|
|
--- |
|
|
base_model: meta-llama/Llama-3.2-1B-Instruct |
|
|
library_name: peft |
|
|
license: llama3.2 |
|
|
tags: |
|
|
- trl |
|
|
- sft |
|
|
- generated_from_trainer |
|
|
- lora |
|
|
model-index: |
|
|
- name: llama-3.2-1B-it-Procurtech-Assistant |
|
|
results: [] |
|
|
datasets: |
|
|
- Victorano/procurtech-assistant-training-dataset |
|
|
language: |
|
|
- en |
|
|
pipeline_tag: text2text-generation |
|
|
--- |
|
|
|
|
|
|
|
|
# llama-3.2-1B-it-Procurtech-Assistant |
|
|
|
|
|
This model is a fine-tuned version of [meta-llama/Llama-3.2-1B-Instruct](https://huggingface.co/meta-llama/Llama-3.2-1B-Instruct) on [Procurtech Assistant dataset](https://huggingface.co/datasets/Victorano/procurtech-assistant-training-dataset). |
|
|
|
|
|
## Model description |
|
|
|
|
|
A customer support model to help customers with their orders, incase they encounter any difficulty. |
|
|
|
|
|
## Intended uses & limitations |
|
|
|
|
|
The training dataset can be modified, see original at [customer support dataset](https://huggingface.co/bitext/Bitext-customer-support-llm-chatbot-training-dataset) .. I edited the system message with a bit of prompt engineering, included additional details about the eCommerce company. |
|
|
You can decide what you want and further fine tune the model... |
|
|
|
|
|
## Training and evaluation data |
|
|
|
|
|
[Training data](https://huggingface.co/datasets/Victorano/procurtech-assistant-training-dataset). |
|
|
Used the complete dataset for training, no evaluation data, I evaluated with random prompts... |
|
|
|
|
|
## Training procedure |
|
|
|
|
|
### Training hyperparameters |
|
|
|
|
|
The following hyperparameters were used during training: |
|
|
- learning_rate: 0.0002 |
|
|
- train_batch_size: 1 |
|
|
- eval_batch_size: 8 |
|
|
- seed: 42 |
|
|
- gradient_accumulation_steps: 2 |
|
|
- total_train_batch_size: 2 |
|
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
|
- lr_scheduler_type: cosine |
|
|
- lr_scheduler_warmup_steps: 682 |
|
|
- num_epochs: 1 |
|
|
|
|
|
### Training results |
|
|
[Training Loss from wandb](https://api.wandb.ai/links/victordareai/o4f88gmp) |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- PEFT 0.13.2 |
|
|
- Transformers 4.45.2 |
|
|
- Pytorch 2.4.1+cu121 |
|
|
- Datasets 3.0.1 |
|
|
- Tokenizers 0.20.1 |
