jacob-mathison/doczyai-pipelines
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Activity

find page no. using llm

Browse Source
This commit is contained in:
Mayank Aamseek
2024-04-06 17:38:53 +05:30
parent b461a6d342
commit 07070840bc
1 changed files with 11 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files
streamlit/interface_3.py
+11 -14
View File
@@ -123,7 +123,7 @@ if st.session_state.user_info['mail'] in user_list:
with field_row[0]:
st.write("**Field Group**")
with field_row[1]:
field_group = st.selectbox('Field Group',('Unique Key', 'Contract Related', 'Pricing Before Carveouts - I'
field_group = st.selectbox('Field Group',('Unique and Contract Related', 'Pricing Before Carveouts - I'
, 'Pricing Before Carveouts - II', 'Carveout Indicator, Code Type and Code #s - I'
, 'Carveout Indicator, Code Type and Code #s - II', 'Carveout Indicator, Code Type and Code #s - III'
, 'Optimize Carving Indic.', 'Carveout Method - I', 'Carveout Method - II', 'Provider'
@@ -131,10 +131,8 @@ if st.session_state.user_info['mail'] in user_list:
# priorty column will be relaced by group_id in snowflake db
if field_group == 'Unique Key':
fields = fields[fields['PRIORITY'] == 'A']
elif field_group == 'Contract Related':
fields = fields[fields['PRIORITY'] == 'C']
if field_group == 'Unique and Contract Related':
fields = fields[fields['PRIORITY'].isin(['A','C'])]
elif field_group == 'Pricing Before Carveouts - I':
fields = fields[fields['PRIORITY'] == 'B']
fields = np.array_split(fields, 2)[0]
@@ -326,7 +324,7 @@ if st.session_state.user_info['mail'] in user_list:
Question: {question}
Answer: Answer in JSON format: {{"""
parameters = {
"maxTokenCount":1280,
"maxTokenCount":2048,
"stopSequences":[],
"temperature":0,
"topP":0.9
@@ -346,7 +344,7 @@ if st.session_state.user_info['mail'] in user_list:
Answer: Answer in JSON format: {{"""
payload={
"prompt":"[INST]"+ prompt_data +"[/INST]",
"max_gen_len":1280,
"max_gen_len":2048,
"temperature":0.0,
"top_p":0.9
}
@@ -368,7 +366,7 @@ if st.session_state.user_info['mail'] in user_list:
Assistant: Answer in JSON format: {{"""
body = json.dumps(
{"prompt": anthropic.HUMAN_PROMPT + prompt_data + anthropic.AI_PROMPT,
"max_tokens_to_sample": 1280,
"max_tokens_to_sample": 2048,
"temperature":0.0,
"top_p":1,
"top_k":250,
@@ -571,8 +569,8 @@ if st.session_state.user_info['mail'] in user_list:
df, history, attempt, raw_response_text = run_llm(attempt, bucket, contract_list, llm_selected, field, field_values, history)
df_1 = df[['Contract Name','Contract ID', 'Actual Value Stored','New Extracted value','Confidence Level'
,'Snippet','Original Page Number', 'New Page Number', 'Revised Prompt', 'Result']]
# df.to_csv('results.csv', index=False)
# history.to_csv('history.csv', index=False)
df.to_csv('results.csv', index=False)
history.to_csv('history.csv', index=False)
# s3_client.upload_file('results.csv', bucket, key)
csv_buf = StringIO()
df_1.to_csv(csv_buf, header=True, index=False)
@@ -585,10 +583,9 @@ if st.session_state.user_info['mail'] in user_list:
df = df[['Contract Name','Contract ID', 'SF_DB_COL_NAME', 'Actual Value Stored','New Extracted value','Confidence Level'
,'Snippet','Original Page Number', 'New Page Number', 'Revised Prompt', 'Result']]
# df = pd.read_csv('results.csv')
# df['Result'] = df['Result'].astype('str')
# history = pd.read_csv('history.csv')
df = pd.read_csv('results.csv')
df['Result'] = df['Result'].astype('str')
history = pd.read_csv('history.csv')
st.dataframe(df)
st.dataframe(history)