File size: 1,415 Bytes
e7d50e2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2d55c94
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
pipeline_tag: text-generation
inference: true
widget:
- text: 'Hello!'
  example_title: Hello world
  group: Python
library_name: transformers
---

This model is randomly initialized, using the config from [deepseek-ai/deepseek-llm-67b-chat](https://huggingface.co/deepseek-ai/deepseek-llm-67b-chat) but with smaller size. 
Note the model is in float16.

Codes:
```python
import transformers
import torch
import os
from huggingface_hub import create_repo, upload_folder

source_model_id = 'deepseek-ai/deepseek-llm-67b-chat'
save_path = '/tmp/yujiepan/deepseek-llm-tiny-random'
repo_id = 'yujiepan/deepseek-llm-tiny-random'

config = transformers.AutoConfig.from_pretrained(
    source_model_id, trust_remote_code=True)
config.hidden_size = 2
config.intermediate_size = 4
config.num_attention_heads = 2
config.num_hidden_layers = 2
config.num_key_value_heads = 1

model = transformers.AutoModelForCausalLM.from_config(
    config, trust_remote_code=True)
model = model.half()
model.save_pretrained(save_path)

tokenizer = transformers.AutoTokenizer.from_pretrained(
    source_model_id, trust_remote_code=True)
tokenizer.save_pretrained(save_path)

result = transformers.pipelines.pipeline(
    'text-generation',
    model=model.float(), tokenizer=tokenizer)('Hello')
print(result)

os.system(f'ls -alh {save_path}')
create_repo(repo_id, exist_ok=True)
upload_folder(repo_id=repo_id, folder_path=save_path)
```