Image-Text-to-Text
PEFT
Safetensors
English
gemma4
lora
adapter
math
code
reasoning
step-by-step
verified-data
autoscientist
adaption
negative-result
conversational
Eval Results (legacy)
Instructions to use manifesta/adaption_verified_math_code_instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use manifesta/adaption_verified_math_code_instruct with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("google/gemma-4-31B-it") model = PeftModel.from_pretrained(base_model, "manifesta/adaption_verified_math_code_instruct") - Notebooks
- Google Colab
- Kaggle
Link The Verdict write-up
Browse files
README.md
CHANGED
|
@@ -51,6 +51,7 @@ model-index:
|
|
| 51 |
</p>
|
| 52 |
|
| 53 |
[](https://manifesta.adaptionlabs.app/)
|
|
|
|
| 54 |
[](https://huggingface.co/google/gemma-4-31B-it)
|
| 55 |
[](https://huggingface.co/datasets/manifesta/verified-math-code-17k)
|
| 56 |
[](#evaluation)
|
|
@@ -338,6 +339,9 @@ That +2.5% is the smallest lift of the four datasets built in this project, and
|
|
| 338 |
|
| 339 |
## Live interface
|
| 340 |
|
|
|
|
|
|
|
|
|
|
| 341 |
**Show Your Work** runs at [manifesta.adaptionlabs.app](https://manifesta.adaptionlabs.app/). Give it a problem, get a step by step derivation with a verification panel, and use the "Change a number" button to check that the derivation actually recomputes rather than pattern matching to a memorised answer.
|
| 342 |
|
| 343 |
<p align="center">
|
|
|
|
| 51 |
</p>
|
| 52 |
|
| 53 |
[](https://manifesta.adaptionlabs.app/)
|
| 54 |
+
[](https://demo-theta-one-40.vercel.app)
|
| 55 |
[](https://huggingface.co/google/gemma-4-31B-it)
|
| 56 |
[](https://huggingface.co/datasets/manifesta/verified-math-code-17k)
|
| 57 |
[](#evaluation)
|
|
|
|
| 339 |
|
| 340 |
## Live interface
|
| 341 |
|
| 342 |
+
The full written case for this entry, every claim paired with the command that checks it, is at
|
| 343 |
+
[The Verdict](https://demo-theta-one-40.vercel.app).
|
| 344 |
+
|
| 345 |
**Show Your Work** runs at [manifesta.adaptionlabs.app](https://manifesta.adaptionlabs.app/). Give it a problem, get a step by step derivation with a verification panel, and use the "Change a number" button to check that the derivation actually recomputes rather than pattern matching to a memorised answer.
|
| 346 |
|
| 347 |
<p align="center">
|