:sparkles: initial commit - llm calculator

.github/workflows/push_to_hub.yml

@@ -0,0 +1,20 @@
+name: Sync to Hugging Face hub
+on:
+  push:
+    branches: [main]
+
+  # to run this workflow manually from the Actions tab
+  workflow_dispatch:
+
+jobs:
+  sync-to-hub:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v3
+        with:
+          fetch-depth: 0
+          lfs: true
+      - name: Push to hub
+        env:
+          HF_TOKEN: ${{ secrets.HF_TOKEN }}
+        run: git push https://willsh1997:[email protected]/spaces/willsh1997/llm-calculator main

README.md

@@ -0,0 +1,14 @@
+---
+title: llm calculator
+emoji: 🐨
+colorFrom: red
+colorTo: green
+sdk: gradio
+sdk_version: 5.23.3
+app_file: llmcalc_gradio.py
+pinned: false
+license: apache-2.0
+short_description: llm calculator with llama backend
+---
+
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

llmcalc_gradio.py

@@ -0,0 +1,106 @@
+import spaces
+from transformers import AutoModelForSeq2SeqLM, AutoTokenizer, AutoModelForCausalLM
+import torch
+from transformers import pipeline
+import pandas as pd
+import gradio as gr
+
+#Llama 3.2 1b setup
+# quantization_config = BitsAndBytesConfig(load_in_4bit=True)
+torch_device = "cuda" if torch.cuda.is_available() else ("mps" if torch.mps.is_available() else "cpu")
+
+torch_dtype = torch.float16 if torch_device in ["cuda", "mps"] else torch.float32
+
+llama_model=AutoModelForCausalLM.from_pretrained("meta-llama/Llama-3.2-3B-Instruct", 
+                                           #  quantization_config=quantization_config, 
+                                           torch_dtype=torch_dtype, 
+                                           device_map=torch_device,
+                                            load_in_4bit=True) #for puny devices like mine.
+
+llama_tokenizer=AutoTokenizer.from_pretrained("meta-llama/Llama-3.2-3B-Instruct")
+
+
+model_id = "meta-llama/Llama-3.2-3B-Instruct"
+pipe = pipeline(
+    "text-generation",
+    model=llama_model,
+    tokenizer=llama_tokenizer,
+    torch_dtype=torch.bfloat16,
+    device_map="auto",
+    #load_in_4bit = True #for lil machines like minegit statu
+)
+
+@spaces.GPU
+def llama_calculator(left_num, operation, right_num):
+    """
+    stupid func for asking llama a question and then getting an answer
+    inputs:
+    - input_question [str]: question for llama to answer
+    outputs:
+    - response [str]: llama's response
+    """
+    
+    messages = [
+    {"role": "system", "content": """You are a calculator, which generates correct answers to arithmetic questions posed by the user. Answer with ONLY the numerical output of the answer - no text. Answer as accurately as possible, to as many decimal places as required. Here are some examples of questions and answers you may see:
+    
+USER: 4+8
+CHATBOT: 12
+
+---
+
+USER: 182*526
+CHATBOT: 95732
+
+---
+
+USER: 97.35/15.02
+CHATBOT: 6.48135818908
+
+    """},
+    {"role": "user", "content": str(left_num)+' '+str(operation)+' '+str(right_num)},
+    ]
+    outputs = pipe(
+        messages,
+        max_new_tokens=512
+    )
+    response = outputs[0]["generated_text"][-1]['content']
+    return response
+
+
+# QA translation roundtrip
+
+@spaces.GPU
+def gradio_func(left_num, operation, right_num):
+    """
+    silly wrapper function for gradio that turns all inputs into a single func. runs both the LHS and RHS of teh 'app' in order to let gradio work correctly.
+    """
+    left_output = llama_calculator(left_num, operation, right_num)
+    right_output = eval(str(left_num)+' '+str(operation)+' '+str(right_num))
+    return left_output, right_output
+
+# Create the Gradio interface
+def create_interface():
+    # Get available languages from the flores_dict
+    
+    with gr.Blocks() as demo:
+        with gr.Row():
+            left_num = gr.Number(label="left number", interactive=True)
+            operation = gr.Dropdown(["+","-","*","/"], multiselect = False)
+            right_num = gr.Number(label="right number", interactive=True)
+        with gr.Row():
+            submit_btn = gr.Button("calculate")
+        with gr.Row():
+            left_output = gr.Textbox(label="LLM output", interactive=False)
+            right_output = gr.Textbox(label="actual calculated answer", interactive=False)
+            
+        submit_btn.click(
+            fn=gradio_func,
+            inputs=[left_num, operation, right_num],
+            outputs=[left_output, right_output]
+        )
+    
+    return demo
+
+# Launch the app
+demo = create_interface()
+demo.launch()

requirements.txt

@@ -0,0 +1,87 @@
+accelerate==1.4.0
+aiofiles==23.2.1
+annotated-types==0.7.0
+anyio==4.8.0
+asttokens==3.0.0
+bitsandbytes==0.45.4
+certifi==2025.1.31
+charset-normalizer==3.4.1
+click==8.1.8
+comm==0.2.2
+debugpy==1.8.12
+decorator==5.1.1
+exceptiongroup==1.2.2
+executing==2.2.0
+fastapi==0.115.8
+ffmpy==0.5.0
+filelock==3.17.0
+fsspec==2025.2.0
+gradio==5.16.1
+gradio_client==1.7.0
+h11==0.14.0
+httpcore==1.0.7
+httpx==0.28.1
+huggingface-hub==0.28.1
+idna==3.10
+ipykernel==6.29.5
+ipython==8.32.0
+jedi==0.19.2
+Jinja2==3.1.5
+jupyter_client==8.6.3
+jupyter_core==5.7.2
+markdown-it-py==3.0.0
+MarkupSafe==2.1.5
+matplotlib-inline==0.1.7
+mdurl==0.1.2
+mpmath==1.3.0
+nest-asyncio==1.6.0
+networkx==3.4.2
+numpy==2.2.3
+orjson==3.10.15
+packaging==24.2
+pandas==2.2.3
+parso==0.8.4
+pexpect==4.9.0
+pillow==11.1.0
+platformdirs==4.3.6
+prompt_toolkit==3.0.50
+psutil==7.0.0
+ptyprocess==0.7.0
+pure_eval==0.2.3
+pydantic==2.10.6
+pydantic_core==2.27.2
+pydub==0.25.1
+Pygments==2.19.1
+python-dateutil==2.9.0.post0
+python-multipart==0.0.20
+pytz==2025.1
+PyYAML==6.0.2
+pyzmq==26.2.1
+regex==2024.11.6
+requests==2.32.3
+rich==13.9.4
+ruff==0.9.6
+safehttpx==0.1.6
+safetensors==0.5.2
+semantic-version==2.10.0
+shellingham==1.5.4
+six==1.17.0
+sniffio==1.3.1
+stack-data==0.6.3
+starlette==0.45.3
+sympy==1.13.1
+tokenizers==0.21.0
+tomlkit==0.13.2
+torch==2.4.0
+tornado==6.4.2
+tqdm==4.67.1
+traitlets==5.14.3
+transformers==4.49.0
+typer==0.15.1
+typing_extensions==4.12.2
+tzdata==2025.1
+urllib3==2.3.0
+uvicorn==0.34.0
+wcwidth==0.2.13
+websockets==14.2
+