Update app.py

app.py

@@ -1,54 +1,93 @@
-"""Python file to serve as the frontend"""
-import streamlit as st
-from streamlit_chat import message
-
-from langchain.chains import ConversationChain, LLMChain
-from langchain import PromptTemplate
-from langchain.llms.base import LLM
-from langchain.memory import ConversationBufferWindowMemory
-from typing import Optional, List, Mapping, Any
-
 import torch
 from peft import PeftModel
 import transformers
+import gradio as gr
 
+assert (
+    "LlamaTokenizer" in transformers._import_structure["models.llama"]
+), "LLaMA is now in HuggingFace's main branch.\nPlease reinstall it: pip uninstall transformers && pip install git+https://github.com/huggingface/transformers.git"
 from transformers import LlamaTokenizer, LlamaForCausalLM, GenerationConfig
-from transformers import BitsAndBytesConfig
 
 tokenizer = LlamaTokenizer.from_pretrained("decapoda-research/llama-7b-hf")
 
+BASE_MODEL = "decapoda-research/llama-7b-hf"
+LORA_WEIGHTS = "tloen/alpaca-lora-7b"
+
+if torch.cuda.is_available():
+    device = "cuda"
+else:
+    device = "cpu"
+
+try:
+    if torch.backends.mps.is_available():
+        device = "mps"
+except:
+    pass
+
+if device == "cuda":
+    model = LlamaForCausalLM.from_pretrained(
+        BASE_MODEL,
+        load_in_8bit=False,
+        torch_dtype=torch.float16,
+        device_map="auto",
+    )
+    model = PeftModel.from_pretrained(
+        model, LORA_WEIGHTS, torch_dtype=torch.float16, force_download=True
+    )
+elif device == "mps":
+    model = LlamaForCausalLM.from_pretrained(
+        BASE_MODEL,
+        device_map={"": device},
+        torch_dtype=torch.float16,
+    )
+    model = PeftModel.from_pretrained(
+        model,
+        LORA_WEIGHTS,
+        device_map={"": device},
+        torch_dtype=torch.float16,
+    )
+else:
+    model = LlamaForCausalLM.from_pretrained(
+        BASE_MODEL, device_map={"": device}, low_cpu_mem_usage=True
+    )
+    model = PeftModel.from_pretrained(
+        model,
+        LORA_WEIGHTS,
+        device_map={"": device},
+    )
 
 
-quantization_config = BitsAndBytesConfig(llm_int8_enable_fp32_cpu_offload=True)
-
-model = LlamaForCausalLM.from_pretrained(
-    "decapoda-research/llama-7b-hf",
-    # load_in_8bit=True,
-    # torch_dtype=torch.float16,
-    #device_map="auto",
-    device_map={"":"cpu"},
-    max_memory={"cpu":"15GiB"},
-    quantization_config=quantization_config
-)
-model = PeftModel.from_pretrained(
-    model, "tloen/alpaca-lora-7b",
-    # torch_dtype=torch.float16,
-    device_map={"":"cpu"},
-)
+def generate_prompt(instruction, input=None):
+    if input:
+        return f"""Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.
+### Instruction:
+{instruction}
+### Input:
+{input}
+### Response:"""
+    else:
+        return f"""Below is an instruction that describes a task. Write a response that appropriately completes the request.
+### Instruction:
+{instruction}
+### Response:"""
 
-device = "cpu"
-print("model device :", model.device, flush=True)
-# model.to(device)
+model.half()
 model.eval()
-
-def evaluate_raw_prompt(
-        prompt:str,
-        temperature=0.1,
-        top_p=0.75,
-        top_k=40,
-        num_beams=4,
-        **kwargs,
+if torch.__version__ >= "2":
+    model = torch.compile(model)
+
+
+def evaluate(
+    instruction,
+    input=None,
+    temperature=0.1,
+    top_p=0.75,
+    top_k=40,
+    num_beams=4,
+    max_new_tokens=128,
+    **kwargs,
 ):
+    prompt = generate_prompt(instruction, input)
     inputs = tokenizer(prompt, return_tensors="pt")
     input_ids = inputs["input_ids"].to(device)
     generation_config = GenerationConfig(
@@ -64,93 +103,57 @@ def evaluate_raw_prompt(
             generation_config=generation_config,
             return_dict_in_generate=True,
             output_scores=True,
-            max_new_tokens=256,
+            max_new_tokens=max_new_tokens,
         )
     s = generation_output.sequences[0]
     output = tokenizer.decode(s)
-    # return output
     return output.split("### Response:")[1].strip()
 
-class AlpacaLLM(LLM):
-    temperature: float
-    top_p: float
-    top_k: int
-    num_beams: int
-    @property
-    def _llm_type(self) -> str:
-        return "custom"
-    
-    def _call(self, prompt: str, stop: Optional[List[str]] = None) -> str:
-        if stop is not None:
-            raise ValueError("stop kwargs are not permitted.")
-        answer = evaluate_raw_prompt(prompt, 
-                            top_p= self.top_p,
-                            top_k= self.top_k,
-                            num_beams= self.num_beams,
-                            temperature= self.temperature
-                            )
-        return answer
-    
-    @property
-    def _identifying_params(self) -> Mapping[str, Any]:
-        """Get the identifying parameters."""
-        return {
-            "top_p": self.top_p,
-            "top_k": self.top_k,
-            "num_beams": self.num_beams,
-            "temperature": self.temperature
-            }
-
 
-template = """Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.
-### Instruction:
-You are a chatbot, you should answer my last question very briefly. You are consistent and non repetitive. 
-### Chat:
-{history}
-Human: {human_input}
-### Response:"""
-
-prompt = PromptTemplate(
-    input_variables=["history","human_input"],
-    template=template,
+g = gr.Interface(
+    fn=evaluate,
+    inputs=[
+        gr.components.Textbox(
+            lines=2, label="Instruction", placeholder="Tell me about alpacas."
+        ),
+        gr.components.Textbox(lines=2, label="Input", placeholder="none"),
+        gr.components.Slider(minimum=0, maximum=1, value=0.1, label="Temperature"),
+        gr.components.Slider(minimum=0, maximum=1, value=0.75, label="Top p"),
+        gr.components.Slider(minimum=0, maximum=100, step=1, value=40, label="Top k"),
+        gr.components.Slider(minimum=1, maximum=4, step=1, value=4, label="Beams"),
+        gr.components.Slider(
+            minimum=1, maximum=512, step=1, value=128, label="Max tokens"
+        ),
+    ],
+    outputs=[
+        gr.inputs.Textbox(
+            lines=5,
+            label="Output",
+        )
+    ],
+    title="🦙🌲 Alpaca-LoRA",
+    description="Alpaca-LoRA is a 7B-parameter LLaMA model finetuned to follow instructions. It is trained on the [Stanford Alpaca](https://github.com/tatsu-lab/stanford_alpaca) dataset and makes use of the Huggingface LLaMA implementation. For more information, please visit [the project's website](https://github.com/tloen/alpaca-lora).",
 )
-
-
-def load_chain():
-    """Logic for loading the chain you want to use should go here."""
-    llm = AlpacaLLM(top_p=0.75, top_k=40, num_beams=4, temperature=0.1)
-    # chain = ConversationChain(llm=llm)
-    chain = LLMChain(llm=llm, prompt=prompt, memory=ConversationBufferWindowMemory(k=2))
-    return chain
-
-chain = load_chain()
-
-# From here down is all the StreamLit UI.
-st.set_page_config(page_title="LangChain Demo", page_icon=":robot:")
-st.header("LangChain Demo")
-
-if "generated" not in st.session_state:
-    st.session_state["generated"] = []
-
-if "past" not in st.session_state:
-    st.session_state["past"] = []
-
-
-def get_text():
-    input_text = st.text_input("Human: ", "Hello, how are you?", key="input")
-    return input_text
-
-
-user_input = get_text()
-
-if user_input:
-    output = chain.predict(human_input=user_input)
-
-    st.session_state.past.append(user_input)
-    st.session_state.generated.append(output)
-
-if st.session_state["generated"]:
-
-    for i in range(len(st.session_state["generated"]) - 1, -1, -1):
-        message(st.session_state["generated"][i], key=str(i))
-        message(st.session_state["past"][i], is_user=True, key=str(i) + "_user")
+g.queue(concurrency_count=1)
+g.launch()
+
+# Old testing code follows.
+
+"""
+if __name__ == "__main__":
+    # testing code for readme
+    for instruction in [
+        "Tell me about alpacas.",
+        "Tell me about the president of Mexico in 2019.",
+        "Tell me about the king of France in 2019.",
+        "List all Canadian provinces in alphabetical order.",
+        "Write a Python program that prints the first 10 Fibonacci numbers.",
+        "Write a program that prints the numbers from 1 to 100. But for multiples of three print 'Fizz' instead of the number and for the multiples of five print 'Buzz'. For numbers which are multiples of both three and five print 'FizzBuzz'.",
+        "Tell me five words that rhyme with 'shock'.",
+        "Translate the sentence 'I have no mouth but I must scream' into Spanish.",
+        "Count up from 1 to 500.",
+    ]:
+        print("Instruction:", instruction)
+        print("Response:", evaluate(instruction))
+        print()
+"""