File size: 2,772 Bytes
a25c1ff c49c1a9 9c95f36 c49c1a9 9c95f36 c49c1a9 ace2f35 563477e c49c1a9 563477e 4e3899c ace2f35 563477e ace2f35 563477e ace2f35 549684b 4e3899c 549684b 4e3899c 549684b 4e3899c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 |
import gradio as gr
import os
from pathlib import Path
import argparse
from huggingface_hub import snapshot_download
# repo_name = "TheBloke/Mistral-7B-v0.1-GGUF"
# model_file = "mistral-7b-v0.1.Q6_K.gguf"
#repo_name = 'HumanityFTW/so_rude'
#model_file = "mistral-comedy-2.0-ckpt-600.Q6_K.gguf"
repo_name = 'TheBloke/OpenHermes-2.5-Mistral-7B-GGUF'
model_file = "openhermes-2.5-mistral-7b.Q4_K_M.gguf"
print('Fetching model:', repo_name, model_file)
snapshot_download(repo_id=repo_name, local_dir=".", allow_patterns=model_file)
print('Done fetching model:')
DEFAULT_MODEL_PATH = model_file
from llama_cpp import Llama
llm = Llama(model_path=model_file, model_type="mistral")
def predict(input, chatbot, max_length, top_p, temperature, history):
chatbot.append((input, ""))
response = ""
history.append(input)
for output in llm(input, stream=True, temperature=temperature, top_p=top_p, max_tokens=max_length, ):
piece = output['choices'][0]['text']
response += piece
chatbot[-1] = (chatbot[-1][0], response)
yield chatbot, history
history.append(response)
yield chatbot, history
def reset_user_input():
return gr.update(value="")
def reset_state():
return [], []
def AIPatient(message):
global isFirstRun, history,context
if isFirstRun:
context = initContext
isFirstRun = False
#else:
#for turn in history:
# context += f"\n<|im_start|> Nurse: {turn[0]}\n<|im_start|> Barry: {turn[1]}"
context += """
<|im_start|>nurse
Nurse: """+message+"""
<|im_start|>barry
Barry:
"""
response = ""
# Here, you should add the code to generate the response using your model
# For example:
while(len(response) < 1):
print("here")
output = llm(context, max_tokens=400, stop=["Nurse:"], echo=False)
response = output["choices"][0]["text"]
response = response.strip()
context += response
print (context)
history.append((message,response))
return history
with gr.Blocks() as demo:
gr.Markdown("# AI Patient Chatbot")
with gr.Group():
with gr.Tab("Patient Chatbot"):
chatbot = gr.Chatbot()
message = gr.Textbox(label="Enter your message to Barry", placeholder="Type here...", lines=2)
send_message = gr.Button("Submit")
send_message.click(AIPatient, inputs=[message], outputs=[chatbot])
save_chatlog = gr.Button("Save Chatlog")
#send_message.click(SaveChatlog, inputs=[message], outputs=[chatbot])
#message.submit(AIPatient, inputs=[message], outputs=[chatbot])
demo.launch(debug=True,share=False,inbrowser=True)
|