Upload README.md
Browse files
README.md
ADDED
@@ -0,0 +1,90 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: other
|
3 |
+
language:
|
4 |
+
- en
|
5 |
+
pipeline_tag: text-generation
|
6 |
+
tags:
|
7 |
+
- gguf
|
8 |
+
- imatrix
|
9 |
+
- stable-code-3b
|
10 |
+
- stabilityai
|
11 |
+
---
|
12 |
+
Quantizations of https://huggingface.co/stabilityai/stable-code-3b
|
13 |
+
|
14 |
+
# From original readme
|
15 |
+
|
16 |
+
## Usage
|
17 |
+
|
18 |
+
Get started generating text with `stable-code-3b` by using the following code snippet:
|
19 |
+
|
20 |
+
```python
|
21 |
+
import torch
|
22 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
23 |
+
tokenizer = AutoTokenizer.from_pretrained("stabilityai/stable-code-3b")
|
24 |
+
model = AutoModelForCausalLM.from_pretrained(
|
25 |
+
"stabilityai/stable-code-3b",
|
26 |
+
torch_dtype="auto",
|
27 |
+
)
|
28 |
+
model.cuda()
|
29 |
+
inputs = tokenizer("import torch\nimport torch.nn as nn", return_tensors="pt").to(model.device)
|
30 |
+
tokens = model.generate(
|
31 |
+
**inputs,
|
32 |
+
max_new_tokens=48,
|
33 |
+
temperature=0.2,
|
34 |
+
do_sample=True,
|
35 |
+
)
|
36 |
+
print(tokenizer.decode(tokens[0], skip_special_tokens=True))
|
37 |
+
```
|
38 |
+
|
39 |
+
### Run with Fill in Middle (FIM) ⚡️
|
40 |
+
|
41 |
+
<details>
|
42 |
+
<summary> Click to expand </summary>
|
43 |
+
|
44 |
+
```python
|
45 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
46 |
+
tokenizer = AutoTokenizer.from_pretrained("stabilityai/stable-code-3b")
|
47 |
+
model = AutoModelForCausalLM.from_pretrained(
|
48 |
+
"stabilityai/stable-code-3b",
|
49 |
+
torch_dtype="auto",
|
50 |
+
attn_implementation="flash_attention_2",
|
51 |
+
)
|
52 |
+
model.cuda()
|
53 |
+
inputs = tokenizer("<fim_prefix>def fib(n):<fim_suffix> else:\n return fib(n - 2) + fib(n - 1)<fim_middle>", return_tensors="pt").to(model.device)
|
54 |
+
tokens = model.generate(
|
55 |
+
**inputs,
|
56 |
+
max_new_tokens=48,
|
57 |
+
temperature=0.2,
|
58 |
+
do_sample=True,
|
59 |
+
)
|
60 |
+
print(tokenizer.decode(tokens[0], skip_special_tokens=True))
|
61 |
+
```
|
62 |
+
|
63 |
+
</details>
|
64 |
+
|
65 |
+
### Run with Flash Attention 2 ⚡️
|
66 |
+
|
67 |
+
<details>
|
68 |
+
<summary> Click to expand </summary>
|
69 |
+
|
70 |
+
```python
|
71 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
72 |
+
tokenizer = AutoTokenizer.from_pretrained("stabilityai/stable-code-3b", trust_remote_code=True)
|
73 |
+
model = AutoModelForCausalLM.from_pretrained(
|
74 |
+
"stabilityai/stable-code-3b",
|
75 |
+
trust_remote_code=True,
|
76 |
+
torch_dtype="auto",
|
77 |
+
+ attn_implementation="flash_attention_2",
|
78 |
+
)
|
79 |
+
model.cuda()
|
80 |
+
inputs = tokenizer("import torch\nimport torch.nn as nn", return_tensors="pt").to(model.device)
|
81 |
+
tokens = model.generate(
|
82 |
+
**inputs,
|
83 |
+
max_new_tokens=48,
|
84 |
+
temperature=0.2,
|
85 |
+
do_sample=True,
|
86 |
+
)
|
87 |
+
print(tokenizer.decode(tokens[0], skip_special_tokens=True))
|
88 |
+
```
|
89 |
+
|
90 |
+
</details>
|