Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,85 @@
|
|
1 |
-
---
|
2 |
-
license: mit
|
3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: mit
|
3 |
+
---
|
4 |
+
# 0502
|
5 |
+
|
6 |
+
This model is a fine-tuned version of [/datas/huggingface/Qwen1.5-7B](https://huggingface.co//datas/huggingface/Qwen1.5-7B) on the alpaca_formatted_ift_eft_dft_rft_2048 dataset.
|
7 |
+
It achieves the following results on the evaluation set:
|
8 |
+
|
9 |
+
- Loss: 0.8510
|
10 |
+
|
11 |
+
## Model description
|
12 |
+
|
13 |
+
Qwen1.5 is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:
|
14 |
+
|
15 |
+
* 8 model sizes, including 0.5B, 1.8B, 4B, 7B, 14B, 32B and 72B dense models, and an MoE model of 14B with 2.7B activated;
|
16 |
+
* Significant performance improvement in Chat models;
|
17 |
+
* Multilingual support of both base and chat models;
|
18 |
+
* Stable support of 32K context length for models of all sizes
|
19 |
+
* No need of `trust_remote_code`.
|
20 |
+
|
21 |
+
For more details, please refer to the [blog post](https://qwenlm.github.io/blog/qwen1.5/) and [GitHub repo](https://github.com/QwenLM/Qwen1.5).
|
22 |
+
|
23 |
+
## Intended uses & limitations
|
24 |
+
|
25 |
+
More information needed
|
26 |
+
|
27 |
+
## Training and evaluation data
|
28 |
+
|
29 |
+
More information needed
|
30 |
+
|
31 |
+
## Training procedure
|
32 |
+
|
33 |
+
### Training hyperparameters
|
34 |
+
|
35 |
+
The following hyperparameters were used during training:
|
36 |
+
|
37 |
+
- learning_rate: 5.5e-06
|
38 |
+
- train_batch_size: 2
|
39 |
+
- eval_batch_size: 1
|
40 |
+
- seed: 42
|
41 |
+
- distributed_type: multi-GPU
|
42 |
+
- num_devices: 2
|
43 |
+
- gradient_accumulation_steps: 2
|
44 |
+
- total_train_batch_size: 8
|
45 |
+
- total_eval_batch_size: 2
|
46 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
47 |
+
- lr_scheduler_type: cosine
|
48 |
+
- lr_scheduler_warmup_steps: 200
|
49 |
+
- num_epochs: 5.0
|
50 |
+
- mixed_precision_training: Native AMP
|
51 |
+
|
52 |
+
### Training results
|
53 |
+
|
54 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
55 |
+
| :-----------: | :----: | :--: | :-------------: |
|
56 |
+
| 1.1252 | 0.2335 | 200 | 1.0653 |
|
57 |
+
| 1.0075 | 0.4670 | 400 | 0.9458 |
|
58 |
+
| 1.2782 | 0.7005 | 600 | 0.9099 |
|
59 |
+
| 0.8558 | 0.9340 | 800 | 0.8929 |
|
60 |
+
| 0.922 | 1.1675 | 1000 | 0.8817 |
|
61 |
+
| 0.8985 | 1.4011 | 1200 | 0.8758 |
|
62 |
+
| 0.8273 | 1.6346 | 1400 | 0.8700 |
|
63 |
+
| 0.9136 | 1.8681 | 1600 | 0.8655 |
|
64 |
+
| 0.9963 | 2.1016 | 1800 | 0.8614 |
|
65 |
+
| 1.0214 | 2.3351 | 2000 | 0.8597 |
|
66 |
+
| 0.8823 | 2.5686 | 2200 | 0.8569 |
|
67 |
+
| 0.9265 | 2.8021 | 2400 | 0.8557 |
|
68 |
+
| 0.8033 | 3.0356 | 2600 | 0.8541 |
|
69 |
+
| 0.992 | 3.2691 | 2800 | 0.8527 |
|
70 |
+
| 0.7903 | 3.5026 | 3000 | 0.8522 |
|
71 |
+
| 0.8686 | 3.7361 | 3200 | 0.8518 |
|
72 |
+
| 0.954 | 3.9696 | 3400 | 0.8515 |
|
73 |
+
| 0.6472 | 4.2032 | 3600 | 0.8513 |
|
74 |
+
| 0.8799 | 4.4367 | 3800 | 0.8510 |
|
75 |
+
| 0.9454 | 4.6702 | 4000 | 0.8510 |
|
76 |
+
| 0.9496 | 4.9037 | 4200 | 0.8510 |
|
77 |
+
|
78 |
+
|
79 |
+
### Framework versions
|
80 |
+
|
81 |
+
- PEFT 0.10.0
|
82 |
+
- Transformers 4.40.0
|
83 |
+
- Pytorch 2.1.0+cu121
|
84 |
+
- Datasets 2.14.5
|
85 |
+
- Tokenizers 0.19.1
|