File size: 1,759 Bytes
f413841
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
library_name: transformers
tags:
- generated_from_trainer
datasets:
- Alant2000/taiga_cleared
metrics:
- accuracy
model-index:
- name: output
  results:
  - task:
      name: Causal Language Modeling
      type: text-generation
    dataset:
      name: Alant2000/taiga_cleared
      type: Alant2000/taiga_cleared
    metrics:
    - name: Accuracy
      type: accuracy
      value: 0.3070757857382329
---

Модель обучена в рамках практического задания по курсу "Практические аспекты обучения LLM": https://github.com/RefalMachine/rcc_msu_llm_course_autumn_2024/tree/main

This model is a fine-tuned version of ruadapt_qwen2.5_1.5B_ext_u48_mean_init on the [Alant2000/taiga_cleared](https://huggingface.co/datasets/Alant2000/taiga_cleared) dataset.
It achieves the following results on the evaluation set:
- Loss: 4.0484
- Accuracy: 0.3071

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 0.0001
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- distributed_type: multi-GPU
- num_devices: 16
- gradient_accumulation_steps: 4
- total_train_batch_size: 1024
- total_eval_batch_size: 256
- optimizer: Adam with betas=(0.9,0.95) and epsilon=1e-05
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- num_epochs: 1.0

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Accuracy |
|:-------------:|:------:|:----:|:---------------:|:--------:|
| No log        | 0.0008 | 1    | 7.8632          | 0.1493   |
| 4.0503        | 0.7593 | 1000 | 4.0488          | 0.3070   |


### Framework versions

- Transformers 4.45.2
- Pytorch 2.3.0a0+6ddf5cf85e.nv24.04
- Datasets 2.18.0
- Tokenizers 0.20.3