--- language: - en license: apache-2.0 tags: - lora-alpaca - alpaca - lora - LLaMA - Stanford Alpaca datasets: - mrzlab630/trading-candles pipeline_tag: question-answering widget: - text: "identify candle" context: "open: 38752.71, close: 38843.7, high: 38847.4, low: 38752.71" example_title: "identify candle" - text: "find candle" context: "38811.24,38838.41,38846.71,38736.24,234.00,45275276.00,59816.00,441285.00,645.00,84176.00,1694619.00,15732335.00" example_title: "find candle" - text: "find candle: Bullish" context: "38751.32,38818.6,38818.6,38695.03,62759348.00,2605789.00,71030.00,820738.00,59659.00,724738.00,7368363.00,50654.00" example_title: "find candle: Bullish" --- ## About: The model was fine-tuned on the LLaMA 7B. [weights_Llama_7b](https://huggingface.co/mrzlab630/weights_Llama_7b) the model is able to identify trading candles. the model knows about: - Four Price Doji, - Inverted Hammer, - Hammer, - Hanging Man, - Doji, - Long-legged doji, - Dragonfly doji, - Inverted Doji, - Bullish, - Bearish ## Prompts: ``` Instruction: identify candle Input: open:241.5,close:232.9, high:241.7, low:230.8 or Input: 241.5,232.9,241.7,230.8 Output: Bearish ``` ``` Instruction: identify candle Input: open:241.5,close:232.9, high:241.7, low:230.8 or Input: 241.5,232.9, 241.7,230.8 Output: Doji ``` ``` Instruction: identify candle:open:241.5,close:232.9, high:241.7, low:230.8 or Instruction: identify candle:241.5,232.9,241.7, 230.8 Output: Bearish:241.5,close:232.9, high:241.7, low:230.8 ``` ``` Instruction: find candle Input: 38811.24,38838.41,38846.71,38736.24,234.00,45275276.00,59816.00,441285.00,645.00,84176.00,1694619.00,15732335.00 Output: Dragonfly doji:38811.24,38838.41,38846.71,38736.24 ``` Instruction: find candle: {%candleName%} ``` Instruction: find candle: Bullish Input: 38751.32,38818.6,38818.6,38695.03,62759348.00,2605789.00,71030.00,820738.00,59659.00,724738.00,7368363.00,50654.00 Output: Bullish:38751.32,38818.6,38818.6,38695.03 ``` ### RUN ``` import sys 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 SHARE_GRADIO=True LOAD_8BIT = False BASE_MODEL = "mrzlab630/weights_Llama_7b" LORA_WEIGHTS = "mrzlab630/lora-alpaca-trading-candles" tokenizer = LlamaTokenizer.from_pretrained(BASE_MODEL) 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=LOAD_8BIT, torch_dtype=torch.float16, device_map="auto", ) model = PeftModel.from_pretrained( model, LORA_WEIGHTS, torch_dtype=torch.float16, ) 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}, ) 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 not LOAD_8BIT: model.half() # seems to fix bugs for some users. model.eval() if torch.__version__ >= "2" and sys.platform != "win32": 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() 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=2000, step=1, value=128, label="Max tokens" ), ], outputs=[ gr.inputs.Textbox( lines=5, label="Output", ) ], title="💹 🕯 Alpaca-LoRA-Trading-Candles", description="Alpaca-LoRA-Trading-Candles is a 7B-parameter LLaMA model tuned to execute instructions. It is trained on the [trading candles] dataset(https://huggingface.co/datasets/mrzlab630/trading-candles) and uses the Huggingface LLaMA implementation. For more information, visit [project website](https://huggingface.co/mrzlab630/lora-alpaca-trading-candles).\nPrompts:\nInstruction: identify candle, Input: open:241.5,close:232.9, high:241.7, low:230.8\nInstruction: find candle, Input: 38811.24,38838.41,38846.71,38736.24,234.00,45275276.00,59816.00,441285.00,645.00,84176.00,1694619.00,15732335.00\nInstruction: find candle: Bullish, Input: 38751.32,38818.6,38818.6,38695.03,62759348.00,2605789.00,71030.00,820738.00,59659.00,724738.00,7368363.00,50654.00", ).launch(server_name="0.0.0.0", share=SHARE_GRADIO) # 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() """ ```