monitory and alerts to avoid unexpected charges.
+Make sure you set up budget monitoring and alerts to avoid unexpected charges.
 </Warning>
 
-
 ## Troubleshooting Deployments
 
 Common issues and their solutions:
@@ -225,4 +225,4 @@ Common issues and their solutions:
    - Verify model ID and version
    - Check instance memory requirements
    - Validate Hugging Face token if required
-   - Endpoing deployed but deployment failed. Check the logs, and do report this to us if you see this issue.
+   - Endpoint deployed but deployment failed. Check the logs, and report this to us if you see this issue.

concepts/fine-tuning.mdx

@@ -5,7 +5,7 @@ description: Guide to fine-tuning models with Magemaker
 
 ## Fine-tuning Overview
 
-Fine-tuning allows you to adapt pre-trained models to your specific use case. Magemaker simplifies this process through YAML configuration.
+Fine-tuning allows you to adapt pre-trained models to your specific use case. Currently, Magemaker supports fine-tuning **on AWS SageMaker only** (GCP and Azure fine-tuning are on the roadmap). The workflow is entirely YAML-driven so it can be automated in CI/CD pipelines.
 
 ### Basic Command
 
@@ -19,19 +19,25 @@ magemaker --train .magemaker_config/train-config.yaml
 
 ```yaml
 training: !Training
-  destination: aws
-  instance_type: ml.p3.2xlarge
+  destination: aws                 # only "aws" is supported for now
+  instance_type: ml.p3.2xlarge      # GPU instance for training
   instance_count: 1
-  training_input_path: s3://your-bucket/training-data.csv
+  training_input_path: s3://your-bucket/training-data.csv  # points to your dataset
 
 models:
-- !Model
-  id: your-model-id
-  source: huggingface
+  - !Model
+    id: your-model-id
+    source: huggingface
 ```
 
+• **destination** – must be `aws` at the moment.  
+• **training_input_path** – S3 URI that Magemaker will pass directly to SageMaker.  
+• **instance_type / instance_count** – any SageMaker training instance type is supported.
+
 ### Advanced Configuration
 
+Beyond the basics, you can supply custom hyperparameters. If omitted, Magemaker will attempt to infer sensible defaults based on the model family (see `get_hyperparameters_for_model()` in the codebase).
+
 ```yaml
 training: !Training
   destination: aws
@@ -49,20 +55,38 @@ training: !Training
     save_steps: 1000
 ```
 
+<Note>
+If you omit `hyperparameters`, Magemaker will fall back to task-specific defaults. For example, text-generation models automatically receive the hyperparameters returned by `get_hyperparameters_for_model()`.
+</Note>
+
+### Optional Parameters
+
+The `Training` schema also supports the following optional fields (all are optional and have sensible defaults):
+
+| Field                | Description                                                               |
+| -------------------- | ------------------------------------------------------------------------- |
+| `output_path`        | S3 URI where training artifacts should be stored                          |
+| `max_run`            | Maximum training job runtime in seconds                                   |
+| `volume_size_in_gb`  | Size of the EBS volume attached to the training instance                  |
+| `spot`               | `true/false` – use SageMaker Spot Training to save costs                  |
+| `checkpoint_s3_uri`  | S3 URI for incremental checkpoints (only relevant if `spot: true`)        |
+
+*(See the `Training` Pydantic model in `magemaker/schemas/training.py` for the full list.)*
+
 ## Data Preparation
 
 ### Supported Formats
 
 <CardGroup>
   <Card title="CSV Format" icon="file-csv">
-    - Simple tabular data
-    - Easy to prepare
+    - Simple tabular data<br />
+    - Easy to prepare<br />
     - Good for classification tasks
   </Card>
 
   <Card title="JSON Lines" icon="file-code">
-    - Flexible data format
-    - Good for complex inputs
+    - Flexible data format<br />
+    - Good for complex inputs<br />
     - Supports nested structures
   </Card>
 </CardGroup>
@@ -71,48 +95,43 @@ training: !Training
 
 <Steps>
   <Step title="Prepare Data">
-    Format your data according to model requirements
+    Format your dataset according to the model requirements (e.g., one JSON line per training example).
   </Step>
   <Step title="Upload to S3">
-    Use AWS CLI or console to upload data
+    Use the AWS CLI or console to upload your dataset:
+    ```bash
+    aws s3 cp local/path/to/data.csv s3://your-bucket/data.csv
+    ```
   </Step>
   <Step title="Configure Path">
-    Specify S3 path in training configuration
+    Reference the S3 URI in `training_input_path` of your YAML file.
   </Step>
 </Steps>
 
 ## Instance Selection
 
-### Training Instance Types
-
-Choose based on:
-- Dataset size
-- Model size
-- Training time requirements
-- Cost constraints
+Choosing the right training instance impacts both training time and cost.
 
 Popular choices:
-- ml.p3.2xlarge (1 GPU)
-- ml.p3.8xlarge (4 GPUs)
-- ml.p3.16xlarge (8 GPUs)
 
-## Hyperparameter Tuning
+| Instance            | GPUs | Typical use-case                              |
+| ------------------- | ---- | -------------------------------------------- |
+| `ml.p3.2xlarge`     | 1    | Small to medium models (<7B parameters)       |
+| `ml.p3.8xlarge`     | 4    | Larger models / shorter turnaround            |
+| `ml.p3.16xlarge`    | 8    | Large-scale training / distributed workloads  |
 
-### Basic Parameters
+<Warning>
+Always check your current SageMaker GPU quota and request increases if necessary.
+</Warning>
 
-```yaml
-hyperparameters: !Hyperparameters
-  epochs: 3
-  learning_rate: 2e-5
-  batch_size: 32
-```
+## Hyperparameter Tuning
 
-### Advanced Tuning
+While you can hard-code values, you can also pass ranges to let SageMaker perform hyperparameter tuning (HPO). Use the `min`, `max`, or `values` keys as shown below:
 
 ```yaml
 hyperparameters: !Hyperparameters
   epochs: 3
-  learning_rate: 
+  learning_rate:
     min: 1e-5
     max: 1e-4
     scaling: log
@@ -122,9 +141,27 @@ hyperparameters: !Hyperparameters
 
 ## Monitoring Training
 
-### CloudWatch Metrics
+Magemaker streams CloudWatch metrics for every training job. Key metrics include:
+
+- `Train/Loss`
+- `Eval/Loss`
+- `LearningRate`
+- `GPUUtilization`
+
+You can access logs directly in the SageMaker console or via the AWS CLI:
 
-Available metrics:
-- Loss
-- Learning rate
-- GPU utilization
+```bash
+aws logs tail /aws/sagemaker/TrainingJobs --follow --since 1h
+```
+
+<Note>
+Job status (Started, InProgress, Completed, Failed) is also surfaced in the CLI output.
+</Note>
+
+## Cleaning Up
+
+Training jobs store artifacts (model checkpoints, logs) in S3. Delete these objects when no longer needed to avoid storage costs:
+
+```bash
+aws s3 rm --recursive s3://your-bucket/<training-job-name>
+```

concepts/jumpstart-custom-models.mdx

@@ -0,0 +1,88 @@
+---
+title: JumpStart & Custom Models
+description: Deploy AWS JumpStart marketplace and custom models with Magemaker
+---
+
+## Overview
+Besides Hugging Face models, Magemaker can now deploy two additional model types **on AWS SageMaker**:
+
+1. **JumpStart models** – pre-packaged models provided by Amazon or third-party sellers.
+2. **Custom models** – your own fine-tuned artefacts (.tar.gz or directory) stored locally or in S3.
+
+<GithubContribTag />
+
+---
+## 1 · Deploying a JumpStart Model
+
+### 1.1 Find the *model_id*
+Browse the [JumpStart Model Zoo](https://docs.aws.amazon.com/sagemaker/latest/dg/jumpstart-model-zoo.html) or call:
+```python
+from sagemaker.jumpstart.notebook_utils import list_jumpstart_models
+list_jumpstart_models()
+```
+
+### 1.2 Create a YAML
+```yaml
+models:
+- !Model
+  id: huggingface-text2text-flan-t5-large  # JumpStart model_id
+  source: sagemaker
+
+deployment: !Deployment
+  destination: aws
+  instance_type: ml.g5.2xlarge
+  instance_count: 1
+```
+
+### 1.3 Deploy
+```bash
+magemaker --deploy .magemaker_config/flan-t5.yaml
+```
+Magemaker handles the EULA acceptance (`accept_eula=True`) automatically.
+
+---
+## 2 · Deploying a Custom Model
+Custom models are useful when you have already trained a model locally or with SageMaker Training.
+
+### 2.1 Package Artifacts
+- For **Hugging Face** style, create a directory containing `config.json`, `tokenizer.json`, etc.  
+- Optionally compress to `model.tar.gz`.
+
+### 2.2 Upload (optional)
+If your artefact is not yet on S3, Magemaker will upload it for you.
+
+### 2.3 YAML Example
+```yaml
+models:
+- !Model
+  id: my-distilbert-finetuned
+  source: huggingface
+  location: ./artifacts/distilbert.tar.gz  # or s3://bucket/key
+
+deployment: !Deployment
+  destination: aws
+  instance_type: ml.m5.xlarge
+```
+
+### 2.4 Deploy
+```bash
+magemaker --deploy .magemaker_config/my-distilbert.yaml
+```
+Magemaker will:
+1. Upload the local artefact to `s3://<default-bucket>/models/my-distilbert-finetuned/` (if needed)
+2. Create a `HuggingFaceModel` pointing to that S3 path
+3. Spin up the endpoint.
+
+---
+## Tips & Quotas
+- **GPU quota** – JumpStart LLMs often need a GPU instance (`ml.g5.*`). Request a quota increase first.
+- **Endpoint names** – If `endpoint_name` is omitted, Magemaker appends a timestamp to ensure uniqueness.
+- **Cost control** – Delete endpoints when not in use: `magemaker --cloud aws → Delete a Model Endpoint`.
+
+---
+## FAQ
+**Q : Do JumpStart models work on GCP / Azure?**  
+*A : Not yet. JumpStart is an AWS-specific marketplace.*
+
+**Q : Can I pass inference parameters to a JumpStart model?**  
+*A : Yes. Use the same `predict:` block as with Hugging Face models.*