File size: 1,501 Bytes
d576ac4 a1c453a d576ac4 0b2bc4c d576ac4 0b2bc4c d576ac4 0b2bc4c d576ac4 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 |
---
language:
- en
license: apache-2.0
tags:
- text-generation-inference
- transformers
- unsloth
- mistral
- trl
- sft
base_model: unsloth/mistral-7b-bnb-4bit
datasets:
- m-a-p/Code-Feedback
- knowrohit07/saraswati-stem
---
# **StemCoder-7B Test**
_This is not a full, comprehensive finetune. Stem-Coder-7B was a test to finetune Mistral-7B on 9k instructions (albeit, many are multi-turn) via a free Google Colab instance with [unsloth](https://github.com/unslothai/unsloth)._
<img src="https://cdn-uploads.huggingface.co/production/uploads/648a374f00f7a3374ee64b99/FRI6tBRlPqJb_qncxtFxO.png" width="500">
This model is finetuned for:
- Math
- Multi-Turn Coding
- STEM / Science / Medicine
StemCoder gives really descriptive answers, particularly for science, math based questions.
# **Examples**
## **Science**
<img src="https://cdn-uploads.huggingface.co/production/uploads/648a374f00f7a3374ee64b99/UZQy1Otn7xmjMKx0UG1_4.png" width="800">
<img src="https://cdn-uploads.huggingface.co/production/uploads/648a374f00f7a3374ee64b99/Z1hiK-YrWt_A5q21DWL3v.png" width="800">
## **Coding**
<img src="https://cdn-uploads.huggingface.co/production/uploads/648a374f00f7a3374ee64b99/XttCe5JvpXgdwDn8xERWR.png" width="800">
<img src="https://cdn-uploads.huggingface.co/production/uploads/648a374f00f7a3374ee64b99/UL0I2q8YjrSooQsxUb27S.png" width="800">
## **Medicine**
<img src="https://cdn-uploads.huggingface.co/production/uploads/648a374f00f7a3374ee64b99/SfyvKBAUqba03ggHyxS52.png" width="800">
|