Update app.py

app.py

@@ -1,60 +1,160 @@
-import gradio as gr
-from torch import cuda, bfloat16
 import torch
+from peft import PeftModel
 import transformers
-from transformers import StoppingCriteria, StoppingCriteriaList
-
-def chatbot(input):
-    device = f'cuda:{cuda.current_device()}' if cuda.is_available() else 'cpu'
-    
-    model = transformers.AutoModelForCausalLM.from_pretrained(
-        'mosaicml/mpt-7b-instruct',
-        trust_remote_code=True,
-        torch_dtype=bfloat16,
-        max_seq_len=2048
+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
+
+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,
     )
-    model.eval()
-    model.to(device)
-    print(f"Model loaded on {device}")
-    
-    
-    tokenizer = transformers.AutoTokenizer.from_pretrained("EleutherAI/gpt-neox-20b")
-
-    for i in range(50):
-        # mtp-7b is trained to add "<|endoftext|>" at the end of generations
-        stop_token_ids = tokenizer.convert_tokens_to_ids(["<|endoftext|>"])
-        
-        # define custom stopping criteria object
-        class StopOnTokens(StoppingCriteria):
-            def __call__(self, input_ids: torch.LongTensor, scores: torch.FloatTensor, **kwargs) -> bool:
-                for stop_id in stop_token_ids:
-                    if input_ids[0][-1] == stop_id:
-                        return True
-                return False
-        
-        stopping_criteria = StoppingCriteriaList([StopOnTokens()])
-        
-        generate_text = transformers.pipeline(
-            model=model, tokenizer=tokenizer,
-            return_full_text=True,  # langchain expects the full text
-            task='text-generation',
-            device=device,
-            # we pass model parameters here too
-            stopping_criteria=stopping_criteria,  # without this model will ramble
-            temperature=0.1,  # 'randomness' of outputs, 0.0 is the min and 1.0 the max
-            top_p=0.15,  # select from top tokens whose probability add up to 15%
-            top_k=0,  # select from top 0 tokens (because zero, relies on top_p)
-            max_new_tokens=64,  # mex number of tokens to generate in the output
-            repetition_penalty=1.1  # without this output begins repeating
+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},
+    )
+
+
+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:"""
+
+if device != "cpu":
+    model.half()
+model.eval()
+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(
+        temperature=temperature,
+        top_p=top_p,
+        top_k=top_k,
+        num_beams=num_beams,
+        **kwargs,
+    )
+    with torch.no_grad():
+        generation_output = model.generate(
+            input_ids=input_ids,
+            generation_config=generation_config,
+            return_dict_in_generate=True,
+            output_scores=True,
+            max_new_tokens=max_new_tokens,
+        )
+    s = generation_output.sequences[0]
+    output = tokenizer.decode(s)
+    return output.split("### Response:")[1].strip()
+
+
+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",
         )
-        
-        res = generate_text(input)
-        output = res[0]["generated_text"]
-        return 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).",
+)
+g.queue(concurrency_count=1)
+g.launch()
 
-inputs = gr.inputs.Textbox(lines=7, label="Chat with AI")
-outputs = gr.outputs.Textbox(label="Reply")
+# Old testing code follows.
 
-gr.Interface(fn=chatbot, inputs=inputs, outputs=outputs, title="Self_Trained_V2",
-             description="Ask anything you want",
-             ).launch()  
+"""
+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()
+"""