Spaces:
Running
Running
Itaykhealth
commited on
Commit
β’
ea9ac36
1
Parent(s):
4a6d668
Add medAlpaca and change to ICL
Browse files- app.py +6 -5
- data/{GPT-3.5+PE+RAG β GPT-3.5+ICL+RAG}/dataset_statistics.csv +0 -0
- data/{GPT-3.5+PE+RAG β GPT-3.5+ICL+RAG}/params.json +0 -0
- data/{GPT-3.5+PE+RAG β GPT-3.5+ICL+RAG}/prompt.txt +0 -0
- data/{GPT-3.5+PE+RAG β GPT-3.5+ICL+RAG}/reasoning.csv +0 -0
- data/{GPT-3.5+PE β GPT-3.5+ICL}/dataset_statistics.csv +0 -0
- data/{GPT-3.5+PE β GPT-3.5+ICL}/params.json +0 -0
- data/{GPT-3.5+PE β GPT-3.5+ICL}/prompt.txt +0 -0
- data/{GPT-3.5+PE β GPT-3.5+ICL}/reasoning.csv +0 -0
- data/{GPT-4+PE+RAG β GPT-4+ICL+RAG}/dataset_statistics.csv +0 -0
- data/{GPT-4+PE+RAG β GPT-4+ICL+RAG}/params.json +0 -0
- data/{GPT-4+PE+RAG β GPT-4+ICL+RAG}/reasoning.csv +0 -0
- data/{GPT-4+PE β GPT-4+ICL}/dataset_statistics.csv +0 -0
- data/{GPT-4+PE β GPT-4+ICL}/params.json +0 -0
- data/{GPT-4+PE β GPT-4+ICL}/prompt.txt +0 -0
- data/{GPT-4+PE β GPT-4+ICL}/reasoning.csv +0 -0
- data/MedAlpaca/dataset_statistics.csv +0 -0
- data/MedAlpaca/params.json +20 -0
- data/MedAlpaca/prompt.txt +9 -0
- data/MedAlpaca/reasoning.csv +0 -0
app.py
CHANGED
@@ -20,13 +20,14 @@ VERSIONS = (
|
|
20 |
"Mistral",
|
21 |
"Bing",
|
22 |
"GPT-3.5",
|
23 |
-
"GPT-3.5+
|
24 |
"GPT-3.5+RAG",
|
25 |
-
"GPT-3.5+
|
26 |
"GPT-4",
|
27 |
-
"GPT-4+
|
28 |
"GPT-4+RAG",
|
29 |
-
"GPT-4+
|
|
|
30 |
)
|
31 |
|
32 |
|
@@ -87,7 +88,7 @@ def main():
|
|
87 |
st.set_page_config(
|
88 |
layout="wide", page_title="AskMed Evaluation", initial_sidebar_state="expanded"
|
89 |
)
|
90 |
-
st.write(""" #
|
91 |
current_versions = st.sidebar.multiselect("select_version", VERSIONS)
|
92 |
if len(current_versions) == 1:
|
93 |
current_version = current_versions[0]
|
|
|
20 |
"Mistral",
|
21 |
"Bing",
|
22 |
"GPT-3.5",
|
23 |
+
"GPT-3.5+ICL",
|
24 |
"GPT-3.5+RAG",
|
25 |
+
"GPT-3.5+ICL+RAG",
|
26 |
"GPT-4",
|
27 |
+
"GPT-4+ICL",
|
28 |
"GPT-4+RAG",
|
29 |
+
"GPT-4+ICL+RAG",
|
30 |
+
"MedAlpaca"
|
31 |
)
|
32 |
|
33 |
|
|
|
88 |
st.set_page_config(
|
89 |
layout="wide", page_title="AskMed Evaluation", initial_sidebar_state="expanded"
|
90 |
)
|
91 |
+
st.write(""" # K-QA Evaluation""")
|
92 |
current_versions = st.sidebar.multiselect("select_version", VERSIONS)
|
93 |
if len(current_versions) == 1:
|
94 |
current_version = current_versions[0]
|
data/{GPT-3.5+PE+RAG β GPT-3.5+ICL+RAG}/dataset_statistics.csv
RENAMED
File without changes
|
data/{GPT-3.5+PE+RAG β GPT-3.5+ICL+RAG}/params.json
RENAMED
File without changes
|
data/{GPT-3.5+PE+RAG β GPT-3.5+ICL+RAG}/prompt.txt
RENAMED
File without changes
|
data/{GPT-3.5+PE+RAG β GPT-3.5+ICL+RAG}/reasoning.csv
RENAMED
File without changes
|
data/{GPT-3.5+PE β GPT-3.5+ICL}/dataset_statistics.csv
RENAMED
File without changes
|
data/{GPT-3.5+PE β GPT-3.5+ICL}/params.json
RENAMED
File without changes
|
data/{GPT-3.5+PE β GPT-3.5+ICL}/prompt.txt
RENAMED
File without changes
|
data/{GPT-3.5+PE β GPT-3.5+ICL}/reasoning.csv
RENAMED
File without changes
|
data/{GPT-4+PE+RAG β GPT-4+ICL+RAG}/dataset_statistics.csv
RENAMED
File without changes
|
data/{GPT-4+PE+RAG β GPT-4+ICL+RAG}/params.json
RENAMED
File without changes
|
data/{GPT-4+PE+RAG β GPT-4+ICL+RAG}/reasoning.csv
RENAMED
File without changes
|
data/{GPT-4+PE β GPT-4+ICL}/dataset_statistics.csv
RENAMED
File without changes
|
data/{GPT-4+PE β GPT-4+ICL}/params.json
RENAMED
File without changes
|
data/{GPT-4+PE β GPT-4+ICL}/prompt.txt
RENAMED
File without changes
|
data/{GPT-4+PE β GPT-4+ICL}/reasoning.csv
RENAMED
File without changes
|
data/MedAlpaca/dataset_statistics.csv
ADDED
The diff for this file is too large to render.
See raw diff
|
|
data/MedAlpaca/params.json
ADDED
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"git_branch": "ask-k-exp",
|
3 |
+
"git_commit_hash": "cfa8b7b9fe596c2e1b154c48e5bbe76b83272841",
|
4 |
+
"skip_empty_respnes": true,
|
5 |
+
"llm_validator": "gpt-4-8k",
|
6 |
+
"generative_model": "med-llama",
|
7 |
+
"version": "medalpaca/medalpaca-7b",
|
8 |
+
"tasks": [
|
9 |
+
"is_entails",
|
10 |
+
"is_contradict"
|
11 |
+
],
|
12 |
+
"date": "2023-12-20T12:50:06.783489",
|
13 |
+
"recall_cols": "Must_have",
|
14 |
+
"precision_cols": [
|
15 |
+
"Must_have",
|
16 |
+
"Nice_to_have"
|
17 |
+
],
|
18 |
+
"count_question": 201,
|
19 |
+
"count_w_answer": 199
|
20 |
+
}
|
data/MedAlpaca/prompt.txt
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.,
|
2 |
+
|
3 |
+
### Instruction:
|
4 |
+
Answer the following question.
|
5 |
+
|
6 |
+
### Input:
|
7 |
+
{question}
|
8 |
+
|
9 |
+
### Response:
|
data/MedAlpaca/reasoning.csv
ADDED
The diff for this file is too large to render.
See raw diff
|
|