Update README.md
Browse files
README.md
CHANGED
|
@@ -1,5 +1,60 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: other
|
| 3 |
-
license_name: lfm1.0
|
| 4 |
-
license_link: https://huggingface.co/LiquidAI/LFM2-350M/blob/main/LICENSE
|
| 5 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: other
|
| 3 |
+
license_name: lfm1.0
|
| 4 |
+
license_link: https://huggingface.co/LiquidAI/LFM2-350M/blob/main/LICENSE
|
| 5 |
+
datasets:
|
| 6 |
+
- Archi-medes/LabGuide_Preview
|
| 7 |
+
base_model:
|
| 8 |
+
- LiquidAI/LFM2-1.2B
|
| 9 |
+
pipeline_tag: question-answering
|
| 10 |
+
---
|
| 11 |
+
# LabGuide Preview Model
|
| 12 |
+
|
| 13 |
+
## Model Summary
|
| 14 |
+
The **LabGuide Preview Model** is a demonstration release built entirely with **Madlab**, using its synthetic dataset generator and training workflow.
|
| 15 |
+
It is based on [LiquidAI/LFM2-1.2B](https://huggingface.co/LiquidAI/LFM2-1.2B), adapted to showcase Madlab’s end-to-end capabilities for dataset creation, model training, and assistant deployment.
|
| 16 |
+
|
| 17 |
+
This model illustrates how applications can leverage Madlab to train their own assistants in a reproducible and accessible way.
|
| 18 |
+
It is **not intended for production use**, but rather as a preview for contributors, collaborators, and community feedback.
|
| 19 |
+
|
| 20 |
+
---
|
| 21 |
+
|
| 22 |
+
## Training Data
|
| 23 |
+
- **Source**: Synthetic dataset generated entirely with Madlab’s dataset generator.
|
| 24 |
+
- **Purpose**: Designed to demonstrate Madlab’s ability to produce structured, reproducible training data.
|
| 25 |
+
- **Scope**: Preview-scale dataset, not representative of real-world or production-ready corpora.
|
| 26 |
+
|
| 27 |
+
---
|
| 28 |
+
|
| 29 |
+
## Training Process
|
| 30 |
+
- **Framework**: Madlab training pipeline.
|
| 31 |
+
- **Base Model**: [LiquidAI/LFM2-1.2B](https://huggingface.co/LiquidAI/LFM2-1.2B).
|
| 32 |
+
- **Workflow**: Synthetic dataset generation → Madlab training loop → Magic Judge Evaluation → Preview model release.
|
| 33 |
+
- **Objective**: Demonstrate Madlab’s integrated workflow for building application-specific assistants.
|
| 34 |
+
|
| 35 |
+
---
|
| 36 |
+
|
| 37 |
+
## Intended Uses
|
| 38 |
+
- Contributor onboarding and workflow validation.
|
| 39 |
+
- Demonstration of Madlab’s synthetic dataset generator and training pipeline.
|
| 40 |
+
- Benchmarking and experimentation in controlled preview settings.
|
| 41 |
+
|
| 42 |
+
---
|
| 43 |
+
|
| 44 |
+
## Limitations
|
| 45 |
+
- **Demo-only**: Not suitable for production or deployment in real-world applications.
|
| 46 |
+
- **Synthetic data**: Training data is fully synthetic and may not reflect natural language distributions.
|
| 47 |
+
- **Preview scale**: Model performance is illustrative, not optimized for accuracy or robustness.
|
| 48 |
+
|
| 49 |
+
---
|
| 50 |
+
|
| 51 |
+
## Ethical Considerations
|
| 52 |
+
- This model is provided for demonstration and educational purposes.
|
| 53 |
+
- It should not be used in applications where accuracy, safety, or reliability are critical.
|
| 54 |
+
- Contributors are encouraged to treat outputs as illustrative examples only.
|
| 55 |
+
|
| 56 |
+
---
|
| 57 |
+
|
| 58 |
+
## Acknowledgements
|
| 59 |
+
- Base model: [LiquidAI/LFM2-1.2B](https://huggingface.co/LiquidAI/LFM2-1.2B).
|
| 60 |
+
- Built and trained with [**Madlab**](https://github.com/archimedes1618/madlab).
|