|
--- |
|
library_name: peft |
|
tags: |
|
- trl |
|
- sft |
|
- unsloth |
|
- generated_from_trainer |
|
datasets: |
|
- generator |
|
base_model: unsloth/tinyllama-chat-bnb-4bit |
|
model-index: |
|
- name: outputs |
|
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. --> |
|
|
|
# outputs |
|
|
|
This model is a fine-tuned version of [unsloth/tinyllama-chat-bnb-4bit](https://huggingface.co/unsloth/tinyllama-chat-bnb-4bit) on the generator dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.4551 |
|
|
|
## 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: 2e-05 |
|
- train_batch_size: 2 |
|
- eval_batch_size: 2 |
|
- seed: 3407 |
|
- gradient_accumulation_steps: 4 |
|
- total_train_batch_size: 8 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_ratio: 0.1 |
|
- training_steps: 100 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 1.9638 | 0.0 | 1 | 1.9504 | |
|
| 1.9792 | 0.0 | 2 | 1.9467 | |
|
| 1.9213 | 0.0 | 3 | 1.9308 | |
|
| 2.0225 | 0.0 | 4 | 1.8936 | |
|
| 1.8364 | 0.0 | 5 | 1.8278 | |
|
| 1.7729 | 0.0 | 6 | 1.7366 | |
|
| 1.9591 | 0.01 | 7 | 1.6324 | |
|
| 1.6693 | 0.01 | 8 | 1.5278 | |
|
| 1.6387 | 0.01 | 9 | 1.4367 | |
|
| 1.5681 | 0.01 | 10 | 1.3741 | |
|
| 1.3459 | 0.01 | 11 | 1.3300 | |
|
| 1.311 | 0.01 | 12 | 1.2931 | |
|
| 1.2721 | 0.01 | 13 | 1.2534 | |
|
| 1.353 | 0.01 | 14 | 1.2140 | |
|
| 1.1664 | 0.01 | 15 | 1.1727 | |
|
| 1.27 | 0.01 | 16 | 1.1344 | |
|
| 1.1007 | 0.01 | 17 | 1.0966 | |
|
| 1.1035 | 0.01 | 18 | 1.0608 | |
|
| 1.0744 | 0.01 | 19 | 1.0278 | |
|
| 1.0491 | 0.02 | 20 | 0.9973 | |
|
| 1.0057 | 0.02 | 21 | 0.9688 | |
|
| 0.9435 | 0.02 | 22 | 0.9423 | |
|
| 0.9612 | 0.02 | 23 | 0.9169 | |
|
| 0.9811 | 0.02 | 24 | 0.8932 | |
|
| 0.9263 | 0.02 | 25 | 0.8700 | |
|
| 0.8581 | 0.02 | 26 | 0.8468 | |
|
| 0.8351 | 0.02 | 27 | 0.8237 | |
|
| 0.8019 | 0.02 | 28 | 0.8008 | |
|
| 0.8526 | 0.02 | 29 | 0.7786 | |
|
| 0.773 | 0.02 | 30 | 0.7571 | |
|
| 0.7436 | 0.02 | 31 | 0.7365 | |
|
| 0.7455 | 0.03 | 32 | 0.7172 | |
|
| 0.747 | 0.03 | 33 | 0.6995 | |
|
| 0.727 | 0.03 | 34 | 0.6834 | |
|
| 0.6859 | 0.03 | 35 | 0.6687 | |
|
| 0.6642 | 0.03 | 36 | 0.6552 | |
|
| 0.6715 | 0.03 | 37 | 0.6428 | |
|
| 0.6538 | 0.03 | 38 | 0.6311 | |
|
| 0.5947 | 0.03 | 39 | 0.6202 | |
|
| 0.6537 | 0.03 | 40 | 0.6102 | |
|
| 0.601 | 0.03 | 41 | 0.6008 | |
|
| 0.5956 | 0.03 | 42 | 0.5921 | |
|
| 0.5875 | 0.03 | 43 | 0.5842 | |
|
| 0.5737 | 0.03 | 44 | 0.5769 | |
|
| 0.5618 | 0.04 | 45 | 0.5701 | |
|
| 0.546 | 0.04 | 46 | 0.5638 | |
|
| 0.5908 | 0.04 | 47 | 0.5578 | |
|
| 0.6172 | 0.04 | 48 | 0.5520 | |
|
| 0.5652 | 0.04 | 49 | 0.5467 | |
|
| 0.5357 | 0.04 | 50 | 0.5417 | |
|
| 0.5524 | 0.04 | 51 | 0.5370 | |
|
| 0.5352 | 0.04 | 52 | 0.5326 | |
|
| 0.5356 | 0.04 | 53 | 0.5283 | |
|
| 0.518 | 0.04 | 54 | 0.5242 | |
|
| 0.5273 | 0.04 | 55 | 0.5201 | |
|
| 0.5099 | 0.04 | 56 | 0.5161 | |
|
| 0.5158 | 0.04 | 57 | 0.5123 | |
|
| 0.521 | 0.05 | 58 | 0.5084 | |
|
| 0.5177 | 0.05 | 59 | 0.5047 | |
|
| 0.4964 | 0.05 | 60 | 0.5010 | |
|
| 0.502 | 0.05 | 61 | 0.4974 | |
|
| 0.5078 | 0.05 | 62 | 0.4942 | |
|
| 0.4814 | 0.05 | 63 | 0.4913 | |
|
| 0.4863 | 0.05 | 64 | 0.4887 | |
|
| 0.4998 | 0.05 | 65 | 0.4864 | |
|
| 0.5106 | 0.05 | 66 | 0.4842 | |
|
| 0.5273 | 0.05 | 67 | 0.4822 | |
|
| 0.4874 | 0.05 | 68 | 0.4803 | |
|
| 0.4697 | 0.05 | 69 | 0.4785 | |
|
| 0.4796 | 0.05 | 70 | 0.4768 | |
|
| 0.4767 | 0.06 | 71 | 0.4753 | |
|
| 0.4582 | 0.06 | 72 | 0.4739 | |
|
| 0.5084 | 0.06 | 73 | 0.4725 | |
|
| 0.4566 | 0.06 | 74 | 0.4712 | |
|
| 0.4583 | 0.06 | 75 | 0.4700 | |
|
| 0.4753 | 0.06 | 76 | 0.4689 | |
|
| 0.4528 | 0.06 | 77 | 0.4678 | |
|
| 0.4617 | 0.06 | 78 | 0.4667 | |
|
| 0.499 | 0.06 | 79 | 0.4656 | |
|
| 0.4368 | 0.06 | 80 | 0.4646 | |
|
| 0.4939 | 0.06 | 81 | 0.4637 | |
|
| 0.4446 | 0.06 | 82 | 0.4627 | |
|
| 0.4428 | 0.07 | 83 | 0.4618 | |
|
| 0.4737 | 0.07 | 84 | 0.4611 | |
|
| 0.4391 | 0.07 | 85 | 0.4603 | |
|
| 0.4985 | 0.07 | 86 | 0.4597 | |
|
| 0.45 | 0.07 | 87 | 0.4590 | |
|
| 0.4642 | 0.07 | 88 | 0.4585 | |
|
| 0.4633 | 0.07 | 89 | 0.4579 | |
|
| 0.4233 | 0.07 | 90 | 0.4574 | |
|
| 0.4478 | 0.07 | 91 | 0.4570 | |
|
| 0.4768 | 0.07 | 92 | 0.4565 | |
|
| 0.4665 | 0.07 | 93 | 0.4562 | |
|
| 0.4504 | 0.07 | 94 | 0.4560 | |
|
| 0.4692 | 0.07 | 95 | 0.4557 | |
|
| 0.4326 | 0.08 | 96 | 0.4555 | |
|
| 0.4727 | 0.08 | 97 | 0.4554 | |
|
| 0.4658 | 0.08 | 98 | 0.4553 | |
|
| 0.4905 | 0.08 | 99 | 0.4552 | |
|
| 0.4501 | 0.08 | 100 | 0.4551 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.8.2 |
|
- Transformers 4.37.2 |
|
- Pytorch 2.1.0+cu121 |
|
- Datasets 2.17.1 |
|
- Tokenizers 0.15.2 |
