shekharp77 commited on
Commit
7efc171
1 Parent(s): fb4c123

first commit

Browse files
Files changed (4) hide show
  1. .gitignore +1 -0
  2. Dockerfile +16 -0
  3. main.py +24 -0
  4. requrement.txt +56 -0
.gitignore ADDED
@@ -0,0 +1 @@
 
 
1
+ .idea/
Dockerfile ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # read the doc: https://huggingface.co/docs/hub/spaces-sdks-docker
2
+ # you will also find guides on how best to write your Dockerfile
3
+
4
+ FROM python:3.9
5
+
6
+ WORKDIR /code
7
+
8
+ COPY ./requirements.txt /code/requirements.txt
9
+
10
+ # COPY /Users/shekharsingh/code/nsph/models/llemma_7b /code/llemma_7b
11
+
12
+ RUN pip install --no-cache-dir --upgrade -r /code/requirements.txt
13
+
14
+ COPY . .
15
+
16
+ CMD ["uvicorn", "app.main:app", "--host", "0.0.0.0", "--port", "7860"]
main.py ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from transformers import AutoTokenizer, LlamaForCausalLM
2
+ from fastapi import FastAPI
3
+ from pydantic import BaseModel
4
+
5
+ model_path = '"EleutherAI/llemma_7b"'
6
+
7
+ model = LlamaForCausalLM.from_pretrained(model_path)
8
+ tokenizer = AutoTokenizer.from_pretrained(model_path)
9
+
10
+ class validation(BaseModel):
11
+ prompt: str
12
+
13
+
14
+ app = FastAPI()
15
+
16
+
17
+ @app.post("/prompt")
18
+ async def stream(item: validation):
19
+ inputs = tokenizer(item.prompt, return_tensors="pt")
20
+
21
+ generate_ids = model.generate(inputs.input_ids, max_length=30)
22
+
23
+ var = tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
24
+ return var
requrement.txt ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ annotated-types==0.6.0
2
+ anyio==4.3.0
3
+ certifi==2024.2.2
4
+ charset-normalizer==3.3.2
5
+ click==8.1.7
6
+ dnspython==2.6.1
7
+ email_validator==2.1.1
8
+ exceptiongroup==1.2.1
9
+ fastapi==0.111.0
10
+ fastapi-cli==0.0.2
11
+ filelock==3.14.0
12
+ fsspec==2024.3.1
13
+ h11==0.14.0
14
+ httpcore==1.0.5
15
+ httptools==0.6.1
16
+ httpx==0.27.0
17
+ huggingface-hub==0.23.0
18
+ idna==3.7
19
+ Jinja2==3.1.4
20
+ markdown-it-py==3.0.0
21
+ MarkupSafe==2.1.5
22
+ mdurl==0.1.2
23
+ mpmath==1.3.0
24
+ networkx==3.2.1
25
+ numpy==1.26.4
26
+ orjson==3.10.3
27
+ packaging==24.0
28
+ pillow==10.3.0
29
+ pydantic==2.7.1
30
+ pydantic_core==2.18.2
31
+ Pygments==2.18.0
32
+ python-dotenv==1.0.1
33
+ python-multipart==0.0.9
34
+ PyYAML==6.0.1
35
+ regex==2024.4.28
36
+ requests==2.31.0
37
+ rich==13.7.1
38
+ safetensors==0.4.3
39
+ shellingham==1.5.4
40
+ sniffio==1.3.1
41
+ starlette==0.37.2
42
+ sympy==1.12
43
+ tokenizers==0.19.1
44
+ torch==2.3.0
45
+ torchaudio==2.3.0
46
+ torchvision==0.18.0
47
+ tqdm==4.66.4
48
+ transformers==4.40.2
49
+ typer==0.12.3
50
+ typing_extensions==4.11.0
51
+ ujson==5.9.0
52
+ urllib3==2.2.1
53
+ uvicorn==0.29.0
54
+ uvloop==0.19.0
55
+ watchfiles==0.21.0
56
+ websockets==12.0