CourseGPT-Pro-DSAI-Lab-Group-6
/

router-gemma3-peft

@@ -1,208 +1,92 @@
 ---
 base_model: google/gemma-3-27b-it
 library_name: peft
-pipeline_tag: text-generation
 tags:
-- axolotl
-- base_model:adapter:/tmp/base_model_dir/vertex-model-garden-restricted-us/gemma3/gemma-3-27b-it
 - lora
-- transformers
 ---
-# Model Card for Model ID
-<!-- Provide a quick summary of what the model is/does. -->
 ## Model Details
-### Model Description
-<!-- Provide a longer summary of what this model is. -->
-- **Developed by:** [More Information Needed]
-- **Funded by [optional]:** [More Information Needed]
-- **Shared by [optional]:** [More Information Needed]
-- **Model type:** [More Information Needed]
-- **Language(s) (NLP):** [More Information Needed]
-- **License:** [More Information Needed]
-- **Finetuned from model [optional]:** [More Information Needed]
-### Model Sources [optional]
-<!-- Provide the basic links for the model. -->
-- **Repository:** [More Information Needed]
-- **Paper [optional]:** [More Information Needed]
-- **Demo [optional]:** [More Information Needed]
-## Uses
-<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
-### Direct Use
-<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
-[More Information Needed]
-### Downstream Use [optional]
-<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
-[More Information Needed]
-### Out-of-Scope Use
-<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
-[More Information Needed]
-## Bias, Risks, and Limitations
-<!-- This section is meant to convey both technical and sociotechnical limitations. -->
-[More Information Needed]
-### Recommendations
-<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
-Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
-## How to Get Started with the Model
-Use the code below to get started with the model.
-[More Information Needed]
-## Training Details
-### Training Data
-<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
-[More Information Needed]
-### Training Procedure
-<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
-#### Preprocessing [optional]
-[More Information Needed]
-#### Training Hyperparameters
-- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
-#### Speeds, Sizes, Times [optional]
-<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
-[More Information Needed]
-## Evaluation
-<!-- This section describes the evaluation protocols and provides the results. -->
-### Testing Data, Factors & Metrics
-#### Testing Data
-<!-- This should link to a Dataset Card if possible. -->
-[More Information Needed]
-#### Factors
-<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
-[More Information Needed]
-#### Metrics
-<!-- These are the evaluation metrics being used, ideally with a description of why. -->
-[More Information Needed]
-### Results
-[More Information Needed]
-#### Summary
-## Model Examination [optional]
-<!-- Relevant interpretability work for the model goes here -->
-[More Information Needed]
-## Environmental Impact
-<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
-Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
-- **Hardware Type:** [More Information Needed]
-- **Hours used:** [More Information Needed]
-- **Cloud Provider:** [More Information Needed]
-- **Compute Region:** [More Information Needed]
-- **Carbon Emitted:** [More Information Needed]
-## Technical Specifications [optional]
-### Model Architecture and Objective
-[More Information Needed]
-### Compute Infrastructure
-[More Information Needed]
-#### Hardware
-[More Information Needed]
-#### Software
-[More Information Needed]
-## Citation [optional]
-<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
-**BibTeX:**
-[More Information Needed]
-**APA:**
-[More Information Needed]
-## Glossary [optional]
-<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
-[More Information Needed]
-## More Information [optional]
-[More Information Needed]
-## Model Card Authors [optional]
-[More Information Needed]
-## Model Card Contact
-[More Information Needed]
-### Framework versions
-- PEFT 0.17.1

 ---
 base_model: google/gemma-3-27b-it
+language:
+- en
 library_name: peft
+license: gemma
 tags:
 - lora
+- peft
+- router-agent
+- vertex-ai
+datasets:
+- CourseGPT-Pro-DSAI-Lab-Group-6/router-dataset
+metrics:
+- name: eval_loss
+  type: loss
+  value: 0.6080
+- name: eval_perplexity
+  type: perplexity
+  value: 1.8368
 ---
+# Router Gemma 3 27B PEFT Adapter
+This repository hosts the LoRA adapter for **google/gemma-3-27b-it**, tuned as a tool-routing brain with strong reasoning headroom. The model reads user queries, decides which agents (e.g., `/math`, `/code`, `/general-search`) should run, and emits strict JSON aligned with our Milestone 2 router schema.
 ## Model Details
+- **Base model:** [google/gemma-3-27b-it](https://huggingface.co/google/gemma-3-27b-it)
+- **Adapter type:** QLoRA rank 16 on attention + MLP projections (Vertex AI managed tuning)
+- **Training:** 3 epochs, micro-batch size 2, cosine LR with warmup, gradient checkpointing enabled
+- **Hardware:** NVIDIA H100/A3 (Vertex managed OSS tuning)
+- **Context length:** 128K tokens
+- **Validation metrics:** loss ≈ 0.6080, perplexity ≈ 1.84, eval runtime ≈ 15.4 s
+Gemma’s larger capacity gives higher-quality routing decisions, especially for multi-step orchestration across math/code/search specialists.
+## Intended Use
+Use this adapter when you need premium routing quality and can afford deploying on higher-memory GPUs (L4 with quantization or A100/H100 in full precision). It is well-suited for research copilots, analytics assistants, and multilingual routing scenarios.
+### Out-of-scope
+- Direct Q/A without tool execution
+- High-risk/sensitive domains without additional alignment checks
+## Quick Start
+```python
+from peft import PeftModel
+from transformers import AutoTokenizer, AutoModelForCausalLM
+base = "google/gemma-3-27b-it"
+adapter = "CourseGPT-Pro-DSAI-Lab-Group-6/router-gemma3-peft"
+tokenizer = AutoTokenizer.from_pretrained(base, trust_remote_code=True)
+model = AutoModelForCausalLM.from_pretrained(
+    base,
+    device_map="auto",
+    torch_dtype="auto",
+    trust_remote_code=True,
+)
+model = PeftModel.from_pretrained(model, adapter)
+prompt = (
+    "System: Emit strict JSON with route_plan, route_rationale, thinking_outline, and handoff_plan.\n"
+    "User: Draft a pipeline combining symbolic integration, Python experimentation, and literature review."
+)
+inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
+output = model.generate(**inputs, max_new_tokens=1200)
+print(tokenizer.decode(output[0], skip_special_tokens=True))
+```
+## Training Data
+- CourseGPT router dataset (Milestone 2), converted to Vertex supervised JSONL (prompt/completion pairs)
+- Structured completions include route plan arrays, rationale, acceptance criteria, metrics, and more
+## Evaluation Summary
+- Held-out validation subset (≈10%)
+- Metrics: loss ≈ 0.6080, perplexity ≈ 1.84
+- Qualitative review shows consistent JSON structure and accurate tool choices on complex problems
+## Deployment Tips
+- Quantize adapters for L4 deployment (bitsandbytes 4-bit)
+- Validate JSON outputs and retry when necessary
+- Extend the prompt with custom tool definitions if your stack differs from `/math`, `/code`, `/general-search`
+## Citation
+```
+@software{CourseGPTRouterGemma3,
+  title  = {Router Gemma 3 27B PEFT Adapter},
+  author = {CourseGPT Pro DSAI Lab Group 6},
+  year   = {2025},
+  url    = {https://huggingface.co/CourseGPT-Pro-DSAI-Lab-Group-6/router-gemma3-peft}
+}
+```