---
library_name: peft
license: mit
base_model: microsoft/Phi-3.5-mini-instruct
tags:
- axolotl
- generated_from_trainer
model-index:
- name: curator_math_phase1_sn_ensemble7_90325
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

[<img src="https://raw.githubusercontent.com/axolotl-ai-cloud/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/axolotl-ai-cloud/axolotl)
<details><summary>See axolotl config</summary>

axolotl version: `0.5.0`


</details><br>

# Collinear Curator 1:

This is an open-source fine-tuned reasoning adapter of [microsoft/Phi-3.5-mini-instruct](https://huggingface.co/microsoft/Phi-3.5-mini-instruct), transformed into a math reasoning model using data curated from [collinear-ai/R1-Distill-SFT-Curated](https://huggingface.co/datasets/collinear-ai/R1-Distill-SFT-Curated).
It achieves the following results on the evaluation set:
- Loss: 0.3203

## Model description

This model is a LoRA adaptor and for best results merge it with base model [microsoft/Phi-3.5-mini-instruct](https://huggingface.co/microsoft/Phi-3.5-mini-instruct) before use. 


## Training and evaluation data

- Training data: [collinear-ai/R1-Distill-SFT-Curated](https://huggingface.co/datasets/collinear-ai/R1-Distill-SFT-Curated)
- Evaluation data: [HuggingFaceH4/MATH-500](https://huggingface.co/datasets/HuggingFaceH4/MATH-500)


## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 5e-06
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 2
- total_train_batch_size: 128
- total_eval_batch_size: 64
- optimizer: Use OptimizerNames.PAGED_ADAMW_8BIT with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 50
- num_epochs: 1

### Training results

| Training Loss | Epoch  | Step | Validation Loss |
|:-------------:|:------:|:----:|:---------------:|
| No log        | 0.0003 | 1    | 0.6714          |
| 0.337         | 0.3335 | 1243 | 0.3361          |
| 0.3248        | 0.6669 | 2486 | 0.3203          |


### Evaluation Results on Math500
The following figure shows the accuracy and the speedup of Collinear Curators C1 and C2 when compared to training on unfiltered dataset.
![Math Reasoning Evaluation](https://huggingface.co/collinear-ai/math_reasoning_phi_c1/raw/main/math500_eval_c1_c2.png)


### Framework versions

- PEFT 0.13.2
- Transformers 4.46.1
- Pytorch 2.3.1+cu121
- Datasets 3.0.1
- Tokenizers 0.20.3