File size: 1,753 Bytes
e952c6c ebb0bda e952c6c ebb0bda e952c6c ebb0bda e952c6c ebb0bda e952c6c ebb0bda e952c6c |
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 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 |
---
base_model: unsloth/llama-3.2-1b-instruct-bnb-4bit
library_name: peft
license: llama3.2
tags:
- trl
- sft
- unsloth
- generated_from_trainer
model-index:
- name: Llama-3.2-1B-Instruct
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. -->
# Llama-3.2-1B-Instruct
This model is a fine-tuned version of [unsloth/llama-3.2-1b-instruct-bnb-4bit](https://huggingface.co/unsloth/llama-3.2-1b-instruct-bnb-4bit) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.7877
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0002
- train_batch_size: 5
- eval_batch_size: 8
- seed: 3407
- gradient_accumulation_steps: 4
- total_train_batch_size: 20
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- num_epochs: 1
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:------:|:----:|:---------------:|
| 1.7636 | 0.1992 | 50 | 1.8301 |
| 1.6351 | 0.3984 | 100 | 1.8318 |
| 1.6559 | 0.5976 | 150 | 1.8157 |
| 1.6928 | 0.7968 | 200 | 1.7976 |
| 1.7535 | 0.9960 | 250 | 1.7877 |
### Framework versions
- PEFT 0.13.2
- Transformers 4.44.2
- Pytorch 2.4.1+cu121
- Datasets 3.0.1
- Tokenizers 0.19.1 |