Model save
Browse files
README.md
ADDED
|
@@ -0,0 +1,117 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: other
|
| 3 |
+
base_model: Qwen/Qwen1.5-4B
|
| 4 |
+
tags:
|
| 5 |
+
- generated_from_trainer
|
| 6 |
+
metrics:
|
| 7 |
+
- accuracy
|
| 8 |
+
model-index:
|
| 9 |
+
- name: lmind_nq_train6000_eval6489_v1_qa_3e-4_lora2
|
| 10 |
+
results: []
|
| 11 |
+
library_name: peft
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 15 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 16 |
+
|
| 17 |
+
# lmind_nq_train6000_eval6489_v1_qa_3e-4_lora2
|
| 18 |
+
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen1.5-4B](https://huggingface.co/Qwen/Qwen1.5-4B) on an unknown dataset.
|
| 20 |
+
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 2.5245
|
| 22 |
+
- Accuracy: 0.5456
|
| 23 |
+
|
| 24 |
+
## Model description
|
| 25 |
+
|
| 26 |
+
More information needed
|
| 27 |
+
|
| 28 |
+
## Intended uses & limitations
|
| 29 |
+
|
| 30 |
+
More information needed
|
| 31 |
+
|
| 32 |
+
## Training and evaluation data
|
| 33 |
+
|
| 34 |
+
More information needed
|
| 35 |
+
|
| 36 |
+
## Training procedure
|
| 37 |
+
|
| 38 |
+
### Training hyperparameters
|
| 39 |
+
|
| 40 |
+
The following hyperparameters were used during training:
|
| 41 |
+
- learning_rate: 0.0003
|
| 42 |
+
- train_batch_size: 1
|
| 43 |
+
- eval_batch_size: 2
|
| 44 |
+
- seed: 42
|
| 45 |
+
- distributed_type: multi-GPU
|
| 46 |
+
- num_devices: 4
|
| 47 |
+
- gradient_accumulation_steps: 8
|
| 48 |
+
- total_train_batch_size: 32
|
| 49 |
+
- total_eval_batch_size: 8
|
| 50 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 51 |
+
- lr_scheduler_type: constant
|
| 52 |
+
- lr_scheduler_warmup_ratio: 0.05
|
| 53 |
+
- num_epochs: 50.0
|
| 54 |
+
|
| 55 |
+
### Training results
|
| 56 |
+
|
| 57 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 58 |
+
|:-------------:|:-------:|:----:|:---------------:|:--------:|
|
| 59 |
+
| 1.7259 | 0.9973 | 187 | 1.6105 | 0.5753 |
|
| 60 |
+
| 1.3358 | 2.0 | 375 | 1.6465 | 0.5723 |
|
| 61 |
+
| 0.9854 | 2.9973 | 562 | 1.7621 | 0.5708 |
|
| 62 |
+
| 0.7508 | 4.0 | 750 | 1.9381 | 0.5663 |
|
| 63 |
+
| 0.6396 | 4.9973 | 937 | 1.9926 | 0.5634 |
|
| 64 |
+
| 0.5833 | 6.0 | 1125 | 2.1015 | 0.5612 |
|
| 65 |
+
| 0.5567 | 6.9973 | 1312 | 2.1645 | 0.5607 |
|
| 66 |
+
| 0.5411 | 8.0 | 1500 | 2.2040 | 0.5614 |
|
| 67 |
+
| 0.5028 | 8.9973 | 1687 | 2.2365 | 0.5608 |
|
| 68 |
+
| 0.5041 | 10.0 | 1875 | 2.2862 | 0.5605 |
|
| 69 |
+
| 0.4991 | 10.9973 | 2062 | 2.2851 | 0.5603 |
|
| 70 |
+
| 0.5048 | 12.0 | 2250 | 2.2455 | 0.5610 |
|
| 71 |
+
| 0.5067 | 12.9973 | 2437 | 2.2589 | 0.5592 |
|
| 72 |
+
| 0.508 | 14.0 | 2625 | 2.2631 | 0.5584 |
|
| 73 |
+
| 0.514 | 14.9973 | 2812 | 2.2773 | 0.5564 |
|
| 74 |
+
| 0.5149 | 16.0 | 3000 | 2.2861 | 0.5576 |
|
| 75 |
+
| 0.4835 | 16.9973 | 3187 | 2.2663 | 0.5588 |
|
| 76 |
+
| 0.484 | 18.0 | 3375 | 2.3145 | 0.5575 |
|
| 77 |
+
| 0.4862 | 18.9973 | 3562 | 2.2949 | 0.5559 |
|
| 78 |
+
| 0.4871 | 20.0 | 3750 | 2.3217 | 0.5581 |
|
| 79 |
+
| 0.4902 | 20.9973 | 3937 | 2.3256 | 0.5572 |
|
| 80 |
+
| 0.492 | 22.0 | 4125 | 2.3584 | 0.5558 |
|
| 81 |
+
| 0.4937 | 22.9973 | 4312 | 2.3608 | 0.5558 |
|
| 82 |
+
| 0.492 | 24.0 | 4500 | 2.3685 | 0.5552 |
|
| 83 |
+
| 0.4728 | 24.9973 | 4687 | 2.3752 | 0.5543 |
|
| 84 |
+
| 0.4753 | 26.0 | 4875 | 2.3276 | 0.5557 |
|
| 85 |
+
| 0.4788 | 26.9973 | 5062 | 2.4252 | 0.5542 |
|
| 86 |
+
| 0.4812 | 28.0 | 5250 | 2.3812 | 0.5551 |
|
| 87 |
+
| 0.4849 | 28.9973 | 5437 | 2.4413 | 0.5523 |
|
| 88 |
+
| 0.4872 | 30.0 | 5625 | 2.3946 | 0.5526 |
|
| 89 |
+
| 0.488 | 30.9973 | 5812 | 2.3911 | 0.5526 |
|
| 90 |
+
| 0.4864 | 32.0 | 6000 | 2.4076 | 0.5517 |
|
| 91 |
+
| 0.4667 | 32.9973 | 6187 | 2.4808 | 0.5505 |
|
| 92 |
+
| 0.4694 | 34.0 | 6375 | 2.4784 | 0.5523 |
|
| 93 |
+
| 0.4703 | 34.9973 | 6562 | 2.4760 | 0.5521 |
|
| 94 |
+
| 0.4704 | 36.0 | 6750 | 2.5062 | 0.5519 |
|
| 95 |
+
| 0.4761 | 36.9973 | 6937 | 2.4947 | 0.5525 |
|
| 96 |
+
| 0.4802 | 38.0 | 7125 | 2.4657 | 0.5496 |
|
| 97 |
+
| 0.4861 | 38.9973 | 7312 | 2.4472 | 0.5504 |
|
| 98 |
+
| 0.4875 | 40.0 | 7500 | 2.4841 | 0.5489 |
|
| 99 |
+
| 0.4681 | 40.9973 | 7687 | 2.4855 | 0.5484 |
|
| 100 |
+
| 0.4661 | 42.0 | 7875 | 2.5166 | 0.5491 |
|
| 101 |
+
| 0.47 | 42.9973 | 8062 | 2.5159 | 0.5487 |
|
| 102 |
+
| 0.4679 | 44.0 | 8250 | 2.5625 | 0.5490 |
|
| 103 |
+
| 0.4688 | 44.9973 | 8437 | 2.4849 | 0.5482 |
|
| 104 |
+
| 0.4699 | 46.0 | 8625 | 2.5193 | 0.5486 |
|
| 105 |
+
| 0.4724 | 46.9973 | 8812 | 2.5711 | 0.5462 |
|
| 106 |
+
| 0.4753 | 48.0 | 9000 | 2.5664 | 0.5465 |
|
| 107 |
+
| 0.4577 | 48.9973 | 9187 | 2.5205 | 0.5466 |
|
| 108 |
+
| 0.4642 | 49.8667 | 9350 | 2.5245 | 0.5456 |
|
| 109 |
+
|
| 110 |
+
|
| 111 |
+
### Framework versions
|
| 112 |
+
|
| 113 |
+
- PEFT 0.5.0
|
| 114 |
+
- Transformers 4.41.1
|
| 115 |
+
- Pytorch 2.1.0+cu121
|
| 116 |
+
- Datasets 2.19.1
|
| 117 |
+
- Tokenizers 0.19.1
|