Spaces:
Sleeping
Sleeping
File size: 8,117 Bytes
18c0acd |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 |
import logging
import os
import streamlit as st
from langchain.schema import AIMessage, HumanMessage
from langchain_core.pydantic_v1 import SecretStr
from src.models.model_factory import EvaluationChatModelFactory, GeneratorModelFactory
from src.utils import convert_json_to_str, read_json, read_plain_text
from src.whisper_transcription import whisper_stt
logging.basicConfig(level=os.getenv("ENV", "INFO"))
start_container = st.container()
with start_container:
col1_head, col2_head = st.columns([0.15, 0.85])
col1_head.image("./docs/images/logo.png", width=80)
col2_head.title("LinguAIcoach")
st.subheader("_Your :red[AI] English Teacher_")
st.divider()
model_factory = EvaluationChatModelFactory()
gen_factory = GeneratorModelFactory()
exam_guides = read_json("./exam_guides/lessons.json")
config = read_json("./src/config.json")
input_text = None
input_voice = None
with st.sidebar as sidebar:
exam_selection = st.sidebar.selectbox("Select Exam type", list(exam_guides.keys()))
exam_selection = exam_selection if exam_selection else str(next(iter(exam_guides.keys())))
exam_info = exam_guides[exam_selection]
st.session_state.openai_api_key = SecretStr(
os.getenv(
"OPENAI_API_KEY",
st.text_input("OpenAI API Key", key="chatbot_api_key", type="password"),
)
)
st.markdown("[:red[Get your OpenAI API key]](https://platform.openai.com/account/api-keys)")
st.markdown(
"[](https://github.com/alvaroalon2/LinguAIcoach)"
)
logging.debug(f"Selected exam: {exam_selection}")
logging.debug(f"Exam guides: {exam_guides}")
exam_content = read_plain_text(exam_info["file"])
exam_desc = exam_info["description"]
init_prompt = exam_info["init_prompt"]
level = exam_info["level"]
if "current_exam" not in st.session_state:
st.session_state["current_exam"] = ""
if st.session_state["current_exam"] != exam_selection:
logging.debug("Resetting state")
input_text = None
input_voice = None
st.session_state["current_exam"] = exam_selection
st.session_state["messages"] = [AIMessage(content=init_prompt)]
st.session_state.first_run = True
st.session_state["image_url"] = None
st.session_state["question"] = []
st.session_state["exam_type"] = exam_info["type"]
response_container = st.container()
start_container = st.container()
with start_container:
col1_start, col2_start, col3_start = st.columns([0.4, 0.2, 0.4])
voice_container = st.container()
with voice_container:
col1_voice, col2_voice, col3_voice = st.columns([0.39, 0.22, 0.39])
if not st.session_state.openai_api_key:
logging.warning("Please add your OpenAI API key to continue.")
st.info("Please add your OpenAI API key to continue.")
st.stop()
model_eval = model_factory.create_model(
model_class=st.session_state["exam_type"],
openai_api_key=st.session_state.openai_api_key,
level=level,
chat_temperature=config["OPENAI_TEMPERATURE_EVAL"],
)
generator = gen_factory.create_model(
model_class=st.session_state["exam_type"],
openai_api_key=st.session_state.openai_api_key,
exam_prompt=exam_content,
level=level,
description=exam_desc,
chat_temperature=config["OPENAI_TEMPERATURE_GEN"],
history_chat=[
AIMessage(content=f"Previous question (Don't repeat): {q.content}")
for q in st.session_state["question"][-config["N_MAX_HISTORY"] :]
],
img_size=config["IMG_GEN_SIZE"],
)
if "messages" in st.session_state:
logging.debug(f"Starting exercises for exam_type: {st.session_state['exam_type']}")
placeholder_start = col2_start.empty()
start_button = placeholder_start.button("Start exercises!", disabled=not (st.session_state.first_run))
if start_button:
logging.debug("Start button clicked, running exercise")
st.session_state.first_run = False
if st.session_state["exam_type"] == "qa":
start_response = generator.generate()
logging.info(f"Generated first question: {start_response}")
st.session_state["question"].append(AIMessage(content=start_response))
st.session_state["messages"].append(st.session_state["question"][-1])
elif st.session_state["exam_type"] == "img_desc":
st.session_state["image_url"] = generator.generate()
logging.debug(f"Generated first image URL: {st.session_state['image_url']}")
st.session_state["messages"].append(
AIMessage(
content=st.session_state["image_url"],
response_metadata={"type": "image"},
)
)
if not st.session_state.first_run:
placeholder_start.empty()
for msg in st.session_state.messages:
if isinstance(msg, HumanMessage):
if getattr(msg, "response_metadata", None) and msg.response_metadata["type"] == "image":
with response_container.chat_message("user"):
response_container.image(str(msg.content), caption="AI generated image")
else:
response_container.chat_message("user").write(msg.content)
elif getattr(msg, "response_metadata", None) and msg.response_metadata["type"] == "image":
with response_container.chat_message("assistant"):
response_container.write("Describe what you can see in the following image: ")
response_container.image(msg.content, caption="AI generated image")
else:
response_container.chat_message("assistant").write(msg.content)
placeholder_input = st.empty()
if st.session_state.first_run:
placeholder_input.empty()
col2_voice.empty()
else:
input_text = placeholder_input.chat_input(disabled=st.session_state.first_run) or None
logging.debug(f"Input text: {input_text}")
with col2_voice:
input_voice = whisper_stt(language="en", n_max_retry=config["N_MAX_RETRY"])
logging.debug(f"Input voice: {input_voice}")
input_prompt = input_text or input_voice
if input_prompt := input_text or input_voice:
logging.info(f"Processing input: {input_prompt}")
match st.session_state["exam_type"]:
case "qa":
response = model_eval.predict(input_prompt, st.session_state["question"][-1].content)
logging.info(f"QA model response: {response}")
case "img_desc":
response = model_eval.predict(input_prompt, st.session_state["image_url"])
logging.info(f"Image description model response: {response}")
response_str = convert_json_to_str(response)
with response_container:
st.session_state.messages.append(HumanMessage(content=input_prompt))
logging.debug(f"Adding user message to session state: {input_prompt}")
response_container.chat_message("user").write(input_prompt)
st.session_state.messages.append(AIMessage(content=response_str))
logging.debug(f"Adding AI message to session state: {response_str}")
response_container.chat_message("assistant").write(response_str)
if st.session_state["exam_type"] == "qa":
new_question = generator.generate()
logging.info(f"Generated new question: {new_question}")
st.session_state["question"].append(AIMessage(content=new_question))
st.session_state.messages.append(AIMessage(content=new_question))
response_container.chat_message("assistant").write(new_question)
elif st.session_state["exam_type"] == "img_desc":
new_image = generator.generate()
st.session_state["image_url"] = new_image
st.session_state.messages.append(AIMessage(content=new_image))
logging.info(f"Generated new image URL: {st.session_state['image_url']}")
response_container.chat_message("assistant").write("Describe what you can see in the following image: ")
with response_container.chat_message("assistant"):
response_container.image(st.session_state["image_url"], caption="AI generated image")
|