diff --git a/README.md b/README.md
old mode 100644
new mode 100755
index 913de00581edbc7ba6478a568f9c21e89e09e59d..8e470f09733b99cf459d011359c0f0911e659d4d
--- a/README.md
+++ b/README.md
@@ -1,13 +1,17 @@
---
title: Self Improving Leaderboard
-emoji: 🦀
-colorFrom: purple
-colorTo: green
+emoji: 🔄
+colorFrom: green
+colorTo: indigo
sdk: gradio
-sdk_version: 4.40.0
+sdk_version: 4.36.0
app_file: app.py
-pinned: false
+pinned: true
license: apache-2.0
+duplicated_from: upstage/open-ko-llm-leaderboard
+fullWidth: true
+tags:
+ - leaderboard
---
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
diff --git a/app.py b/app.py
new file mode 100755
index 0000000000000000000000000000000000000000..0c3250565871e6c16d5e80310a91cda8f8ec9e37
--- /dev/null
+++ b/app.py
@@ -0,0 +1,400 @@
+import gradio as gr
+import pandas as pd
+from apscheduler.schedulers.background import BackgroundScheduler
+from huggingface_hub import snapshot_download
+from gradio_space_ci import configure_space_ci # FOR CI
+
+from src.display.about import (
+ EVALUATION_QUEUE_TEXT,
+ INTRODUCTION_TEXT,
+ LLM_BENCHMARKS_TEXT,
+ TITLE,
+)
+from src.display.css_html_js import custom_css
+from src.display.utils import (
+ BENCHMARK_COLS,
+ COLS,
+ EVAL_COLS,
+ EVAL_TYPES,
+ NUMERIC_INTERVALS,
+ TYPES,
+ AutoEvalColumn,
+ ModelType,
+ fields,
+ WeightType,
+ Precision
+)
+from src.envs import API, EVAL_REQUESTS_PATH, EVAL_RESULTS_PATH, H4_TOKEN, IS_PUBLIC, QUEUE_REPO, REPO_ID, RESULTS_REPO
+from src.populate import get_evaluation_queue_df, get_leaderboard_df
+from src.submission.submit import add_new_eval
+from src.tools.collections import update_collections
+from src.tools.plots import (
+ create_metric_plot_obj,
+ create_plot_df,
+ create_scores_df,
+)
+
+
+def restart_space():
+ API.restart_space(repo_id=REPO_ID, token=H4_TOKEN)
+
+try:
+ print(EVAL_REQUESTS_PATH)
+ snapshot_download(
+ repo_id=QUEUE_REPO, local_dir=EVAL_REQUESTS_PATH, repo_type="dataset", tqdm_class=None, etag_timeout=30
+ )
+except Exception:
+ restart_space()
+try:
+ print(EVAL_RESULTS_PATH)
+ snapshot_download(
+ repo_id=RESULTS_REPO, local_dir=EVAL_RESULTS_PATH, repo_type="dataset", tqdm_class=None, etag_timeout=30
+ )
+except Exception:
+ restart_space()
+
+
+_, original_df = get_leaderboard_df(EVAL_RESULTS_PATH, EVAL_REQUESTS_PATH, COLS, BENCHMARK_COLS)
+leaderboard_df = original_df.copy()
+
+(
+ finished_eval_queue_df,
+ running_eval_queue_df,
+ pending_eval_queue_df,
+ failed_eval_queue_df,
+) = get_evaluation_queue_df(EVAL_REQUESTS_PATH, EVAL_COLS)
+
+
+# Searching and filtering
+def update_table(
+ hidden_df: pd.DataFrame,
+ columns: list,
+ type_query: list,
+ precision_query: str,
+ size_query: list,
+ show_deleted: bool,
+ show_merges: bool,
+ show_flagged: bool,
+ query: str,
+):
+ filtered_df = filter_models(hidden_df, type_query, size_query, precision_query)
+ filtered_df = filter_queries(query, filtered_df)
+ df = select_columns(filtered_df, columns)
+ return df
+
+def quarter_update_table(
+ hidden_df: pd.DataFrame,
+ columns: list,
+ type_query: list,
+ precision_query: str,
+ size_query: list,
+ show_deleted: bool,
+ show_merges: bool,
+ show_flagged: bool,
+ query: str,
+):
+ filtered_df = filter_models(hidden_df, type_query, size_query, precision_query)
+ filtered_df = filter_queries(query, filtered_df)
+ df = quarter_select_columns(filtered_df, columns)
+ return df
+
+
+def load_query(request: gr.Request): # triggered only once at startup => read query parameter if it exists
+ query = request.query_params.get("query") or ""
+ return query, query # return one for the "search_bar", one for a hidden component that triggers a reload only if value has changed
+
+
+def search_table(df: pd.DataFrame, query: str) -> pd.DataFrame:
+ return df[(df[AutoEvalColumn.dummy.name].str.contains(query, case=False))]
+
+
+def select_columns(df: pd.DataFrame, columns: list) -> pd.DataFrame:
+ always_here_cols = [
+ AutoEvalColumn.model_type_symbol.name,
+ AutoEvalColumn.model.name,
+ ]
+ # We use COLS to maintain sorting
+ filtered_df = df[
+ always_here_cols + [c for c in COLS if c in df.columns and c in columns] + [AutoEvalColumn.dummy.name]
+ ]
+ return filtered_df
+
+
+def filter_queries(query: str, filtered_df: pd.DataFrame):
+ """Added by Abishek"""
+ final_df = []
+ if query != "":
+ queries = [q.strip() for q in query.split(";")]
+ for _q in queries:
+ _q = _q.strip()
+ if _q != "":
+ temp_filtered_df = search_table(filtered_df, _q)
+ if len(temp_filtered_df) > 0:
+ final_df.append(temp_filtered_df)
+ if len(final_df) > 0:
+ filtered_df = pd.concat(final_df)
+ filtered_df = filtered_df.drop_duplicates(
+ subset=[AutoEvalColumn.model.name, AutoEvalColumn.precision.name, AutoEvalColumn.revision.name]
+ )
+
+ return filtered_df
+
+
+def filter_models(
+ df: pd.DataFrame, type_query: list, size_query: list, precision_query: list
+) -> pd.DataFrame:
+
+ type_emoji = [t[0] for t in type_query]
+ df = df.loc[df[AutoEvalColumn.model_type_symbol.name].isin(type_emoji)]
+ df = df.loc[df[AutoEvalColumn.precision.name].isin(precision_query + ["None"])]
+ numeric_interval = pd.IntervalIndex(sorted([NUMERIC_INTERVALS[s] for s in size_query]))
+ params_column = pd.to_numeric(df[AutoEvalColumn.params.name], errors="coerce")
+ mask = params_column.apply(lambda x: any(numeric_interval.contains(x)))
+ df = df.loc[mask]
+
+ return df
+
+
+leaderboard_df = filter_models(leaderboard_df, [t.to_str(" : ") for t in ModelType], list(NUMERIC_INTERVALS.keys()), [i.value.name for i in Precision])
+
+print(leaderboard_df)
+
+demo = gr.Blocks(css=custom_css)
+with demo:
+ gr.HTML(TITLE)
+ gr.Markdown(INTRODUCTION_TEXT, elem_classes="markdown-text")
+
+ with gr.Tabs(elem_classes="tab-buttons") as tabs:
+ with gr.TabItem("🔄 Self-Improving Benchmark", elem_id="llm-benchmark-tab-table", id=0):
+ with gr.Row():
+ with gr.Column():
+ with gr.Row():
+ search_bar = gr.Textbox(
+ placeholder=" 🔍 Search for your model (separate multiple queries with `;`) and press ENTER...",
+ show_label=False,
+ elem_id="search-bar",
+ )
+ with gr.Row():
+ shown_columns = gr.CheckboxGroup(
+ choices=[
+ c.name
+ for c in fields(AutoEvalColumn)
+ if not c.hidden and not c.never_hidden and not c.dummy
+ ],
+ value=[
+ c.name
+ for c in fields(AutoEvalColumn)
+ if c.displayed_by_default and not c.hidden and not c.never_hidden
+ ],
+ label="Select columns to show",
+ elem_id="column-select",
+ interactive=True,
+ )
+
+ with gr.Column(min_width=320):
+ #with gr.Box(elem_id="box-filter"):
+ filter_columns_type = gr.CheckboxGroup(
+ label="Model types",
+ choices=[t.to_str() for t in ModelType],
+ value=[t.to_str() for t in ModelType],
+ interactive=True,
+ elem_id="filter-columns-type",
+ )
+ filter_columns_precision = gr.CheckboxGroup(
+ label="Precision",
+ choices=[i.value.name for i in Precision],
+ value=[i.value.name for i in Precision],
+ interactive=True,
+ elem_id="filter-columns-precision",
+ )
+ filter_columns_size = gr.CheckboxGroup(
+ label="Model sizes (in billions of parameters)",
+ choices=list(NUMERIC_INTERVALS.keys()),
+ value=list(NUMERIC_INTERVALS.keys()),
+ interactive=True,
+ elem_id="filter-columns-size",
+ )
+
+ leaderboard_table = gr.components.Dataframe(
+ value=leaderboard_df[
+ [c.name for c in fields(AutoEvalColumn) if c.never_hidden]
+ + shown_columns.value
+ + [AutoEvalColumn.dummy.name]
+ ],
+ headers=[c.name for c in fields(AutoEvalColumn) if c.never_hidden] + shown_columns.value,
+ datatype=TYPES,
+ elem_id="leaderboard-table",
+ interactive=False,
+ visible=True,
+ #column_widths=["2%", "33%"]
+ )
+
+ # Dummy leaderboard for handling the case when the user uses backspace key
+ hidden_leaderboard_table_for_search = gr.components.Dataframe(
+ value=original_df[COLS],
+ headers=COLS,
+ datatype=TYPES,
+ visible=False,
+ )
+ search_bar.submit(
+ update_table,
+ [
+ hidden_leaderboard_table_for_search,
+ shown_columns,
+ filter_columns_type,
+ filter_columns_precision,
+ filter_columns_size,
+ search_bar,
+ ],
+ leaderboard_table,
+ )
+
+ # Define a hidden component that will trigger a reload only if a query parameter has be set
+ hidden_search_bar = gr.Textbox(value="", visible=False)
+ hidden_search_bar.change(
+ update_table,
+ [
+ hidden_leaderboard_table_for_search,
+ shown_columns,
+ filter_columns_type,
+ filter_columns_precision,
+ filter_columns_size,
+ search_bar,
+ ],
+ leaderboard_table,
+ )
+ # Check query parameter once at startup and update search bar + hidden component
+ demo.load(load_query, inputs=[], outputs=[search_bar, hidden_search_bar])
+
+ for selector in [shown_columns, filter_columns_type, filter_columns_precision, filter_columns_size]:
+ selector.change(
+ update_table,
+ [
+ hidden_leaderboard_table_for_search,
+ shown_columns,
+ filter_columns_type,
+ filter_columns_precision,
+ filter_columns_size,
+ search_bar,
+ ],
+ leaderboard_table,
+ queue=True,
+ )
+
+ with gr.TabItem("📝 About", elem_id="llm-benchmark-tab-table", id=1):
+ gr.Markdown(LLM_BENCHMARKS_TEXT, elem_classes="markdown-text")
+
+ with gr.TabItem("🚀 Submit here! ", elem_id="llm-benchmark-tab-table", id=2):
+ with gr.Column():
+ with gr.Row():
+ gr.Markdown(EVALUATION_QUEUE_TEXT, elem_classes="markdown-text")
+
+ with gr.Column():
+ with gr.Accordion(
+ f"✅ Finished Evaluations ({len(finished_eval_queue_df)})",
+ open=False,
+ ):
+ with gr.Row():
+ finished_eval_table = gr.components.Dataframe(
+ value=finished_eval_queue_df,
+ headers=EVAL_COLS,
+ datatype=EVAL_TYPES,
+ row_count=5,
+ )
+ with gr.Accordion(
+ f"🔄 Running Evaluation Queue ({len(running_eval_queue_df)})",
+ open=False,
+ ):
+ with gr.Row():
+ running_eval_table = gr.components.Dataframe(
+ value=running_eval_queue_df,
+ headers=EVAL_COLS,
+ datatype=EVAL_TYPES,
+ row_count=5,
+ )
+
+ with gr.Accordion(
+ f"⏳ Pending Evaluation Queue ({len(pending_eval_queue_df)})",
+ open=False,
+ ):
+ with gr.Row():
+ pending_eval_table = gr.components.Dataframe(
+ value=pending_eval_queue_df,
+ headers=EVAL_COLS,
+ datatype=EVAL_TYPES,
+ row_count=5,
+ )
+ with gr.Accordion(
+ f"❌ Failed Evaluations ({len(failed_eval_queue_df)})",
+ open=False,
+ ):
+ with gr.Row():
+ pending_eval_table = gr.components.Dataframe(
+ value=failed_eval_queue_df,
+ headers=EVAL_COLS,
+ datatype=EVAL_TYPES,
+ row_count=5,
+ )
+ with gr.Row():
+ gr.Markdown("# ✉️✨ Submit your model here!", elem_classes="markdown-text")
+
+ with gr.Row():
+ with gr.Column():
+ model_name_textbox = gr.Textbox(label="Model name")
+ revision_name_textbox = gr.Textbox(label="Revision commit", placeholder="main")
+ private = gr.Checkbox(False, label="Private", visible=not IS_PUBLIC)
+ model_type = gr.Dropdown(
+ choices=[t.to_str(" : ") for t in ModelType if t != ModelType.Unknown],
+ label="Model type",
+ multiselect=False,
+ value=ModelType.IFT.to_str(" : "),
+ interactive=True,
+ )
+
+ with gr.Column():
+ precision = gr.Dropdown(
+ choices=[i.value.name for i in Precision if i != Precision.Unknown],
+ label="Precision",
+ multiselect=False,
+ value="float16",
+ interactive=True,
+ )
+ weight_type = gr.Dropdown(
+ choices=[i.value.name for i in WeightType],
+ label="Weights type",
+ multiselect=False,
+ value="Original",
+ interactive=True,
+ )
+ base_model_name_textbox = gr.Textbox(label="Base model (for delta or adapter weights)")
+
+ submit_button = gr.Button("Submit Evalulation!")
+ submission_result = gr.Markdown()
+ submit_button.click(
+ add_new_eval,
+ [
+ model_name_textbox,
+ base_model_name_textbox,
+ revision_name_textbox,
+ precision,
+ private,
+ weight_type,
+ model_type,
+ ],
+ submission_result,
+ )
+
+scheduler = BackgroundScheduler()
+scheduler.add_job(restart_space, "interval", seconds=1800)
+scheduler.start()
+
+# Both launches the space and its CI
+configure_space_ci(
+ demo.queue(default_concurrency_limit=40),
+ trusted_authors=[], # add manually trusted authors
+ private="True", # ephemeral spaces will have same visibility as the main space. Otherwise, set to `True` or `False` explicitly.
+ variables={}, # We overwrite HF_HOME as tmp CI spaces will have no cache
+ secrets=["HF_TOKEN", "H4_TOKEN"], # which secret do I want to copy from the main space? Can be a `List[str]`.
+ hardware=None, # "cpu-basic" by default. Otherwise set to "auto" to have same hardware as the main space or any valid string value.
+ storage=None, # no storage by default. Otherwise set to "auto" to have same storage as the main space or any valid string value.
+).launch()
\ No newline at end of file
diff --git a/eval-queue/.gitattributes b/eval-queue/.gitattributes
new file mode 100644
index 0000000000000000000000000000000000000000..28df5f900b358436f0267334b3e3e9af33f917ba
--- /dev/null
+++ b/eval-queue/.gitattributes
@@ -0,0 +1,55 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.lz4 filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+# Audio files - uncompressed
+*.pcm filter=lfs diff=lfs merge=lfs -text
+*.sam filter=lfs diff=lfs merge=lfs -text
+*.raw filter=lfs diff=lfs merge=lfs -text
+# Audio files - compressed
+*.aac filter=lfs diff=lfs merge=lfs -text
+*.flac filter=lfs diff=lfs merge=lfs -text
+*.mp3 filter=lfs diff=lfs merge=lfs -text
+*.ogg filter=lfs diff=lfs merge=lfs -text
+*.wav filter=lfs diff=lfs merge=lfs -text
+# Image files - uncompressed
+*.bmp filter=lfs diff=lfs merge=lfs -text
+*.gif filter=lfs diff=lfs merge=lfs -text
+*.png filter=lfs diff=lfs merge=lfs -text
+*.tiff filter=lfs diff=lfs merge=lfs -text
+# Image files - compressed
+*.jpg filter=lfs diff=lfs merge=lfs -text
+*.jpeg filter=lfs diff=lfs merge=lfs -text
+*.webp filter=lfs diff=lfs merge=lfs -text
diff --git a/eval-queue/01-ai/Yi-1.5-9B-32K_eval_request_False_float16_Original.json b/eval-queue/01-ai/Yi-1.5-9B-32K_eval_request_False_float16_Original.json
new file mode 100644
index 0000000000000000000000000000000000000000..0e4f2e315d739b98393c777b9a806f1ac5a840a9
--- /dev/null
+++ b/eval-queue/01-ai/Yi-1.5-9B-32K_eval_request_False_float16_Original.json
@@ -0,0 +1,14 @@
+{
+ "model": "01-ai/Yi-1.5-9B-32K",
+ "base_model": "",
+ "revision": "c0239dbc923b8a2b5ca849763bdd592d39c60850",
+ "private": false,
+ "precision": "float16",
+ "weight_type": "Original",
+ "status": "FINISHED",
+ "submitted_time": "2024-07-29T13:10:13Z",
+ "model_type": "\ud83d\udfe2 : pretrained",
+ "likes": 18,
+ "params": 8.829,
+ "license": "apache-2.0"
+}
\ No newline at end of file
diff --git a/eval-queue/BioMistral/BioMistral-7B_eval_request_False_float16_Original.json b/eval-queue/BioMistral/BioMistral-7B_eval_request_False_float16_Original.json
new file mode 100644
index 0000000000000000000000000000000000000000..4f1d37cdaa1a844889dc0b86f1d37987eac18d17
--- /dev/null
+++ b/eval-queue/BioMistral/BioMistral-7B_eval_request_False_float16_Original.json
@@ -0,0 +1,15 @@
+{
+ "model": "BioMistral/BioMistral-7B",
+ "base_model": "",
+ "revision": "main",
+ "private": false,
+ "precision": "float16",
+ "weight_type": "Original",
+ "status": "FINISHED",
+ "submitted_time": "2024-05-30 01:33:58",
+ "model_type": "\u2b55 : instruction-tuned",
+ "job_id": "2031",
+ "params": 7.0,
+ "likes": 354,
+ "license": "apache-2.0"
+}
\ No newline at end of file
diff --git a/eval-queue/EleutherAI/polyglot-ko-1.3b_eval_request_False_float16_Original.json b/eval-queue/EleutherAI/polyglot-ko-1.3b_eval_request_False_float16_Original.json
new file mode 100644
index 0000000000000000000000000000000000000000..e9bdc0c2fb5a5f1325934e777c01ca6674591358
--- /dev/null
+++ b/eval-queue/EleutherAI/polyglot-ko-1.3b_eval_request_False_float16_Original.json
@@ -0,0 +1,14 @@
+{
+ "model": "EleutherAI/polyglot-ko-1.3b",
+ "base_model": "",
+ "revision": "main",
+ "private": false,
+ "precision": "float16",
+ "weight_type": "Original",
+ "status": "FINISHED",
+ "submitted_time": "2024-07-25T11:04:40Z",
+ "model_type": "\ud83d\udfe2 : pretrained",
+ "likes": 71,
+ "params": 1.432,
+ "license": "apache-2.0"
+}
\ No newline at end of file
diff --git a/eval-queue/HuggingFaceH4/zephyr-7b-beta_eval_request_False_float16_Original.json b/eval-queue/HuggingFaceH4/zephyr-7b-beta_eval_request_False_float16_Original.json
new file mode 100644
index 0000000000000000000000000000000000000000..1c06f98202b82aab66fd3f351c7eb3849b9b9d32
--- /dev/null
+++ b/eval-queue/HuggingFaceH4/zephyr-7b-beta_eval_request_False_float16_Original.json
@@ -0,0 +1,15 @@
+{
+ "model": "HuggingFaceH4/zephyr-7b-beta",
+ "base_model": "",
+ "revision": "main",
+ "private": false,
+ "precision": "float16",
+ "weight_type": "Original",
+ "status": "FINISHED",
+ "submitted_time": "2023-11-01 04:21:47",
+ "model_type": "\u2b55 : instruction-tuned",
+ "job_id": "401",
+ "params": 7.242,
+ "likes": 1162,
+ "license": "mit"
+}
\ No newline at end of file
diff --git a/eval-queue/README.md b/eval-queue/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..7b95401dc46245ac339fc25059d4a56d90b4cde5
--- /dev/null
+++ b/eval-queue/README.md
@@ -0,0 +1,3 @@
+---
+license: apache-2.0
+---
diff --git a/eval-queue/nlpai-lab/KULLM3_eval_request_False_float16_Original.json b/eval-queue/nlpai-lab/KULLM3_eval_request_False_float16_Original.json
new file mode 100644
index 0000000000000000000000000000000000000000..c04269a4cb1f73140fcb3ef175dd8a784955b71a
--- /dev/null
+++ b/eval-queue/nlpai-lab/KULLM3_eval_request_False_float16_Original.json
@@ -0,0 +1,15 @@
+{
+ "model": "nlpai-lab/KULLM3",
+ "base_model": "",
+ "revision": "main",
+ "private": false,
+ "precision": "float16",
+ "weight_type": "Original",
+ "status": "FINISHED",
+ "submitted_time": "2024-04-08 05:16:47",
+ "model_type": "\u2b55 : instruction-tuned",
+ "job_id": "1751",
+ "params": 10.732000350952148,
+ "likes": 13,
+ "license": "cc-by-nc-4.0"
+}
diff --git a/eval-queue/x2bee/POLAR-14B-DPO-v1.3_eval_request_False_float16_Original.json b/eval-queue/x2bee/POLAR-14B-DPO-v1.3_eval_request_False_float16_Original.json
new file mode 100644
index 0000000000000000000000000000000000000000..2ef469f131f52aba9274a2681184555924cc23ef
--- /dev/null
+++ b/eval-queue/x2bee/POLAR-14B-DPO-v1.3_eval_request_False_float16_Original.json
@@ -0,0 +1,15 @@
+{
+ "model": "x2bee/POLAR-14B-DPO-v1.3",
+ "base_model": "",
+ "revision": "main",
+ "private": false,
+ "precision": "float16",
+ "weight_type": "Original",
+ "status": "FINISHED",
+ "submitted_time": "2024-05-23 11:59:50",
+ "model_type": "\u2b55 : instruction-tuned",
+ "job_id": "1987",
+ "params": 14.220999717712402,
+ "likes": 0,
+ "license": "apache-2.0"
+}
\ No newline at end of file
diff --git a/eval-queue/x2bee/POLAR-14B-DPO-v1.4_eval_request_False_float16_Original.json b/eval-queue/x2bee/POLAR-14B-DPO-v1.4_eval_request_False_float16_Original.json
new file mode 100644
index 0000000000000000000000000000000000000000..7c9f43b54371ecc812142ccabeb144da0cb8e746
--- /dev/null
+++ b/eval-queue/x2bee/POLAR-14B-DPO-v1.4_eval_request_False_float16_Original.json
@@ -0,0 +1,15 @@
+{
+ "model": "x2bee/POLAR-14B-DPO-v1.4",
+ "base_model": "",
+ "revision": "main",
+ "private": false,
+ "precision": "float16",
+ "weight_type": "Original",
+ "status": "FINISHED",
+ "submitted_time": "2024-05-27 15:02:47",
+ "model_type": "\u2b55 : instruction-tuned",
+ "job_id": "2004",
+ "params": 14.220999717712402,
+ "likes": 0,
+ "license": "apache-2.0"
+}
\ No newline at end of file
diff --git a/eval-queue/x2bee/POLAR-14B-HES-DPO-v1.5_eval_request_False_float16_Original.json b/eval-queue/x2bee/POLAR-14B-HES-DPO-v1.5_eval_request_False_float16_Original.json
new file mode 100644
index 0000000000000000000000000000000000000000..1ab7a7823d68a62943fb46fe594bc2af66aea731
--- /dev/null
+++ b/eval-queue/x2bee/POLAR-14B-HES-DPO-v1.5_eval_request_False_float16_Original.json
@@ -0,0 +1,15 @@
+{
+ "model": "x2bee/POLAR-14B-HES-DPO-v1.5",
+ "base_model": "",
+ "revision": "main",
+ "private": false,
+ "precision": "float16",
+ "weight_type": "Original",
+ "status": "FINISHED",
+ "submitted_time": "2024-05-29 23:53:33",
+ "model_type": "\u2b55 : instruction-tuned",
+ "job_id": "2029",
+ "params": 14.220999717712402,
+ "likes": 0,
+ "license": "apache-2.0"
+}
\ No newline at end of file
diff --git a/eval-queue/x2bee/POLAR-14B-SON-SFT-v0.1_eval_request_False_float16_Original.json b/eval-queue/x2bee/POLAR-14B-SON-SFT-v0.1_eval_request_False_float16_Original.json
new file mode 100644
index 0000000000000000000000000000000000000000..e05dcf514240d93da66d455500e83888cb398899
--- /dev/null
+++ b/eval-queue/x2bee/POLAR-14B-SON-SFT-v0.1_eval_request_False_float16_Original.json
@@ -0,0 +1,15 @@
+{
+ "model": "x2bee/POLAR-14B-SON-SFT-v0.1",
+ "base_model": "x2bee/POLAR-14B-v0.2",
+ "revision": "main",
+ "private": false,
+ "precision": "float16",
+ "weight_type": "Original",
+ "status": "FINISHED",
+ "submitted_time": "2024-05-27 13:52:58",
+ "model_type": "\u2b55 : instruction-tuned",
+ "job_id": "2003",
+ "params": 14.220999717712402,
+ "likes": 0,
+ "license": "apache-2.0"
+}
\ No newline at end of file
diff --git a/eval-queue/x2bee/POLAR-14B-v0.2_eval_request_False_float16_Original.json b/eval-queue/x2bee/POLAR-14B-v0.2_eval_request_False_float16_Original.json
new file mode 100644
index 0000000000000000000000000000000000000000..4bd6f0ba567b6ddb914195db30cb389d37b987e1
--- /dev/null
+++ b/eval-queue/x2bee/POLAR-14B-v0.2_eval_request_False_float16_Original.json
@@ -0,0 +1,15 @@
+{
+ "model": "x2bee/POLAR-14B-v0.2",
+ "base_model": "",
+ "revision": "main",
+ "private": false,
+ "precision": "float16",
+ "weight_type": "Original",
+ "status": "FINISHED",
+ "submitted_time": "2024-05-02 00:34:33",
+ "model_type": "\ud83d\udfe2 : pretrained",
+ "job_id": "1874",
+ "params": 14.220999717712402,
+ "likes": 0,
+ "license": "apache-2.0"
+}
\ No newline at end of file
diff --git a/eval-queue/x2bee/POLAR-14B-v0.5_eval_request_False_float16_Original.json b/eval-queue/x2bee/POLAR-14B-v0.5_eval_request_False_float16_Original.json
new file mode 100644
index 0000000000000000000000000000000000000000..4fae812cffc7b05a196cb62a7d3f7bb5d6c1aac8
--- /dev/null
+++ b/eval-queue/x2bee/POLAR-14B-v0.5_eval_request_False_float16_Original.json
@@ -0,0 +1,15 @@
+{
+ "model": "x2bee/POLAR-14B-v0.5",
+ "base_model": "",
+ "revision": "main",
+ "private": false,
+ "precision": "float16",
+ "weight_type": "Original",
+ "status": "FINISHED",
+ "submitted_time": "2024-06-05 00:49:59",
+ "model_type": "\ud83d\udfe2 : pretrained",
+ "job_id": "2041",
+ "params": 14.220999717712402,
+ "likes": 0,
+ "license": "apache-2.0"
+}
\ No newline at end of file
diff --git a/eval-results/.gitattributes b/eval-results/.gitattributes
new file mode 100644
index 0000000000000000000000000000000000000000..28df5f900b358436f0267334b3e3e9af33f917ba
--- /dev/null
+++ b/eval-results/.gitattributes
@@ -0,0 +1,55 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.lz4 filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+# Audio files - uncompressed
+*.pcm filter=lfs diff=lfs merge=lfs -text
+*.sam filter=lfs diff=lfs merge=lfs -text
+*.raw filter=lfs diff=lfs merge=lfs -text
+# Audio files - compressed
+*.aac filter=lfs diff=lfs merge=lfs -text
+*.flac filter=lfs diff=lfs merge=lfs -text
+*.mp3 filter=lfs diff=lfs merge=lfs -text
+*.ogg filter=lfs diff=lfs merge=lfs -text
+*.wav filter=lfs diff=lfs merge=lfs -text
+# Image files - uncompressed
+*.bmp filter=lfs diff=lfs merge=lfs -text
+*.gif filter=lfs diff=lfs merge=lfs -text
+*.png filter=lfs diff=lfs merge=lfs -text
+*.tiff filter=lfs diff=lfs merge=lfs -text
+# Image files - compressed
+*.jpg filter=lfs diff=lfs merge=lfs -text
+*.jpeg filter=lfs diff=lfs merge=lfs -text
+*.webp filter=lfs diff=lfs merge=lfs -text
diff --git a/eval-results/01-ai/Yi-1.5-9B-32K/result_2024_07_30 20:36:30.json b/eval-results/01-ai/Yi-1.5-9B-32K/result_2024_07_30 20:36:30.json
new file mode 100644
index 0000000000000000000000000000000000000000..f01f8a60f89596f15703a48706a96da80d000f73
--- /dev/null
+++ b/eval-results/01-ai/Yi-1.5-9B-32K/result_2024_07_30 20:36:30.json
@@ -0,0 +1,450 @@
+{
+ "results": {
+ "daily": {
+ "daily": 7
+ },
+ "quarterly": {
+ "quarterly": 7
+ },
+ "harness|arc_challenge|25": {
+ "acc": 0.29948805460750855,
+ "acc_stderr": 0.013385021637313567,
+ "acc_norm": 0.3506825938566553,
+ "acc_norm_stderr": 0.013944635930726089
+ },
+ "harness|hellaswag|10": {
+ "acc": 0.3333001394144593,
+ "acc_stderr": 0.004704293898729902,
+ "acc_norm": 0.4137621987651862,
+ "acc_norm_stderr": 0.004915003499517831
+ },
+ "harness|mmlu_world_religions|5": {
+ "acc": 0.47953216374269003,
+ "acc_stderr": 0.0383161053282193,
+ "acc_norm": 0.47953216374269003,
+ "acc_norm_stderr": 0.0383161053282193
+ },
+ "harness|mmlu_management|5": {
+ "acc": 0.5631067961165048,
+ "acc_stderr": 0.049111471073657764,
+ "acc_norm": 0.5631067961165048,
+ "acc_norm_stderr": 0.049111471073657764
+ },
+ "harness|mmlu_miscellaneous|5": {
+ "acc": 0.47509578544061304,
+ "acc_stderr": 0.01785777070490102,
+ "acc_norm": 0.47509578544061304,
+ "acc_norm_stderr": 0.01785777070490102
+ },
+ "harness|mmlu_anatomy|5": {
+ "acc": 0.28888888888888886,
+ "acc_stderr": 0.0391545063041425,
+ "acc_norm": 0.28888888888888886,
+ "acc_norm_stderr": 0.0391545063041425
+ },
+ "harness|mmlu_abstract_algebra|5": {
+ "acc": 0.31,
+ "acc_stderr": 0.04648231987117316,
+ "acc_norm": 0.31,
+ "acc_norm_stderr": 0.04648231987117316
+ },
+ "harness|mmlu_conceptual_physics|5": {
+ "acc": 0.46808510638297873,
+ "acc_stderr": 0.03261936918467382,
+ "acc_norm": 0.46808510638297873,
+ "acc_norm_stderr": 0.03261936918467382
+ },
+ "harness|mmlu_virology|5": {
+ "acc": 0.45180722891566266,
+ "acc_stderr": 0.03874371556587953,
+ "acc_norm": 0.45180722891566266,
+ "acc_norm_stderr": 0.03874371556587953
+ },
+ "harness|mmlu_philosophy|5": {
+ "acc": 0.47266881028938906,
+ "acc_stderr": 0.028355633568328188,
+ "acc_norm": 0.47266881028938906,
+ "acc_norm_stderr": 0.028355633568328188
+ },
+ "harness|mmlu_human_aging|5": {
+ "acc": 0.45739910313901344,
+ "acc_stderr": 0.033435777055830646,
+ "acc_norm": 0.45739910313901344,
+ "acc_norm_stderr": 0.033435777055830646
+ },
+ "harness|mmlu_human_sexuality|5": {
+ "acc": 0.5267175572519084,
+ "acc_stderr": 0.04379024936553894,
+ "acc_norm": 0.5267175572519084,
+ "acc_norm_stderr": 0.04379024936553894
+ },
+ "harness|mmlu_medical_genetics|5": {
+ "acc": 0.39,
+ "acc_stderr": 0.04902071300001975,
+ "acc_norm": 0.39,
+ "acc_norm_stderr": 0.04902071300001975
+ },
+ "harness|mmlu_high_school_geography|5": {
+ "acc": 0.5555555555555556,
+ "acc_stderr": 0.035402943770953675,
+ "acc_norm": 0.5555555555555556,
+ "acc_norm_stderr": 0.035402943770953675
+ },
+ "harness|mmlu_electrical_engineering|5": {
+ "acc": 0.5724137931034483,
+ "acc_stderr": 0.04122737111370332,
+ "acc_norm": 0.5724137931034483,
+ "acc_norm_stderr": 0.04122737111370332
+ },
+ "harness|mmlu_college_physics|5": {
+ "acc": 0.3137254901960784,
+ "acc_stderr": 0.04617034827006716,
+ "acc_norm": 0.3137254901960784,
+ "acc_norm_stderr": 0.04617034827006716
+ },
+ "harness|mmlu_high_school_microeconomics|5": {
+ "acc": 0.5,
+ "acc_stderr": 0.032478490123081544,
+ "acc_norm": 0.5,
+ "acc_norm_stderr": 0.032478490123081544
+ },
+ "harness|mmlu_high_school_macroeconomics|5": {
+ "acc": 0.47692307692307695,
+ "acc_stderr": 0.025323990861736125,
+ "acc_norm": 0.47692307692307695,
+ "acc_norm_stderr": 0.025323990861736125
+ },
+ "harness|mmlu_computer_security|5": {
+ "acc": 0.53,
+ "acc_stderr": 0.05016135580465919,
+ "acc_norm": 0.53,
+ "acc_norm_stderr": 0.05016135580465919
+ },
+ "harness|mmlu_global_facts|5": {
+ "acc": 0.37,
+ "acc_stderr": 0.048523658709391,
+ "acc_norm": 0.37,
+ "acc_norm_stderr": 0.048523658709391
+ },
+ "harness|mmlu_jurisprudence|5": {
+ "acc": 0.5740740740740741,
+ "acc_stderr": 0.047803436269367894,
+ "acc_norm": 0.5740740740740741,
+ "acc_norm_stderr": 0.047803436269367894
+ },
+ "harness|mmlu_high_school_chemistry|5": {
+ "acc": 0.4187192118226601,
+ "acc_stderr": 0.03471192860518468,
+ "acc_norm": 0.4187192118226601,
+ "acc_norm_stderr": 0.03471192860518468
+ },
+ "harness|mmlu_high_school_biology|5": {
+ "acc": 0.47419354838709676,
+ "acc_stderr": 0.02840609505765332,
+ "acc_norm": 0.47419354838709676,
+ "acc_norm_stderr": 0.02840609505765332
+ },
+ "harness|mmlu_marketing|5": {
+ "acc": 0.6752136752136753,
+ "acc_stderr": 0.03067902276549883,
+ "acc_norm": 0.6752136752136753,
+ "acc_norm_stderr": 0.03067902276549883
+ },
+ "harness|mmlu_clinical_knowledge|5": {
+ "acc": 0.44150943396226416,
+ "acc_stderr": 0.030561590426731833,
+ "acc_norm": 0.44150943396226416,
+ "acc_norm_stderr": 0.030561590426731833
+ },
+ "harness|mmlu_public_relations|5": {
+ "acc": 0.4727272727272727,
+ "acc_stderr": 0.04782001791380063,
+ "acc_norm": 0.4727272727272727,
+ "acc_norm_stderr": 0.04782001791380063
+ },
+ "harness|mmlu_high_school_mathematics|5": {
+ "acc": 0.4185185185185185,
+ "acc_stderr": 0.030078013075022066,
+ "acc_norm": 0.4185185185185185,
+ "acc_norm_stderr": 0.030078013075022066
+ },
+ "harness|mmlu_high_school_physics|5": {
+ "acc": 0.304635761589404,
+ "acc_stderr": 0.03757949922943343,
+ "acc_norm": 0.304635761589404,
+ "acc_norm_stderr": 0.03757949922943343
+ },
+ "harness|mmlu_sociology|5": {
+ "acc": 0.6069651741293532,
+ "acc_stderr": 0.0345368246603156,
+ "acc_norm": 0.6069651741293532,
+ "acc_norm_stderr": 0.0345368246603156
+ },
+ "harness|mmlu_college_medicine|5": {
+ "acc": 0.4046242774566474,
+ "acc_stderr": 0.03742461193887248,
+ "acc_norm": 0.4046242774566474,
+ "acc_norm_stderr": 0.03742461193887248
+ },
+ "harness|mmlu_elementary_mathematics|5": {
+ "acc": 0.5476190476190477,
+ "acc_stderr": 0.02563425811555495,
+ "acc_norm": 0.5476190476190477,
+ "acc_norm_stderr": 0.02563425811555495
+ },
+ "harness|mmlu_college_biology|5": {
+ "acc": 0.3472222222222222,
+ "acc_stderr": 0.039812405437178615,
+ "acc_norm": 0.3472222222222222,
+ "acc_norm_stderr": 0.039812405437178615
+ },
+ "harness|mmlu_college_chemistry|5": {
+ "acc": 0.33,
+ "acc_stderr": 0.04725815626252605,
+ "acc_norm": 0.33,
+ "acc_norm_stderr": 0.04725815626252605
+ },
+ "harness|mmlu_us_foreign_policy|5": {
+ "acc": 0.57,
+ "acc_stderr": 0.04975698519562426,
+ "acc_norm": 0.57,
+ "acc_norm_stderr": 0.04975698519562426
+ },
+ "harness|mmlu_moral_disputes|5": {
+ "acc": 0.49710982658959535,
+ "acc_stderr": 0.026918645383239015,
+ "acc_norm": 0.49710982658959535,
+ "acc_norm_stderr": 0.026918645383239015
+ },
+ "harness|mmlu_logical_fallacies|5": {
+ "acc": 0.5276073619631901,
+ "acc_stderr": 0.03922378290610991,
+ "acc_norm": 0.5276073619631901,
+ "acc_norm_stderr": 0.03922378290610991
+ },
+ "harness|mmlu_prehistory|5": {
+ "acc": 0.49691358024691357,
+ "acc_stderr": 0.027820214158594377,
+ "acc_norm": 0.49691358024691357,
+ "acc_norm_stderr": 0.027820214158594377
+ },
+ "harness|mmlu_college_mathematics|5": {
+ "acc": 0.45,
+ "acc_stderr": 0.05,
+ "acc_norm": 0.45,
+ "acc_norm_stderr": 0.05
+ },
+ "harness|mmlu_high_school_government_and_politics|5": {
+ "acc": 0.49222797927461137,
+ "acc_stderr": 0.03608003225569654,
+ "acc_norm": 0.49222797927461137,
+ "acc_norm_stderr": 0.03608003225569654
+ },
+ "harness|mmlu_econometrics|5": {
+ "acc": 0.41228070175438597,
+ "acc_stderr": 0.046306532033665956,
+ "acc_norm": 0.41228070175438597,
+ "acc_norm_stderr": 0.046306532033665956
+ },
+ "harness|mmlu_high_school_psychology|5": {
+ "acc": 0.5027522935779817,
+ "acc_stderr": 0.02143699835976532,
+ "acc_norm": 0.5027522935779817,
+ "acc_norm_stderr": 0.02143699835976532
+ },
+ "harness|mmlu_formal_logic|5": {
+ "acc": 0.40476190476190477,
+ "acc_stderr": 0.04390259265377561,
+ "acc_norm": 0.40476190476190477,
+ "acc_norm_stderr": 0.04390259265377561
+ },
+ "harness|mmlu_nutrition|5": {
+ "acc": 0.49019607843137253,
+ "acc_stderr": 0.028624412550167958,
+ "acc_norm": 0.49019607843137253,
+ "acc_norm_stderr": 0.028624412550167958
+ },
+ "harness|mmlu_business_ethics|5": {
+ "acc": 0.5,
+ "acc_stderr": 0.050251890762960605,
+ "acc_norm": 0.5,
+ "acc_norm_stderr": 0.050251890762960605
+ },
+ "harness|mmlu_international_law|5": {
+ "acc": 0.7355371900826446,
+ "acc_stderr": 0.04026187527591205,
+ "acc_norm": 0.7355371900826446,
+ "acc_norm_stderr": 0.04026187527591205
+ },
+ "harness|mmlu_astronomy|5": {
+ "acc": 0.45394736842105265,
+ "acc_stderr": 0.04051646342874142,
+ "acc_norm": 0.45394736842105265,
+ "acc_norm_stderr": 0.04051646342874142
+ },
+ "harness|mmlu_professional_psychology|5": {
+ "acc": 0.39705882352941174,
+ "acc_stderr": 0.019794488900024113,
+ "acc_norm": 0.39705882352941174,
+ "acc_norm_stderr": 0.019794488900024113
+ },
+ "harness|mmlu_professional_accounting|5": {
+ "acc": 0.40070921985815605,
+ "acc_stderr": 0.029233465745573086,
+ "acc_norm": 0.40070921985815605,
+ "acc_norm_stderr": 0.029233465745573086
+ },
+ "harness|mmlu_machine_learning|5": {
+ "acc": 0.39285714285714285,
+ "acc_stderr": 0.04635550135609976,
+ "acc_norm": 0.39285714285714285,
+ "acc_norm_stderr": 0.04635550135609976
+ },
+ "harness|mmlu_high_school_statistics|5": {
+ "acc": 0.4675925925925926,
+ "acc_stderr": 0.034028015813589656,
+ "acc_norm": 0.4675925925925926,
+ "acc_norm_stderr": 0.034028015813589656
+ },
+ "harness|mmlu_moral_scenarios|5": {
+ "acc": 0.3329608938547486,
+ "acc_stderr": 0.015761716178397552,
+ "acc_norm": 0.3329608938547486,
+ "acc_norm_stderr": 0.015761716178397552
+ },
+ "harness|mmlu_college_computer_science|5": {
+ "acc": 0.43,
+ "acc_stderr": 0.049756985195624284,
+ "acc_norm": 0.43,
+ "acc_norm_stderr": 0.049756985195624284
+ },
+ "harness|mmlu_high_school_computer_science|5": {
+ "acc": 0.76,
+ "acc_stderr": 0.042923469599092816,
+ "acc_norm": 0.76,
+ "acc_norm_stderr": 0.042923469599092816
+ },
+ "harness|mmlu_professional_medicine|5": {
+ "acc": 0.35294117647058826,
+ "acc_stderr": 0.029029422815681404,
+ "acc_norm": 0.35294117647058826,
+ "acc_norm_stderr": 0.029029422815681404
+ },
+ "harness|mmlu_security_studies|5": {
+ "acc": 0.6163265306122448,
+ "acc_stderr": 0.031130880396235943,
+ "acc_norm": 0.6163265306122448,
+ "acc_norm_stderr": 0.031130880396235943
+ },
+ "harness|mmlu_high_school_world_history|5": {
+ "acc": 0.5654008438818565,
+ "acc_stderr": 0.03226759995510145,
+ "acc_norm": 0.5654008438818565,
+ "acc_norm_stderr": 0.03226759995510145
+ },
+ "harness|mmlu_professional_law|5": {
+ "acc": 0.36571056062581486,
+ "acc_stderr": 0.012301028188840567,
+ "acc_norm": 0.36571056062581486,
+ "acc_norm_stderr": 0.012301028188840567
+ },
+ "harness|mmlu_high_school_us_history|5": {
+ "acc": 0.4852941176470588,
+ "acc_stderr": 0.03507793834791324,
+ "acc_norm": 0.4852941176470588,
+ "acc_norm_stderr": 0.03507793834791324
+ },
+ "harness|mmlu_high_school_european_history|5": {
+ "acc": 0.5151515151515151,
+ "acc_stderr": 0.03902551007374448,
+ "acc_norm": 0.5151515151515151,
+ "acc_norm_stderr": 0.03902551007374448
+ },
+ "harness|truthfulqa_mc|0": {
+ "mc1": 0.2937576499388005,
+ "mc1_stderr": 0.015945068581236614,
+ "mc2": 0.4670848140389129,
+ "mc2_stderr": 0.01585178282587417
+ },
+ "harness|commongen_v2|2": {
+ "acc": 0.47107438016528924,
+ "acc_stderr": 0.017161563949916348,
+ "acc_norm": 0.5171192443919717,
+ "acc_norm_stderr": 0.017180275246085626
+ }
+ },
+ "versions": {
+ "all": 0,
+ "harness|arc_challenge|25": 0,
+ "harness|hellaswag|10": 0,
+ "harness|mmlu_world_religions|5": 1,
+ "harness|mmlu_management|5": 1,
+ "harness|mmlu_miscellaneous|5": 1,
+ "harness|mmlu_anatomy|5": 1,
+ "harness|mmlu_abstract_algebra|5": 1,
+ "harness|mmlu_conceptual_physics|5": 1,
+ "harness|mmlu_virology|5": 1,
+ "harness|mmlu_philosophy|5": 1,
+ "harness|mmlu_human_aging|5": 1,
+ "harness|mmlu_human_sexuality|5": 1,
+ "harness|mmlu_medical_genetics|5": 1,
+ "harness|mmlu_high_school_geography|5": 1,
+ "harness|mmlu_electrical_engineering|5": 1,
+ "harness|mmlu_college_physics|5": 1,
+ "harness|mmlu_high_school_microeconomics|5": 1,
+ "harness|mmlu_high_school_macroeconomics|5": 1,
+ "harness|mmlu_computer_security|5": 1,
+ "harness|mmlu_global_facts|5": 1,
+ "harness|mmlu_jurisprudence|5": 1,
+ "harness|mmlu_high_school_chemistry|5": 1,
+ "harness|mmlu_high_school_biology|5": 1,
+ "harness|mmlu_marketing|5": 1,
+ "harness|mmlu_clinical_knowledge|5": 1,
+ "harness|mmlu_public_relations|5": 1,
+ "harness|mmlu_high_school_mathematics|5": 1,
+ "harness|mmlu_high_school_physics|5": 1,
+ "harness|mmlu_sociology|5": 1,
+ "harness|mmlu_college_medicine|5": 1,
+ "harness|mmlu_elementary_mathematics|5": 1,
+ "harness|mmlu_college_biology|5": 1,
+ "harness|mmlu_college_chemistry|5": 1,
+ "harness|mmlu_us_foreign_policy|5": 1,
+ "harness|mmlu_moral_disputes|5": 1,
+ "harness|mmlu_logical_fallacies|5": 1,
+ "harness|mmlu_prehistory|5": 1,
+ "harness|mmlu_college_mathematics|5": 1,
+ "harness|mmlu_high_school_government_and_politics|5": 1,
+ "harness|mmlu_econometrics|5": 1,
+ "harness|mmlu_high_school_psychology|5": 1,
+ "harness|mmlu_formal_logic|5": 1,
+ "harness|mmlu_nutrition|5": 1,
+ "harness|mmlu_business_ethics|5": 1,
+ "harness|mmlu_international_law|5": 1,
+ "harness|mmlu_astronomy|5": 1,
+ "harness|mmlu_professional_psychology|5": 1,
+ "harness|mmlu_professional_accounting|5": 1,
+ "harness|mmlu_machine_learning|5": 1,
+ "harness|mmlu_high_school_statistics|5": 1,
+ "harness|mmlu_moral_scenarios|5": 1,
+ "harness|mmlu_college_computer_science|5": 1,
+ "harness|mmlu_high_school_computer_science|5": 1,
+ "harness|mmlu_professional_medicine|5": 1,
+ "harness|mmlu_security_studies|5": 1,
+ "harness|mmlu_high_school_world_history|5": 1,
+ "harness|mmlu_professional_law|5": 1,
+ "harness|mmlu_high_school_us_history|5": 1,
+ "harness|mmlu_high_school_european_history|5": 1,
+ "harness|truthfulqa_mc|0": 0,
+ "harness|commongen_v2|2": 1
+ },
+ "config_general": {
+ "model_name": "01-ai/Yi-1.5-9B-32K",
+ "model_sha": "c0239dbc923b8a2b5ca849763bdd592d39c60850",
+ "model_dtype": "torch.float16",
+ "lighteval_sha": "",
+ "num_few_shot_default": 0,
+ "num_fewshot_seeds": 1,
+ "override_batch_size": 1,
+ "max_samples": null
+ }
+}
\ No newline at end of file
diff --git a/eval-results/BioMistral/BioMistral-7B/BioMistral_BioMistral-7B_result_2024-05-30 01_33_58.json b/eval-results/BioMistral/BioMistral-7B/BioMistral_BioMistral-7B_result_2024-05-30 01_33_58.json
new file mode 100644
index 0000000000000000000000000000000000000000..84b56b65cf0b5b96038c5186afc4f253243511cb
--- /dev/null
+++ b/eval-results/BioMistral/BioMistral-7B/BioMistral_BioMistral-7B_result_2024-05-30 01_33_58.json
@@ -0,0 +1,450 @@
+{
+ "results": {
+ "daily": {
+ "daily": 10
+ },
+ "quarterly": {
+ "quarterly": 10
+ },
+ "harness|arc_challenge|25": {
+ "acc": 0.257679180887372,
+ "acc_stderr": 0.012780770562768416,
+ "acc_norm": 0.3122866894197952,
+ "acc_norm_stderr": 0.013542598541688065
+ },
+ "harness|hellaswag|10": {
+ "acc": 0.3229436367257518,
+ "acc_stderr": 0.004666457279979418,
+ "acc_norm": 0.39255128460466043,
+ "acc_norm_stderr": 0.004873203269366306
+ },
+ "harness|mmlu_world_religions|5": {
+ "acc": 0.34502923976608185,
+ "acc_stderr": 0.036459813773888065,
+ "acc_norm": 0.34502923976608185,
+ "acc_norm_stderr": 0.036459813773888065
+ },
+ "harness|mmlu_management|5": {
+ "acc": 0.4368932038834951,
+ "acc_stderr": 0.04911147107365778,
+ "acc_norm": 0.4368932038834951,
+ "acc_norm_stderr": 0.04911147107365778
+ },
+ "harness|mmlu_miscellaneous|5": {
+ "acc": 0.3780332056194125,
+ "acc_stderr": 0.017339844462104625,
+ "acc_norm": 0.3780332056194125,
+ "acc_norm_stderr": 0.017339844462104625
+ },
+ "harness|mmlu_anatomy|5": {
+ "acc": 0.3037037037037037,
+ "acc_stderr": 0.039725528847851355,
+ "acc_norm": 0.3037037037037037,
+ "acc_norm_stderr": 0.039725528847851355
+ },
+ "harness|mmlu_abstract_algebra|5": {
+ "acc": 0.37,
+ "acc_stderr": 0.04852365870939099,
+ "acc_norm": 0.37,
+ "acc_norm_stderr": 0.04852365870939099
+ },
+ "harness|mmlu_conceptual_physics|5": {
+ "acc": 0.28085106382978725,
+ "acc_stderr": 0.02937917046412482,
+ "acc_norm": 0.28085106382978725,
+ "acc_norm_stderr": 0.02937917046412482
+ },
+ "harness|mmlu_virology|5": {
+ "acc": 0.3373493975903614,
+ "acc_stderr": 0.03680783690727581,
+ "acc_norm": 0.3373493975903614,
+ "acc_norm_stderr": 0.03680783690727581
+ },
+ "harness|mmlu_philosophy|5": {
+ "acc": 0.3954983922829582,
+ "acc_stderr": 0.027770918531427838,
+ "acc_norm": 0.3954983922829582,
+ "acc_norm_stderr": 0.027770918531427838
+ },
+ "harness|mmlu_human_aging|5": {
+ "acc": 0.34977578475336324,
+ "acc_stderr": 0.03200736719484503,
+ "acc_norm": 0.34977578475336324,
+ "acc_norm_stderr": 0.03200736719484503
+ },
+ "harness|mmlu_human_sexuality|5": {
+ "acc": 0.3969465648854962,
+ "acc_stderr": 0.04291135671009224,
+ "acc_norm": 0.3969465648854962,
+ "acc_norm_stderr": 0.04291135671009224
+ },
+ "harness|mmlu_medical_genetics|5": {
+ "acc": 0.42,
+ "acc_stderr": 0.049604496374885836,
+ "acc_norm": 0.42,
+ "acc_norm_stderr": 0.049604496374885836
+ },
+ "harness|mmlu_high_school_geography|5": {
+ "acc": 0.4292929292929293,
+ "acc_stderr": 0.03526552724601199,
+ "acc_norm": 0.4292929292929293,
+ "acc_norm_stderr": 0.03526552724601199
+ },
+ "harness|mmlu_electrical_engineering|5": {
+ "acc": 0.4,
+ "acc_stderr": 0.04082482904638628,
+ "acc_norm": 0.4,
+ "acc_norm_stderr": 0.04082482904638628
+ },
+ "harness|mmlu_college_physics|5": {
+ "acc": 0.30392156862745096,
+ "acc_stderr": 0.045766654032077636,
+ "acc_norm": 0.30392156862745096,
+ "acc_norm_stderr": 0.045766654032077636
+ },
+ "harness|mmlu_high_school_microeconomics|5": {
+ "acc": 0.40336134453781514,
+ "acc_stderr": 0.031866081214088314,
+ "acc_norm": 0.40336134453781514,
+ "acc_norm_stderr": 0.031866081214088314
+ },
+ "harness|mmlu_high_school_macroeconomics|5": {
+ "acc": 0.40512820512820513,
+ "acc_stderr": 0.024890471769938145,
+ "acc_norm": 0.40512820512820513,
+ "acc_norm_stderr": 0.024890471769938145
+ },
+ "harness|mmlu_computer_security|5": {
+ "acc": 0.48,
+ "acc_stderr": 0.050211673156867795,
+ "acc_norm": 0.48,
+ "acc_norm_stderr": 0.050211673156867795
+ },
+ "harness|mmlu_global_facts|5": {
+ "acc": 0.32,
+ "acc_stderr": 0.04688261722621505,
+ "acc_norm": 0.32,
+ "acc_norm_stderr": 0.04688261722621505
+ },
+ "harness|mmlu_jurisprudence|5": {
+ "acc": 0.49074074074074076,
+ "acc_stderr": 0.04832853553437055,
+ "acc_norm": 0.49074074074074076,
+ "acc_norm_stderr": 0.04832853553437055
+ },
+ "harness|mmlu_high_school_chemistry|5": {
+ "acc": 0.37438423645320196,
+ "acc_stderr": 0.03405155380561952,
+ "acc_norm": 0.37438423645320196,
+ "acc_norm_stderr": 0.03405155380561952
+ },
+ "harness|mmlu_high_school_biology|5": {
+ "acc": 0.36774193548387096,
+ "acc_stderr": 0.027430866579973474,
+ "acc_norm": 0.36774193548387096,
+ "acc_norm_stderr": 0.027430866579973474
+ },
+ "harness|mmlu_marketing|5": {
+ "acc": 0.5598290598290598,
+ "acc_stderr": 0.0325207417206305,
+ "acc_norm": 0.5598290598290598,
+ "acc_norm_stderr": 0.0325207417206305
+ },
+ "harness|mmlu_clinical_knowledge|5": {
+ "acc": 0.3886792452830189,
+ "acc_stderr": 0.030000485448675986,
+ "acc_norm": 0.3886792452830189,
+ "acc_norm_stderr": 0.030000485448675986
+ },
+ "harness|mmlu_public_relations|5": {
+ "acc": 0.44545454545454544,
+ "acc_stderr": 0.047605488214603246,
+ "acc_norm": 0.44545454545454544,
+ "acc_norm_stderr": 0.047605488214603246
+ },
+ "harness|mmlu_high_school_mathematics|5": {
+ "acc": 0.34444444444444444,
+ "acc_stderr": 0.028972648884844267,
+ "acc_norm": 0.34444444444444444,
+ "acc_norm_stderr": 0.028972648884844267
+ },
+ "harness|mmlu_high_school_physics|5": {
+ "acc": 0.3443708609271523,
+ "acc_stderr": 0.038796870240733264,
+ "acc_norm": 0.3443708609271523,
+ "acc_norm_stderr": 0.038796870240733264
+ },
+ "harness|mmlu_sociology|5": {
+ "acc": 0.4577114427860697,
+ "acc_stderr": 0.035228658640995975,
+ "acc_norm": 0.4577114427860697,
+ "acc_norm_stderr": 0.035228658640995975
+ },
+ "harness|mmlu_college_medicine|5": {
+ "acc": 0.3815028901734104,
+ "acc_stderr": 0.03703851193099521,
+ "acc_norm": 0.3815028901734104,
+ "acc_norm_stderr": 0.03703851193099521
+ },
+ "harness|mmlu_elementary_mathematics|5": {
+ "acc": 0.35714285714285715,
+ "acc_stderr": 0.02467786284133278,
+ "acc_norm": 0.35714285714285715,
+ "acc_norm_stderr": 0.02467786284133278
+ },
+ "harness|mmlu_college_biology|5": {
+ "acc": 0.3333333333333333,
+ "acc_stderr": 0.03942082639927213,
+ "acc_norm": 0.3333333333333333,
+ "acc_norm_stderr": 0.03942082639927213
+ },
+ "harness|mmlu_college_chemistry|5": {
+ "acc": 0.47,
+ "acc_stderr": 0.05016135580465919,
+ "acc_norm": 0.47,
+ "acc_norm_stderr": 0.05016135580465919
+ },
+ "harness|mmlu_us_foreign_policy|5": {
+ "acc": 0.54,
+ "acc_stderr": 0.05009082659620333,
+ "acc_norm": 0.54,
+ "acc_norm_stderr": 0.05009082659620333
+ },
+ "harness|mmlu_moral_disputes|5": {
+ "acc": 0.44508670520231214,
+ "acc_stderr": 0.02675625512966377,
+ "acc_norm": 0.44508670520231214,
+ "acc_norm_stderr": 0.02675625512966377
+ },
+ "harness|mmlu_logical_fallacies|5": {
+ "acc": 0.34355828220858897,
+ "acc_stderr": 0.03731133519673893,
+ "acc_norm": 0.34355828220858897,
+ "acc_norm_stderr": 0.03731133519673893
+ },
+ "harness|mmlu_prehistory|5": {
+ "acc": 0.37037037037037035,
+ "acc_stderr": 0.02686949074481525,
+ "acc_norm": 0.37037037037037035,
+ "acc_norm_stderr": 0.02686949074481525
+ },
+ "harness|mmlu_college_mathematics|5": {
+ "acc": 0.33,
+ "acc_stderr": 0.04725815626252605,
+ "acc_norm": 0.33,
+ "acc_norm_stderr": 0.04725815626252605
+ },
+ "harness|mmlu_high_school_government_and_politics|5": {
+ "acc": 0.44559585492227977,
+ "acc_stderr": 0.0358701498607566,
+ "acc_norm": 0.44559585492227977,
+ "acc_norm_stderr": 0.0358701498607566
+ },
+ "harness|mmlu_econometrics|5": {
+ "acc": 0.2719298245614035,
+ "acc_stderr": 0.041857744240220575,
+ "acc_norm": 0.2719298245614035,
+ "acc_norm_stderr": 0.041857744240220575
+ },
+ "harness|mmlu_high_school_psychology|5": {
+ "acc": 0.3798165137614679,
+ "acc_stderr": 0.020808825617866244,
+ "acc_norm": 0.3798165137614679,
+ "acc_norm_stderr": 0.020808825617866244
+ },
+ "harness|mmlu_formal_logic|5": {
+ "acc": 0.3492063492063492,
+ "acc_stderr": 0.04263906892795132,
+ "acc_norm": 0.3492063492063492,
+ "acc_norm_stderr": 0.04263906892795132
+ },
+ "harness|mmlu_nutrition|5": {
+ "acc": 0.4117647058823529,
+ "acc_stderr": 0.02818059632825929,
+ "acc_norm": 0.4117647058823529,
+ "acc_norm_stderr": 0.02818059632825929
+ },
+ "harness|mmlu_business_ethics|5": {
+ "acc": 0.42,
+ "acc_stderr": 0.049604496374885836,
+ "acc_norm": 0.42,
+ "acc_norm_stderr": 0.049604496374885836
+ },
+ "harness|mmlu_international_law|5": {
+ "acc": 0.5619834710743802,
+ "acc_stderr": 0.045291468044357915,
+ "acc_norm": 0.5619834710743802,
+ "acc_norm_stderr": 0.045291468044357915
+ },
+ "harness|mmlu_astronomy|5": {
+ "acc": 0.34868421052631576,
+ "acc_stderr": 0.038781398887976125,
+ "acc_norm": 0.34868421052631576,
+ "acc_norm_stderr": 0.038781398887976125
+ },
+ "harness|mmlu_professional_psychology|5": {
+ "acc": 0.3284313725490196,
+ "acc_stderr": 0.018999707383162666,
+ "acc_norm": 0.3284313725490196,
+ "acc_norm_stderr": 0.018999707383162666
+ },
+ "harness|mmlu_professional_accounting|5": {
+ "acc": 0.2730496453900709,
+ "acc_stderr": 0.026577860943307857,
+ "acc_norm": 0.2730496453900709,
+ "acc_norm_stderr": 0.026577860943307857
+ },
+ "harness|mmlu_machine_learning|5": {
+ "acc": 0.2767857142857143,
+ "acc_stderr": 0.04246624336697627,
+ "acc_norm": 0.2767857142857143,
+ "acc_norm_stderr": 0.04246624336697627
+ },
+ "harness|mmlu_high_school_statistics|5": {
+ "acc": 0.4074074074074074,
+ "acc_stderr": 0.03350991604696043,
+ "acc_norm": 0.4074074074074074,
+ "acc_norm_stderr": 0.03350991604696043
+ },
+ "harness|mmlu_moral_scenarios|5": {
+ "acc": 0.23910614525139665,
+ "acc_stderr": 0.014265554192331149,
+ "acc_norm": 0.23910614525139665,
+ "acc_norm_stderr": 0.014265554192331149
+ },
+ "harness|mmlu_college_computer_science|5": {
+ "acc": 0.33,
+ "acc_stderr": 0.047258156262526045,
+ "acc_norm": 0.33,
+ "acc_norm_stderr": 0.047258156262526045
+ },
+ "harness|mmlu_high_school_computer_science|5": {
+ "acc": 0.4,
+ "acc_stderr": 0.04923659639173309,
+ "acc_norm": 0.4,
+ "acc_norm_stderr": 0.04923659639173309
+ },
+ "harness|mmlu_professional_medicine|5": {
+ "acc": 0.4227941176470588,
+ "acc_stderr": 0.030008562845003483,
+ "acc_norm": 0.4227941176470588,
+ "acc_norm_stderr": 0.030008562845003483
+ },
+ "harness|mmlu_security_studies|5": {
+ "acc": 0.3469387755102041,
+ "acc_stderr": 0.030472526026726492,
+ "acc_norm": 0.3469387755102041,
+ "acc_norm_stderr": 0.030472526026726492
+ },
+ "harness|mmlu_high_school_world_history|5": {
+ "acc": 0.4177215189873418,
+ "acc_stderr": 0.032103530322412685,
+ "acc_norm": 0.4177215189873418,
+ "acc_norm_stderr": 0.032103530322412685
+ },
+ "harness|mmlu_professional_law|5": {
+ "acc": 0.3005215123859192,
+ "acc_stderr": 0.011709918883039124,
+ "acc_norm": 0.3005215123859192,
+ "acc_norm_stderr": 0.011709918883039124
+ },
+ "harness|mmlu_high_school_us_history|5": {
+ "acc": 0.3872549019607843,
+ "acc_stderr": 0.03418931233833344,
+ "acc_norm": 0.3872549019607843,
+ "acc_norm_stderr": 0.03418931233833344
+ },
+ "harness|mmlu_high_school_european_history|5": {
+ "acc": 0.43636363636363634,
+ "acc_stderr": 0.03872592983524753,
+ "acc_norm": 0.43636363636363634,
+ "acc_norm_stderr": 0.03872592983524753
+ },
+ "harness|truthfulqa_mc|0": {
+ "mc1": 0.3072215422276622,
+ "mc1_stderr": 0.016150201321323002,
+ "mc2": 0.4721418472000992,
+ "mc2_stderr": 0.01626625866283201
+ },
+ "harness|commongen_v2|2": {
+ "acc": 0.27863046044864226,
+ "acc_stderr": 0.01541373949434568,
+ "acc_norm": 0.3825265643447462,
+ "acc_norm_stderr": 0.016709165387228803
+ }
+ },
+ "versions": {
+ "all": 0,
+ "harness|arc_challenge|25": 0,
+ "harness|hellaswag|10": 0,
+ "harness|mmlu_world_religions|5": 1,
+ "harness|mmlu_management|5": 1,
+ "harness|mmlu_miscellaneous|5": 1,
+ "harness|mmlu_anatomy|5": 1,
+ "harness|mmlu_abstract_algebra|5": 1,
+ "harness|mmlu_conceptual_physics|5": 1,
+ "harness|mmlu_virology|5": 1,
+ "harness|mmlu_philosophy|5": 1,
+ "harness|mmlu_human_aging|5": 1,
+ "harness|mmlu_human_sexuality|5": 1,
+ "harness|mmlu_medical_genetics|5": 1,
+ "harness|mmlu_high_school_geography|5": 1,
+ "harness|mmlu_electrical_engineering|5": 1,
+ "harness|mmlu_college_physics|5": 1,
+ "harness|mmlu_high_school_microeconomics|5": 1,
+ "harness|mmlu_high_school_macroeconomics|5": 1,
+ "harness|mmlu_computer_security|5": 1,
+ "harness|mmlu_global_facts|5": 1,
+ "harness|mmlu_jurisprudence|5": 1,
+ "harness|mmlu_high_school_chemistry|5": 1,
+ "harness|mmlu_high_school_biology|5": 1,
+ "harness|mmlu_marketing|5": 1,
+ "harness|mmlu_clinical_knowledge|5": 1,
+ "harness|mmlu_public_relations|5": 1,
+ "harness|mmlu_high_school_mathematics|5": 1,
+ "harness|mmlu_high_school_physics|5": 1,
+ "harness|mmlu_sociology|5": 1,
+ "harness|mmlu_college_medicine|5": 1,
+ "harness|mmlu_elementary_mathematics|5": 1,
+ "harness|mmlu_college_biology|5": 1,
+ "harness|mmlu_college_chemistry|5": 1,
+ "harness|mmlu_us_foreign_policy|5": 1,
+ "harness|mmlu_moral_disputes|5": 1,
+ "harness|mmlu_logical_fallacies|5": 1,
+ "harness|mmlu_prehistory|5": 1,
+ "harness|mmlu_college_mathematics|5": 1,
+ "harness|mmlu_high_school_government_and_politics|5": 1,
+ "harness|mmlu_econometrics|5": 1,
+ "harness|mmlu_high_school_psychology|5": 1,
+ "harness|mmlu_formal_logic|5": 1,
+ "harness|mmlu_nutrition|5": 1,
+ "harness|mmlu_business_ethics|5": 1,
+ "harness|mmlu_international_law|5": 1,
+ "harness|mmlu_astronomy|5": 1,
+ "harness|mmlu_professional_psychology|5": 1,
+ "harness|mmlu_professional_accounting|5": 1,
+ "harness|mmlu_machine_learning|5": 1,
+ "harness|mmlu_high_school_statistics|5": 1,
+ "harness|mmlu_moral_scenarios|5": 1,
+ "harness|mmlu_college_computer_science|5": 1,
+ "harness|mmlu_high_school_computer_science|5": 1,
+ "harness|mmlu_professional_medicine|5": 1,
+ "harness|mmlu_security_studies|5": 1,
+ "harness|mmlu_high_school_world_history|5": 1,
+ "harness|mmlu_professional_law|5": 1,
+ "harness|mmlu_high_school_us_history|5": 1,
+ "harness|mmlu_high_school_european_history|5": 1,
+ "harness|truthfulqa_mc|0": 0,
+ "harness|commongen_v2|2": 1
+ },
+ "config_general": {
+ "model_name": "BioMistral/BioMistral-7B",
+ "model_sha": "9a11e1ffa817c211cbb52ee1fb312dc6b61b40a5",
+ "model_dtype": "torch.float16",
+ "lighteval_sha": "",
+ "num_few_shot_default": 0,
+ "num_fewshot_seeds": 1,
+ "override_batch_size": 1,
+ "max_samples": null
+ }
+}
\ No newline at end of file
diff --git a/eval-results/EleutherAI/polyglot-ko-1.3b/EleutherAI_polyglot-ko-1.3b_result_2023-09-24 15_21_38.json b/eval-results/EleutherAI/polyglot-ko-1.3b/EleutherAI_polyglot-ko-1.3b_result_2023-09-24 15_21_38.json
new file mode 100644
index 0000000000000000000000000000000000000000..ac0ac041b644148e7ea7a60a1059234fe3564f1d
--- /dev/null
+++ b/eval-results/EleutherAI/polyglot-ko-1.3b/EleutherAI_polyglot-ko-1.3b_result_2023-09-24 15_21_38.json
@@ -0,0 +1,450 @@
+{
+ "results": {
+ "daily": {
+ "daily": 11
+ },
+ "quarterly": {
+ "quarterly": 11
+ },
+ "harness|arc_challenge|25": {
+ "acc": 0.2235494880546075,
+ "acc_stderr": 0.012174896631202605,
+ "acc_norm": 0.2815699658703072,
+ "acc_norm_stderr": 0.013143376735009015
+ },
+ "harness|hellaswag|10": {
+ "acc": 0.3345947022505477,
+ "acc_stderr": 0.004708842600177431,
+ "acc_norm": 0.4135630352519418,
+ "acc_norm_stderr": 0.0049146550633294974
+ },
+ "harness|mmlu_world_religions|5": {
+ "acc": 0.27485380116959063,
+ "acc_stderr": 0.03424042924691585,
+ "acc_norm": 0.27485380116959063,
+ "acc_norm_stderr": 0.03424042924691585
+ },
+ "harness|mmlu_management|5": {
+ "acc": 0.27184466019417475,
+ "acc_stderr": 0.044052680241409216,
+ "acc_norm": 0.27184466019417475,
+ "acc_norm_stderr": 0.044052680241409216
+ },
+ "harness|mmlu_miscellaneous|5": {
+ "acc": 0.26947637292464877,
+ "acc_stderr": 0.015866243073215065,
+ "acc_norm": 0.26947637292464877,
+ "acc_norm_stderr": 0.015866243073215065
+ },
+ "harness|mmlu_anatomy|5": {
+ "acc": 0.26666666666666666,
+ "acc_stderr": 0.038201699145179055,
+ "acc_norm": 0.26666666666666666,
+ "acc_norm_stderr": 0.038201699145179055
+ },
+ "harness|mmlu_abstract_algebra|5": {
+ "acc": 0.3,
+ "acc_stderr": 0.046056618647183814,
+ "acc_norm": 0.3,
+ "acc_norm_stderr": 0.046056618647183814
+ },
+ "harness|mmlu_conceptual_physics|5": {
+ "acc": 0.2127659574468085,
+ "acc_stderr": 0.026754391348039783,
+ "acc_norm": 0.2127659574468085,
+ "acc_norm_stderr": 0.026754391348039783
+ },
+ "harness|mmlu_virology|5": {
+ "acc": 0.24096385542168675,
+ "acc_stderr": 0.033293941190735296,
+ "acc_norm": 0.24096385542168675,
+ "acc_norm_stderr": 0.033293941190735296
+ },
+ "harness|mmlu_philosophy|5": {
+ "acc": 0.2379421221864952,
+ "acc_stderr": 0.024185150647818707,
+ "acc_norm": 0.2379421221864952,
+ "acc_norm_stderr": 0.024185150647818707
+ },
+ "harness|mmlu_human_aging|5": {
+ "acc": 0.2825112107623318,
+ "acc_stderr": 0.030216831011508766,
+ "acc_norm": 0.2825112107623318,
+ "acc_norm_stderr": 0.030216831011508766
+ },
+ "harness|mmlu_human_sexuality|5": {
+ "acc": 0.21374045801526717,
+ "acc_stderr": 0.0359546161177469,
+ "acc_norm": 0.21374045801526717,
+ "acc_norm_stderr": 0.0359546161177469
+ },
+ "harness|mmlu_medical_genetics|5": {
+ "acc": 0.24,
+ "acc_stderr": 0.042923469599092816,
+ "acc_norm": 0.24,
+ "acc_norm_stderr": 0.042923469599092816
+ },
+ "harness|mmlu_high_school_geography|5": {
+ "acc": 0.2474747474747475,
+ "acc_stderr": 0.03074630074212451,
+ "acc_norm": 0.2474747474747475,
+ "acc_norm_stderr": 0.03074630074212451
+ },
+ "harness|mmlu_electrical_engineering|5": {
+ "acc": 0.22758620689655173,
+ "acc_stderr": 0.03493950380131184,
+ "acc_norm": 0.22758620689655173,
+ "acc_norm_stderr": 0.03493950380131184
+ },
+ "harness|mmlu_college_physics|5": {
+ "acc": 0.22549019607843138,
+ "acc_stderr": 0.041583075330832865,
+ "acc_norm": 0.22549019607843138,
+ "acc_norm_stderr": 0.041583075330832865
+ },
+ "harness|mmlu_high_school_microeconomics|5": {
+ "acc": 0.31512605042016806,
+ "acc_stderr": 0.030176808288974337,
+ "acc_norm": 0.31512605042016806,
+ "acc_norm_stderr": 0.030176808288974337
+ },
+ "harness|mmlu_high_school_macroeconomics|5": {
+ "acc": 0.2205128205128205,
+ "acc_stderr": 0.02102067268082791,
+ "acc_norm": 0.2205128205128205,
+ "acc_norm_stderr": 0.02102067268082791
+ },
+ "harness|mmlu_computer_security|5": {
+ "acc": 0.18,
+ "acc_stderr": 0.038612291966536955,
+ "acc_norm": 0.18,
+ "acc_norm_stderr": 0.038612291966536955
+ },
+ "harness|mmlu_global_facts|5": {
+ "acc": 0.31,
+ "acc_stderr": 0.04648231987117316,
+ "acc_norm": 0.31,
+ "acc_norm_stderr": 0.04648231987117316
+ },
+ "harness|mmlu_jurisprudence|5": {
+ "acc": 0.25,
+ "acc_stderr": 0.04186091791394607,
+ "acc_norm": 0.25,
+ "acc_norm_stderr": 0.04186091791394607
+ },
+ "harness|mmlu_high_school_chemistry|5": {
+ "acc": 0.2660098522167488,
+ "acc_stderr": 0.03108982600293752,
+ "acc_norm": 0.2660098522167488,
+ "acc_norm_stderr": 0.03108982600293752
+ },
+ "harness|mmlu_high_school_biology|5": {
+ "acc": 0.3,
+ "acc_stderr": 0.02606936229533513,
+ "acc_norm": 0.3,
+ "acc_norm_stderr": 0.02606936229533513
+ },
+ "harness|mmlu_marketing|5": {
+ "acc": 0.23076923076923078,
+ "acc_stderr": 0.027601921381417607,
+ "acc_norm": 0.23076923076923078,
+ "acc_norm_stderr": 0.027601921381417607
+ },
+ "harness|mmlu_clinical_knowledge|5": {
+ "acc": 0.25660377358490566,
+ "acc_stderr": 0.026880647889051968,
+ "acc_norm": 0.25660377358490566,
+ "acc_norm_stderr": 0.026880647889051968
+ },
+ "harness|mmlu_public_relations|5": {
+ "acc": 0.2545454545454545,
+ "acc_stderr": 0.04172343038705383,
+ "acc_norm": 0.2545454545454545,
+ "acc_norm_stderr": 0.04172343038705383
+ },
+ "harness|mmlu_high_school_mathematics|5": {
+ "acc": 0.2962962962962963,
+ "acc_stderr": 0.02784081149587194,
+ "acc_norm": 0.2962962962962963,
+ "acc_norm_stderr": 0.02784081149587194
+ },
+ "harness|mmlu_high_school_physics|5": {
+ "acc": 0.304635761589404,
+ "acc_stderr": 0.03757949922943342,
+ "acc_norm": 0.304635761589404,
+ "acc_norm_stderr": 0.03757949922943342
+ },
+ "harness|mmlu_sociology|5": {
+ "acc": 0.25870646766169153,
+ "acc_stderr": 0.03096590312357303,
+ "acc_norm": 0.25870646766169153,
+ "acc_norm_stderr": 0.03096590312357303
+ },
+ "harness|mmlu_college_medicine|5": {
+ "acc": 0.2254335260115607,
+ "acc_stderr": 0.03186209851641144,
+ "acc_norm": 0.2254335260115607,
+ "acc_norm_stderr": 0.03186209851641144
+ },
+ "harness|mmlu_elementary_mathematics|5": {
+ "acc": 0.2566137566137566,
+ "acc_stderr": 0.022494510767503154,
+ "acc_norm": 0.2566137566137566,
+ "acc_norm_stderr": 0.022494510767503154
+ },
+ "harness|mmlu_college_biology|5": {
+ "acc": 0.2638888888888889,
+ "acc_stderr": 0.03685651095897532,
+ "acc_norm": 0.2638888888888889,
+ "acc_norm_stderr": 0.03685651095897532
+ },
+ "harness|mmlu_college_chemistry|5": {
+ "acc": 0.23,
+ "acc_stderr": 0.04229525846816505,
+ "acc_norm": 0.23,
+ "acc_norm_stderr": 0.04229525846816505
+ },
+ "harness|mmlu_us_foreign_policy|5": {
+ "acc": 0.22,
+ "acc_stderr": 0.04163331998932269,
+ "acc_norm": 0.22,
+ "acc_norm_stderr": 0.04163331998932269
+ },
+ "harness|mmlu_moral_disputes|5": {
+ "acc": 0.24855491329479767,
+ "acc_stderr": 0.023267528432100174,
+ "acc_norm": 0.24855491329479767,
+ "acc_norm_stderr": 0.023267528432100174
+ },
+ "harness|mmlu_logical_fallacies|5": {
+ "acc": 0.31901840490797545,
+ "acc_stderr": 0.03661997551073836,
+ "acc_norm": 0.31901840490797545,
+ "acc_norm_stderr": 0.03661997551073836
+ },
+ "harness|mmlu_prehistory|5": {
+ "acc": 0.2623456790123457,
+ "acc_stderr": 0.024477222856135114,
+ "acc_norm": 0.2623456790123457,
+ "acc_norm_stderr": 0.024477222856135114
+ },
+ "harness|mmlu_college_mathematics|5": {
+ "acc": 0.25,
+ "acc_stderr": 0.04351941398892446,
+ "acc_norm": 0.25,
+ "acc_norm_stderr": 0.04351941398892446
+ },
+ "harness|mmlu_high_school_government_and_politics|5": {
+ "acc": 0.33678756476683935,
+ "acc_stderr": 0.03410780251836184,
+ "acc_norm": 0.33678756476683935,
+ "acc_norm_stderr": 0.03410780251836184
+ },
+ "harness|mmlu_econometrics|5": {
+ "acc": 0.20175438596491227,
+ "acc_stderr": 0.037752050135836386,
+ "acc_norm": 0.20175438596491227,
+ "acc_norm_stderr": 0.037752050135836386
+ },
+ "harness|mmlu_high_school_psychology|5": {
+ "acc": 0.24220183486238533,
+ "acc_stderr": 0.01836817630659862,
+ "acc_norm": 0.24220183486238533,
+ "acc_norm_stderr": 0.01836817630659862
+ },
+ "harness|mmlu_formal_logic|5": {
+ "acc": 0.23015873015873015,
+ "acc_stderr": 0.03764950879790606,
+ "acc_norm": 0.23015873015873015,
+ "acc_norm_stderr": 0.03764950879790606
+ },
+ "harness|mmlu_nutrition|5": {
+ "acc": 0.23529411764705882,
+ "acc_stderr": 0.024288619466046102,
+ "acc_norm": 0.23529411764705882,
+ "acc_norm_stderr": 0.024288619466046102
+ },
+ "harness|mmlu_business_ethics|5": {
+ "acc": 0.18,
+ "acc_stderr": 0.03861229196653695,
+ "acc_norm": 0.18,
+ "acc_norm_stderr": 0.03861229196653695
+ },
+ "harness|mmlu_international_law|5": {
+ "acc": 0.256198347107438,
+ "acc_stderr": 0.039849796533028704,
+ "acc_norm": 0.256198347107438,
+ "acc_norm_stderr": 0.039849796533028704
+ },
+ "harness|mmlu_astronomy|5": {
+ "acc": 0.21710526315789475,
+ "acc_stderr": 0.033550453048829226,
+ "acc_norm": 0.21710526315789475,
+ "acc_norm_stderr": 0.033550453048829226
+ },
+ "harness|mmlu_professional_psychology|5": {
+ "acc": 0.24019607843137256,
+ "acc_stderr": 0.01728276069516743,
+ "acc_norm": 0.24019607843137256,
+ "acc_norm_stderr": 0.01728276069516743
+ },
+ "harness|mmlu_professional_accounting|5": {
+ "acc": 0.2553191489361702,
+ "acc_stderr": 0.02601199293090201,
+ "acc_norm": 0.2553191489361702,
+ "acc_norm_stderr": 0.02601199293090201
+ },
+ "harness|mmlu_machine_learning|5": {
+ "acc": 0.21428571428571427,
+ "acc_stderr": 0.03894641120044793,
+ "acc_norm": 0.21428571428571427,
+ "acc_norm_stderr": 0.03894641120044793
+ },
+ "harness|mmlu_high_school_statistics|5": {
+ "acc": 0.46296296296296297,
+ "acc_stderr": 0.03400603625538272,
+ "acc_norm": 0.46296296296296297,
+ "acc_norm_stderr": 0.03400603625538272
+ },
+ "harness|mmlu_moral_scenarios|5": {
+ "acc": 0.24692737430167597,
+ "acc_stderr": 0.014422292204808852,
+ "acc_norm": 0.24692737430167597,
+ "acc_norm_stderr": 0.014422292204808852
+ },
+ "harness|mmlu_college_computer_science|5": {
+ "acc": 0.25,
+ "acc_stderr": 0.04351941398892446,
+ "acc_norm": 0.25,
+ "acc_norm_stderr": 0.04351941398892446
+ },
+ "harness|mmlu_high_school_computer_science|5": {
+ "acc": 0.3,
+ "acc_stderr": 0.046056618647183814,
+ "acc_norm": 0.3,
+ "acc_norm_stderr": 0.046056618647183814
+ },
+ "harness|mmlu_professional_medicine|5": {
+ "acc": 0.4411764705882353,
+ "acc_stderr": 0.030161911930767102,
+ "acc_norm": 0.4411764705882353,
+ "acc_norm_stderr": 0.030161911930767102
+ },
+ "harness|mmlu_security_studies|5": {
+ "acc": 0.3795918367346939,
+ "acc_stderr": 0.03106721126287249,
+ "acc_norm": 0.3795918367346939,
+ "acc_norm_stderr": 0.03106721126287249
+ },
+ "harness|mmlu_high_school_world_history|5": {
+ "acc": 0.2109704641350211,
+ "acc_stderr": 0.02655837250266192,
+ "acc_norm": 0.2109704641350211,
+ "acc_norm_stderr": 0.02655837250266192
+ },
+ "harness|mmlu_professional_law|5": {
+ "acc": 0.23468057366362452,
+ "acc_stderr": 0.010824026872449344,
+ "acc_norm": 0.23468057366362452,
+ "acc_norm_stderr": 0.010824026872449344
+ },
+ "harness|mmlu_high_school_us_history|5": {
+ "acc": 0.25,
+ "acc_stderr": 0.03039153369274154,
+ "acc_norm": 0.25,
+ "acc_norm_stderr": 0.03039153369274154
+ },
+ "harness|mmlu_high_school_european_history|5": {
+ "acc": 0.22424242424242424,
+ "acc_stderr": 0.03256866661681102,
+ "acc_norm": 0.22424242424242424,
+ "acc_norm_stderr": 0.03256866661681102
+ },
+ "harness|truthfulqa_mc|0": {
+ "mc1": 0.25091799265605874,
+ "mc1_stderr": 0.015176985027707682,
+ "mc2": 0.4116568832959107,
+ "mc2_stderr": 0.015044504977529799
+ },
+ "harness|commongen_v2|2": {
+ "acc": 0.27744982290436837,
+ "acc_stderr": 0.015393630236605975,
+ "acc_norm": 0.3400236127508855,
+ "acc_norm_stderr": 0.016286717220737674
+ }
+ },
+ "versions": {
+ "all": 0,
+ "harness|arc_challenge|25": 0,
+ "harness|hellaswag|10": 0,
+ "harness|mmlu_world_religions|5": 1,
+ "harness|mmlu_management|5": 1,
+ "harness|mmlu_miscellaneous|5": 1,
+ "harness|mmlu_anatomy|5": 1,
+ "harness|mmlu_abstract_algebra|5": 1,
+ "harness|mmlu_conceptual_physics|5": 1,
+ "harness|mmlu_virology|5": 1,
+ "harness|mmlu_philosophy|5": 1,
+ "harness|mmlu_human_aging|5": 1,
+ "harness|mmlu_human_sexuality|5": 1,
+ "harness|mmlu_medical_genetics|5": 1,
+ "harness|mmlu_high_school_geography|5": 1,
+ "harness|mmlu_electrical_engineering|5": 1,
+ "harness|mmlu_college_physics|5": 1,
+ "harness|mmlu_high_school_microeconomics|5": 1,
+ "harness|mmlu_high_school_macroeconomics|5": 1,
+ "harness|mmlu_computer_security|5": 1,
+ "harness|mmlu_global_facts|5": 1,
+ "harness|mmlu_jurisprudence|5": 1,
+ "harness|mmlu_high_school_chemistry|5": 1,
+ "harness|mmlu_high_school_biology|5": 1,
+ "harness|mmlu_marketing|5": 1,
+ "harness|mmlu_clinical_knowledge|5": 1,
+ "harness|mmlu_public_relations|5": 1,
+ "harness|mmlu_high_school_mathematics|5": 1,
+ "harness|mmlu_high_school_physics|5": 1,
+ "harness|mmlu_sociology|5": 1,
+ "harness|mmlu_college_medicine|5": 1,
+ "harness|mmlu_elementary_mathematics|5": 1,
+ "harness|mmlu_college_biology|5": 1,
+ "harness|mmlu_college_chemistry|5": 1,
+ "harness|mmlu_us_foreign_policy|5": 1,
+ "harness|mmlu_moral_disputes|5": 1,
+ "harness|mmlu_logical_fallacies|5": 1,
+ "harness|mmlu_prehistory|5": 1,
+ "harness|mmlu_college_mathematics|5": 1,
+ "harness|mmlu_high_school_government_and_politics|5": 1,
+ "harness|mmlu_econometrics|5": 1,
+ "harness|mmlu_high_school_psychology|5": 1,
+ "harness|mmlu_formal_logic|5": 1,
+ "harness|mmlu_nutrition|5": 1,
+ "harness|mmlu_business_ethics|5": 1,
+ "harness|mmlu_international_law|5": 1,
+ "harness|mmlu_astronomy|5": 1,
+ "harness|mmlu_professional_psychology|5": 1,
+ "harness|mmlu_professional_accounting|5": 1,
+ "harness|mmlu_machine_learning|5": 1,
+ "harness|mmlu_high_school_statistics|5": 1,
+ "harness|mmlu_moral_scenarios|5": 1,
+ "harness|mmlu_college_computer_science|5": 1,
+ "harness|mmlu_high_school_computer_science|5": 1,
+ "harness|mmlu_professional_medicine|5": 1,
+ "harness|mmlu_security_studies|5": 1,
+ "harness|mmlu_high_school_world_history|5": 1,
+ "harness|mmlu_professional_law|5": 1,
+ "harness|mmlu_high_school_us_history|5": 1,
+ "harness|mmlu_high_school_european_history|5": 1,
+ "harness|truthfulqa_mc|0": 0,
+ "harness|commongen_v2|2": 1
+ },
+ "config_general": {
+ "model_name": "EleutherAI/polyglot-ko-1.3b",
+ "model_sha": "557e162cf6e944fdbae05bab2e45d066a125eacb",
+ "model_dtype": "torch.float16",
+ "lighteval_sha": "",
+ "num_few_shot_default": 0,
+ "num_fewshot_seeds": 1,
+ "override_batch_size": 1,
+ "max_samples": null
+ }
+}
\ No newline at end of file
diff --git a/eval-results/HuggingFaceH4/.DS_Store b/eval-results/HuggingFaceH4/.DS_Store
new file mode 100644
index 0000000000000000000000000000000000000000..242908a5e43d09a037aba8dd0f0ec7a5e2c351cd
Binary files /dev/null and b/eval-results/HuggingFaceH4/.DS_Store differ
diff --git a/eval-results/HuggingFaceH4/zephyr-7b-beta/result.json b/eval-results/HuggingFaceH4/zephyr-7b-beta/result.json
new file mode 100644
index 0000000000000000000000000000000000000000..03ed7830c4c7cad4daa545fb305d005506cc24c3
--- /dev/null
+++ b/eval-results/HuggingFaceH4/zephyr-7b-beta/result.json
@@ -0,0 +1,450 @@
+{
+ "results": {
+ "daily": {
+ "daily": 8
+ },
+ "quarterly": {
+ "quarterly": 8
+ },
+ "harness|arc_challenge|25": {
+ "acc": 0.33532423208191126,
+ "acc_stderr": 0.01379618294778556,
+ "acc_norm": 0.3848122866894198,
+ "acc_norm_stderr": 0.014218371065251112
+ },
+ "harness|hellaswag|10": {
+ "acc": 0.35480979884485164,
+ "acc_stderr": 0.004774778180345192,
+ "acc_norm": 0.44911372236606256,
+ "acc_norm_stderr": 0.00496387293685794
+ },
+ "harness|mmlu_world_religions|5": {
+ "acc": 0.45614035087719296,
+ "acc_stderr": 0.03820042586602966,
+ "acc_norm": 0.45614035087719296,
+ "acc_norm_stderr": 0.03820042586602966
+ },
+ "harness|mmlu_management|5": {
+ "acc": 0.6019417475728155,
+ "acc_stderr": 0.04846748253977238,
+ "acc_norm": 0.6019417475728155,
+ "acc_norm_stderr": 0.04846748253977238
+ },
+ "harness|mmlu_miscellaneous|5": {
+ "acc": 0.41762452107279696,
+ "acc_stderr": 0.017635637326951534,
+ "acc_norm": 0.41762452107279696,
+ "acc_norm_stderr": 0.017635637326951534
+ },
+ "harness|mmlu_anatomy|5": {
+ "acc": 0.34074074074074073,
+ "acc_stderr": 0.040943762699967946,
+ "acc_norm": 0.34074074074074073,
+ "acc_norm_stderr": 0.040943762699967946
+ },
+ "harness|mmlu_abstract_algebra|5": {
+ "acc": 0.19,
+ "acc_stderr": 0.03942772444036623,
+ "acc_norm": 0.19,
+ "acc_norm_stderr": 0.03942772444036623
+ },
+ "harness|mmlu_conceptual_physics|5": {
+ "acc": 0.2978723404255319,
+ "acc_stderr": 0.029896145682095462,
+ "acc_norm": 0.2978723404255319,
+ "acc_norm_stderr": 0.029896145682095462
+ },
+ "harness|mmlu_virology|5": {
+ "acc": 0.3614457831325301,
+ "acc_stderr": 0.0374005938202932,
+ "acc_norm": 0.3614457831325301,
+ "acc_norm_stderr": 0.0374005938202932
+ },
+ "harness|mmlu_philosophy|5": {
+ "acc": 0.4758842443729904,
+ "acc_stderr": 0.028365041542564584,
+ "acc_norm": 0.4758842443729904,
+ "acc_norm_stderr": 0.028365041542564584
+ },
+ "harness|mmlu_human_aging|5": {
+ "acc": 0.3811659192825112,
+ "acc_stderr": 0.032596251184168284,
+ "acc_norm": 0.3811659192825112,
+ "acc_norm_stderr": 0.032596251184168284
+ },
+ "harness|mmlu_human_sexuality|5": {
+ "acc": 0.3511450381679389,
+ "acc_stderr": 0.04186445163013751,
+ "acc_norm": 0.3511450381679389,
+ "acc_norm_stderr": 0.04186445163013751
+ },
+ "harness|mmlu_medical_genetics|5": {
+ "acc": 0.27,
+ "acc_stderr": 0.0446196043338474,
+ "acc_norm": 0.27,
+ "acc_norm_stderr": 0.0446196043338474
+ },
+ "harness|mmlu_high_school_geography|5": {
+ "acc": 0.494949494949495,
+ "acc_stderr": 0.035621707606254015,
+ "acc_norm": 0.494949494949495,
+ "acc_norm_stderr": 0.035621707606254015
+ },
+ "harness|mmlu_electrical_engineering|5": {
+ "acc": 0.4,
+ "acc_stderr": 0.04082482904638628,
+ "acc_norm": 0.4,
+ "acc_norm_stderr": 0.04082482904638628
+ },
+ "harness|mmlu_college_physics|5": {
+ "acc": 0.3137254901960784,
+ "acc_stderr": 0.04617034827006717,
+ "acc_norm": 0.3137254901960784,
+ "acc_norm_stderr": 0.04617034827006717
+ },
+ "harness|mmlu_high_school_microeconomics|5": {
+ "acc": 0.4957983193277311,
+ "acc_stderr": 0.0324773433444811,
+ "acc_norm": 0.4957983193277311,
+ "acc_norm_stderr": 0.0324773433444811
+ },
+ "harness|mmlu_high_school_macroeconomics|5": {
+ "acc": 0.4256410256410256,
+ "acc_stderr": 0.025069094387296546,
+ "acc_norm": 0.4256410256410256,
+ "acc_norm_stderr": 0.025069094387296546
+ },
+ "harness|mmlu_computer_security|5": {
+ "acc": 0.59,
+ "acc_stderr": 0.049431107042371025,
+ "acc_norm": 0.59,
+ "acc_norm_stderr": 0.049431107042371025
+ },
+ "harness|mmlu_global_facts|5": {
+ "acc": 0.29,
+ "acc_stderr": 0.045604802157206845,
+ "acc_norm": 0.29,
+ "acc_norm_stderr": 0.045604802157206845
+ },
+ "harness|mmlu_jurisprudence|5": {
+ "acc": 0.4537037037037037,
+ "acc_stderr": 0.04812917324536821,
+ "acc_norm": 0.4537037037037037,
+ "acc_norm_stderr": 0.04812917324536821
+ },
+ "harness|mmlu_high_school_chemistry|5": {
+ "acc": 0.35467980295566504,
+ "acc_stderr": 0.03366124489051449,
+ "acc_norm": 0.35467980295566504,
+ "acc_norm_stderr": 0.03366124489051449
+ },
+ "harness|mmlu_high_school_biology|5": {
+ "acc": 0.4290322580645161,
+ "acc_stderr": 0.02815603653823321,
+ "acc_norm": 0.4290322580645161,
+ "acc_norm_stderr": 0.02815603653823321
+ },
+ "harness|mmlu_marketing|5": {
+ "acc": 0.6666666666666666,
+ "acc_stderr": 0.03088273697413865,
+ "acc_norm": 0.6666666666666666,
+ "acc_norm_stderr": 0.03088273697413865
+ },
+ "harness|mmlu_clinical_knowledge|5": {
+ "acc": 0.4188679245283019,
+ "acc_stderr": 0.03036505082911521,
+ "acc_norm": 0.4188679245283019,
+ "acc_norm_stderr": 0.03036505082911521
+ },
+ "harness|mmlu_public_relations|5": {
+ "acc": 0.42727272727272725,
+ "acc_stderr": 0.04738198703545483,
+ "acc_norm": 0.42727272727272725,
+ "acc_norm_stderr": 0.04738198703545483
+ },
+ "harness|mmlu_high_school_mathematics|5": {
+ "acc": 0.34814814814814815,
+ "acc_stderr": 0.029045600290616258,
+ "acc_norm": 0.34814814814814815,
+ "acc_norm_stderr": 0.029045600290616258
+ },
+ "harness|mmlu_high_school_physics|5": {
+ "acc": 0.2913907284768212,
+ "acc_stderr": 0.037101857261199946,
+ "acc_norm": 0.2913907284768212,
+ "acc_norm_stderr": 0.037101857261199946
+ },
+ "harness|mmlu_sociology|5": {
+ "acc": 0.5174129353233831,
+ "acc_stderr": 0.03533389234739245,
+ "acc_norm": 0.5174129353233831,
+ "acc_norm_stderr": 0.03533389234739245
+ },
+ "harness|mmlu_college_medicine|5": {
+ "acc": 0.37572254335260113,
+ "acc_stderr": 0.03692820767264867,
+ "acc_norm": 0.37572254335260113,
+ "acc_norm_stderr": 0.03692820767264867
+ },
+ "harness|mmlu_elementary_mathematics|5": {
+ "acc": 0.3492063492063492,
+ "acc_stderr": 0.024552292209342658,
+ "acc_norm": 0.3492063492063492,
+ "acc_norm_stderr": 0.024552292209342658
+ },
+ "harness|mmlu_college_biology|5": {
+ "acc": 0.3333333333333333,
+ "acc_stderr": 0.039420826399272135,
+ "acc_norm": 0.3333333333333333,
+ "acc_norm_stderr": 0.039420826399272135
+ },
+ "harness|mmlu_college_chemistry|5": {
+ "acc": 0.35,
+ "acc_stderr": 0.04793724854411019,
+ "acc_norm": 0.35,
+ "acc_norm_stderr": 0.04793724854411019
+ },
+ "harness|mmlu_us_foreign_policy|5": {
+ "acc": 0.49,
+ "acc_stderr": 0.05024183937956913,
+ "acc_norm": 0.49,
+ "acc_norm_stderr": 0.05024183937956913
+ },
+ "harness|mmlu_moral_disputes|5": {
+ "acc": 0.47398843930635837,
+ "acc_stderr": 0.026882643434022885,
+ "acc_norm": 0.47398843930635837,
+ "acc_norm_stderr": 0.026882643434022885
+ },
+ "harness|mmlu_logical_fallacies|5": {
+ "acc": 0.44171779141104295,
+ "acc_stderr": 0.039015918258361836,
+ "acc_norm": 0.44171779141104295,
+ "acc_norm_stderr": 0.039015918258361836
+ },
+ "harness|mmlu_prehistory|5": {
+ "acc": 0.42592592592592593,
+ "acc_stderr": 0.027513747284379424,
+ "acc_norm": 0.42592592592592593,
+ "acc_norm_stderr": 0.027513747284379424
+ },
+ "harness|mmlu_college_mathematics|5": {
+ "acc": 0.33,
+ "acc_stderr": 0.04725815626252606,
+ "acc_norm": 0.33,
+ "acc_norm_stderr": 0.04725815626252606
+ },
+ "harness|mmlu_high_school_government_and_politics|5": {
+ "acc": 0.5129533678756477,
+ "acc_stderr": 0.0360722806104775,
+ "acc_norm": 0.5129533678756477,
+ "acc_norm_stderr": 0.0360722806104775
+ },
+ "harness|mmlu_econometrics|5": {
+ "acc": 0.24561403508771928,
+ "acc_stderr": 0.0404933929774814,
+ "acc_norm": 0.24561403508771928,
+ "acc_norm_stderr": 0.0404933929774814
+ },
+ "harness|mmlu_high_school_psychology|5": {
+ "acc": 0.47155963302752296,
+ "acc_stderr": 0.02140261569734804,
+ "acc_norm": 0.47155963302752296,
+ "acc_norm_stderr": 0.02140261569734804
+ },
+ "harness|mmlu_formal_logic|5": {
+ "acc": 0.36507936507936506,
+ "acc_stderr": 0.04306241259127152,
+ "acc_norm": 0.36507936507936506,
+ "acc_norm_stderr": 0.04306241259127152
+ },
+ "harness|mmlu_nutrition|5": {
+ "acc": 0.4117647058823529,
+ "acc_stderr": 0.028180596328259297,
+ "acc_norm": 0.4117647058823529,
+ "acc_norm_stderr": 0.028180596328259297
+ },
+ "harness|mmlu_business_ethics|5": {
+ "acc": 0.44,
+ "acc_stderr": 0.04988876515698589,
+ "acc_norm": 0.44,
+ "acc_norm_stderr": 0.04988876515698589
+ },
+ "harness|mmlu_international_law|5": {
+ "acc": 0.5867768595041323,
+ "acc_stderr": 0.04495087843548408,
+ "acc_norm": 0.5867768595041323,
+ "acc_norm_stderr": 0.04495087843548408
+ },
+ "harness|mmlu_astronomy|5": {
+ "acc": 0.40131578947368424,
+ "acc_stderr": 0.03988903703336284,
+ "acc_norm": 0.40131578947368424,
+ "acc_norm_stderr": 0.03988903703336284
+ },
+ "harness|mmlu_professional_psychology|5": {
+ "acc": 0.32679738562091504,
+ "acc_stderr": 0.018975427920507215,
+ "acc_norm": 0.32679738562091504,
+ "acc_norm_stderr": 0.018975427920507215
+ },
+ "harness|mmlu_professional_accounting|5": {
+ "acc": 0.3333333333333333,
+ "acc_stderr": 0.02812163604063988,
+ "acc_norm": 0.3333333333333333,
+ "acc_norm_stderr": 0.02812163604063988
+ },
+ "harness|mmlu_machine_learning|5": {
+ "acc": 0.3392857142857143,
+ "acc_stderr": 0.04493949068613539,
+ "acc_norm": 0.3392857142857143,
+ "acc_norm_stderr": 0.04493949068613539
+ },
+ "harness|mmlu_high_school_statistics|5": {
+ "acc": 0.41203703703703703,
+ "acc_stderr": 0.03356787758160835,
+ "acc_norm": 0.41203703703703703,
+ "acc_norm_stderr": 0.03356787758160835
+ },
+ "harness|mmlu_moral_scenarios|5": {
+ "acc": 0.329608938547486,
+ "acc_stderr": 0.015721531075183884,
+ "acc_norm": 0.329608938547486,
+ "acc_norm_stderr": 0.015721531075183884
+ },
+ "harness|mmlu_college_computer_science|5": {
+ "acc": 0.39,
+ "acc_stderr": 0.04902071300001975,
+ "acc_norm": 0.39,
+ "acc_norm_stderr": 0.04902071300001975
+ },
+ "harness|mmlu_high_school_computer_science|5": {
+ "acc": 0.61,
+ "acc_stderr": 0.04902071300001975,
+ "acc_norm": 0.61,
+ "acc_norm_stderr": 0.04902071300001975
+ },
+ "harness|mmlu_professional_medicine|5": {
+ "acc": 0.375,
+ "acc_stderr": 0.029408372932278746,
+ "acc_norm": 0.375,
+ "acc_norm_stderr": 0.029408372932278746
+ },
+ "harness|mmlu_security_studies|5": {
+ "acc": 0.43673469387755104,
+ "acc_stderr": 0.03175195237583322,
+ "acc_norm": 0.43673469387755104,
+ "acc_norm_stderr": 0.03175195237583322
+ },
+ "harness|mmlu_high_school_world_history|5": {
+ "acc": 0.4810126582278481,
+ "acc_stderr": 0.03252375148090448,
+ "acc_norm": 0.4810126582278481,
+ "acc_norm_stderr": 0.03252375148090448
+ },
+ "harness|mmlu_professional_law|5": {
+ "acc": 0.29791395045632335,
+ "acc_stderr": 0.011680717340400059,
+ "acc_norm": 0.29791395045632335,
+ "acc_norm_stderr": 0.011680717340400059
+ },
+ "harness|mmlu_high_school_us_history|5": {
+ "acc": 0.29411764705882354,
+ "acc_stderr": 0.03198001660115072,
+ "acc_norm": 0.29411764705882354,
+ "acc_norm_stderr": 0.03198001660115072
+ },
+ "harness|mmlu_high_school_european_history|5": {
+ "acc": 0.30303030303030304,
+ "acc_stderr": 0.03588624800091707,
+ "acc_norm": 0.30303030303030304,
+ "acc_norm_stderr": 0.03588624800091707
+ },
+ "harness|truthfulqa_mc|0": {
+ "mc1": 0.3317013463892289,
+ "mc1_stderr": 0.01648214881024147,
+ "mc2": 0.5171680571717291,
+ "mc2_stderr": 0.01606077987901482
+ },
+ "harness|commongen_v2|2": {
+ "acc": 0.39787485242030696,
+ "acc_stderr": 0.01682795905473339,
+ "acc_norm": 0.4014167650531287,
+ "acc_norm_stderr": 0.01685290785872906
+ }
+ },
+ "versions": {
+ "all": 0,
+ "harness|arc_challenge|25": 0,
+ "harness|hellaswag|10": 0,
+ "harness|mmlu_world_religions|5": 1,
+ "harness|mmlu_management|5": 1,
+ "harness|mmlu_miscellaneous|5": 1,
+ "harness|mmlu_anatomy|5": 1,
+ "harness|mmlu_abstract_algebra|5": 1,
+ "harness|mmlu_conceptual_physics|5": 1,
+ "harness|mmlu_virology|5": 1,
+ "harness|mmlu_philosophy|5": 1,
+ "harness|mmlu_human_aging|5": 1,
+ "harness|mmlu_human_sexuality|5": 1,
+ "harness|mmlu_medical_genetics|5": 1,
+ "harness|mmlu_high_school_geography|5": 1,
+ "harness|mmlu_electrical_engineering|5": 1,
+ "harness|mmlu_college_physics|5": 1,
+ "harness|mmlu_high_school_microeconomics|5": 1,
+ "harness|mmlu_high_school_macroeconomics|5": 1,
+ "harness|mmlu_computer_security|5": 1,
+ "harness|mmlu_global_facts|5": 1,
+ "harness|mmlu_jurisprudence|5": 1,
+ "harness|mmlu_high_school_chemistry|5": 1,
+ "harness|mmlu_high_school_biology|5": 1,
+ "harness|mmlu_marketing|5": 1,
+ "harness|mmlu_clinical_knowledge|5": 1,
+ "harness|mmlu_public_relations|5": 1,
+ "harness|mmlu_high_school_mathematics|5": 1,
+ "harness|mmlu_high_school_physics|5": 1,
+ "harness|mmlu_sociology|5": 1,
+ "harness|mmlu_college_medicine|5": 1,
+ "harness|mmlu_elementary_mathematics|5": 1,
+ "harness|mmlu_college_biology|5": 1,
+ "harness|mmlu_college_chemistry|5": 1,
+ "harness|mmlu_us_foreign_policy|5": 1,
+ "harness|mmlu_moral_disputes|5": 1,
+ "harness|mmlu_logical_fallacies|5": 1,
+ "harness|mmlu_prehistory|5": 1,
+ "harness|mmlu_college_mathematics|5": 1,
+ "harness|mmlu_high_school_government_and_politics|5": 1,
+ "harness|mmlu_econometrics|5": 1,
+ "harness|mmlu_high_school_psychology|5": 1,
+ "harness|mmlu_formal_logic|5": 1,
+ "harness|mmlu_nutrition|5": 1,
+ "harness|mmlu_business_ethics|5": 1,
+ "harness|mmlu_international_law|5": 1,
+ "harness|mmlu_astronomy|5": 1,
+ "harness|mmlu_professional_psychology|5": 1,
+ "harness|mmlu_professional_accounting|5": 1,
+ "harness|mmlu_machine_learning|5": 1,
+ "harness|mmlu_high_school_statistics|5": 1,
+ "harness|mmlu_moral_scenarios|5": 1,
+ "harness|mmlu_college_computer_science|5": 1,
+ "harness|mmlu_high_school_computer_science|5": 1,
+ "harness|mmlu_professional_medicine|5": 1,
+ "harness|mmlu_security_studies|5": 1,
+ "harness|mmlu_high_school_world_history|5": 1,
+ "harness|mmlu_professional_law|5": 1,
+ "harness|mmlu_high_school_us_history|5": 1,
+ "harness|mmlu_high_school_european_history|5": 1,
+ "harness|truthfulqa_mc|0": 0,
+ "harness|commongen_v2|2": 1
+ },
+ "config_general": {
+ "model_name": "HuggingFaceH4/zephyr-7b-beta",
+ "model_sha": "3bac358730f8806e5c3dc7c7e19eb36e045bf720",
+ "model_dtype": "torch.float16",
+ "lighteval_sha": "",
+ "num_few_shot_default": 0,
+ "num_fewshot_seeds": 1,
+ "override_batch_size": 1,
+ "max_samples": null
+ }
+}
\ No newline at end of file
diff --git a/eval-results/README.md b/eval-results/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..7b95401dc46245ac339fc25059d4a56d90b4cde5
--- /dev/null
+++ b/eval-results/README.md
@@ -0,0 +1,3 @@
+---
+license: apache-2.0
+---
diff --git a/eval-results/nlpai-lab/KULLM3/result.json b/eval-results/nlpai-lab/KULLM3/result.json
new file mode 100644
index 0000000000000000000000000000000000000000..be8b17b0dd2fc13e9bab3e5d24fec5086956344b
--- /dev/null
+++ b/eval-results/nlpai-lab/KULLM3/result.json
@@ -0,0 +1,450 @@
+{
+ "results": {
+ "daily": {
+ "daily": 6
+ },
+ "quarterly": {
+ "quarterly": 6
+ },
+ "harness|arc_challenge|25": {
+ "acc": 0.42918088737201365,
+ "acc_stderr": 0.014464085894870651,
+ "acc_norm": 0.46501706484641636,
+ "acc_norm_stderr": 0.014575583922019672
+ },
+ "harness|hellaswag|10": {
+ "acc": 0.445628360884286,
+ "acc_stderr": 0.004960191341430244,
+ "acc_norm": 0.589523999203346,
+ "acc_norm_stderr": 0.004909148239488273
+ },
+ "harness|mmlu_world_religions|5": {
+ "acc": 0.6432748538011696,
+ "acc_stderr": 0.03674013002860954,
+ "acc_norm": 0.6432748538011696,
+ "acc_norm_stderr": 0.03674013002860954
+ },
+ "harness|mmlu_management|5": {
+ "acc": 0.6116504854368932,
+ "acc_stderr": 0.04825729337356389,
+ "acc_norm": 0.6116504854368932,
+ "acc_norm_stderr": 0.04825729337356389
+ },
+ "harness|mmlu_miscellaneous|5": {
+ "acc": 0.6155810983397191,
+ "acc_stderr": 0.01739568874281962,
+ "acc_norm": 0.6155810983397191,
+ "acc_norm_stderr": 0.01739568874281962
+ },
+ "harness|mmlu_anatomy|5": {
+ "acc": 0.4962962962962963,
+ "acc_stderr": 0.04319223625811331,
+ "acc_norm": 0.4962962962962963,
+ "acc_norm_stderr": 0.04319223625811331
+ },
+ "harness|mmlu_abstract_algebra|5": {
+ "acc": 0.26,
+ "acc_stderr": 0.04408440022768077,
+ "acc_norm": 0.26,
+ "acc_norm_stderr": 0.04408440022768077
+ },
+ "harness|mmlu_conceptual_physics|5": {
+ "acc": 0.4553191489361702,
+ "acc_stderr": 0.03255525359340354,
+ "acc_norm": 0.4553191489361702,
+ "acc_norm_stderr": 0.03255525359340354
+ },
+ "harness|mmlu_virology|5": {
+ "acc": 0.5180722891566265,
+ "acc_stderr": 0.038899512528272166,
+ "acc_norm": 0.5180722891566265,
+ "acc_norm_stderr": 0.038899512528272166
+ },
+ "harness|mmlu_philosophy|5": {
+ "acc": 0.5755627009646302,
+ "acc_stderr": 0.028071928247946205,
+ "acc_norm": 0.5755627009646302,
+ "acc_norm_stderr": 0.028071928247946205
+ },
+ "harness|mmlu_human_aging|5": {
+ "acc": 0.5650224215246636,
+ "acc_stderr": 0.033272833702713445,
+ "acc_norm": 0.5650224215246636,
+ "acc_norm_stderr": 0.033272833702713445
+ },
+ "harness|mmlu_human_sexuality|5": {
+ "acc": 0.5877862595419847,
+ "acc_stderr": 0.04317171194870255,
+ "acc_norm": 0.5877862595419847,
+ "acc_norm_stderr": 0.04317171194870255
+ },
+ "harness|mmlu_medical_genetics|5": {
+ "acc": 0.5,
+ "acc_stderr": 0.050251890762960605,
+ "acc_norm": 0.5,
+ "acc_norm_stderr": 0.050251890762960605
+ },
+ "harness|mmlu_high_school_geography|5": {
+ "acc": 0.6515151515151515,
+ "acc_stderr": 0.033948539651564025,
+ "acc_norm": 0.6515151515151515,
+ "acc_norm_stderr": 0.033948539651564025
+ },
+ "harness|mmlu_electrical_engineering|5": {
+ "acc": 0.503448275862069,
+ "acc_stderr": 0.04166567577101579,
+ "acc_norm": 0.503448275862069,
+ "acc_norm_stderr": 0.04166567577101579
+ },
+ "harness|mmlu_college_physics|5": {
+ "acc": 0.2549019607843137,
+ "acc_stderr": 0.043364327079931785,
+ "acc_norm": 0.2549019607843137,
+ "acc_norm_stderr": 0.043364327079931785
+ },
+ "harness|mmlu_high_school_microeconomics|5": {
+ "acc": 0.5756302521008403,
+ "acc_stderr": 0.03210479051015776,
+ "acc_norm": 0.5756302521008403,
+ "acc_norm_stderr": 0.03210479051015776
+ },
+ "harness|mmlu_high_school_macroeconomics|5": {
+ "acc": 0.541025641025641,
+ "acc_stderr": 0.025265525491284295,
+ "acc_norm": 0.541025641025641,
+ "acc_norm_stderr": 0.025265525491284295
+ },
+ "harness|mmlu_computer_security|5": {
+ "acc": 0.54,
+ "acc_stderr": 0.05009082659620332,
+ "acc_norm": 0.54,
+ "acc_norm_stderr": 0.05009082659620332
+ },
+ "harness|mmlu_global_facts|5": {
+ "acc": 0.31,
+ "acc_stderr": 0.04648231987117316,
+ "acc_norm": 0.31,
+ "acc_norm_stderr": 0.04648231987117316
+ },
+ "harness|mmlu_jurisprudence|5": {
+ "acc": 0.5555555555555556,
+ "acc_stderr": 0.04803752235190192,
+ "acc_norm": 0.5555555555555556,
+ "acc_norm_stderr": 0.04803752235190192
+ },
+ "harness|mmlu_high_school_chemistry|5": {
+ "acc": 0.3842364532019704,
+ "acc_stderr": 0.0342239856565755,
+ "acc_norm": 0.3842364532019704,
+ "acc_norm_stderr": 0.0342239856565755
+ },
+ "harness|mmlu_high_school_biology|5": {
+ "acc": 0.5774193548387097,
+ "acc_stderr": 0.02810096472427264,
+ "acc_norm": 0.5774193548387097,
+ "acc_norm_stderr": 0.02810096472427264
+ },
+ "harness|mmlu_marketing|5": {
+ "acc": 0.7777777777777778,
+ "acc_stderr": 0.027236013946196673,
+ "acc_norm": 0.7777777777777778,
+ "acc_norm_stderr": 0.027236013946196673
+ },
+ "harness|mmlu_clinical_knowledge|5": {
+ "acc": 0.4981132075471698,
+ "acc_stderr": 0.030772653642075657,
+ "acc_norm": 0.4981132075471698,
+ "acc_norm_stderr": 0.030772653642075657
+ },
+ "harness|mmlu_public_relations|5": {
+ "acc": 0.5272727272727272,
+ "acc_stderr": 0.04782001791380061,
+ "acc_norm": 0.5272727272727272,
+ "acc_norm_stderr": 0.04782001791380061
+ },
+ "harness|mmlu_high_school_mathematics|5": {
+ "acc": 0.25555555555555554,
+ "acc_stderr": 0.026593939101844082,
+ "acc_norm": 0.25555555555555554,
+ "acc_norm_stderr": 0.026593939101844082
+ },
+ "harness|mmlu_high_school_physics|5": {
+ "acc": 0.33774834437086093,
+ "acc_stderr": 0.038615575462551684,
+ "acc_norm": 0.33774834437086093,
+ "acc_norm_stderr": 0.038615575462551684
+ },
+ "harness|mmlu_sociology|5": {
+ "acc": 0.7064676616915423,
+ "acc_stderr": 0.032200241045342054,
+ "acc_norm": 0.7064676616915423,
+ "acc_norm_stderr": 0.032200241045342054
+ },
+ "harness|mmlu_college_medicine|5": {
+ "acc": 0.4797687861271676,
+ "acc_stderr": 0.03809342081273958,
+ "acc_norm": 0.4797687861271676,
+ "acc_norm_stderr": 0.03809342081273958
+ },
+ "harness|mmlu_elementary_mathematics|5": {
+ "acc": 0.38095238095238093,
+ "acc_stderr": 0.025010749116137602,
+ "acc_norm": 0.38095238095238093,
+ "acc_norm_stderr": 0.025010749116137602
+ },
+ "harness|mmlu_college_biology|5": {
+ "acc": 0.4236111111111111,
+ "acc_stderr": 0.041321250197233685,
+ "acc_norm": 0.4236111111111111,
+ "acc_norm_stderr": 0.041321250197233685
+ },
+ "harness|mmlu_college_chemistry|5": {
+ "acc": 0.31,
+ "acc_stderr": 0.04648231987117316,
+ "acc_norm": 0.31,
+ "acc_norm_stderr": 0.04648231987117316
+ },
+ "harness|mmlu_us_foreign_policy|5": {
+ "acc": 0.71,
+ "acc_stderr": 0.04560480215720683,
+ "acc_norm": 0.71,
+ "acc_norm_stderr": 0.04560480215720683
+ },
+ "harness|mmlu_moral_disputes|5": {
+ "acc": 0.5751445086705202,
+ "acc_stderr": 0.026613350840261733,
+ "acc_norm": 0.5751445086705202,
+ "acc_norm_stderr": 0.026613350840261733
+ },
+ "harness|mmlu_logical_fallacies|5": {
+ "acc": 0.5030674846625767,
+ "acc_stderr": 0.03928297078179662,
+ "acc_norm": 0.5030674846625767,
+ "acc_norm_stderr": 0.03928297078179662
+ },
+ "harness|mmlu_prehistory|5": {
+ "acc": 0.5370370370370371,
+ "acc_stderr": 0.027744313443376536,
+ "acc_norm": 0.5370370370370371,
+ "acc_norm_stderr": 0.027744313443376536
+ },
+ "harness|mmlu_college_mathematics|5": {
+ "acc": 0.33,
+ "acc_stderr": 0.04725815626252606,
+ "acc_norm": 0.33,
+ "acc_norm_stderr": 0.04725815626252606
+ },
+ "harness|mmlu_high_school_government_and_politics|5": {
+ "acc": 0.6217616580310881,
+ "acc_stderr": 0.034998072761933376,
+ "acc_norm": 0.6217616580310881,
+ "acc_norm_stderr": 0.034998072761933376
+ },
+ "harness|mmlu_econometrics|5": {
+ "acc": 0.37719298245614036,
+ "acc_stderr": 0.04559522141958216,
+ "acc_norm": 0.37719298245614036,
+ "acc_norm_stderr": 0.04559522141958216
+ },
+ "harness|mmlu_high_school_psychology|5": {
+ "acc": 0.6385321100917432,
+ "acc_stderr": 0.02059808200993736,
+ "acc_norm": 0.6385321100917432,
+ "acc_norm_stderr": 0.02059808200993736
+ },
+ "harness|mmlu_formal_logic|5": {
+ "acc": 0.4126984126984127,
+ "acc_stderr": 0.04403438954768177,
+ "acc_norm": 0.4126984126984127,
+ "acc_norm_stderr": 0.04403438954768177
+ },
+ "harness|mmlu_nutrition|5": {
+ "acc": 0.5261437908496732,
+ "acc_stderr": 0.028590752958852387,
+ "acc_norm": 0.5261437908496732,
+ "acc_norm_stderr": 0.028590752958852387
+ },
+ "harness|mmlu_business_ethics|5": {
+ "acc": 0.57,
+ "acc_stderr": 0.049756985195624284,
+ "acc_norm": 0.57,
+ "acc_norm_stderr": 0.049756985195624284
+ },
+ "harness|mmlu_international_law|5": {
+ "acc": 0.7520661157024794,
+ "acc_stderr": 0.03941897526516304,
+ "acc_norm": 0.7520661157024794,
+ "acc_norm_stderr": 0.03941897526516304
+ },
+ "harness|mmlu_astronomy|5": {
+ "acc": 0.5789473684210527,
+ "acc_stderr": 0.040179012759817494,
+ "acc_norm": 0.5789473684210527,
+ "acc_norm_stderr": 0.040179012759817494
+ },
+ "harness|mmlu_professional_psychology|5": {
+ "acc": 0.4738562091503268,
+ "acc_stderr": 0.020200164564804588,
+ "acc_norm": 0.4738562091503268,
+ "acc_norm_stderr": 0.020200164564804588
+ },
+ "harness|mmlu_professional_accounting|5": {
+ "acc": 0.3404255319148936,
+ "acc_stderr": 0.02826765748265013,
+ "acc_norm": 0.3404255319148936,
+ "acc_norm_stderr": 0.02826765748265013
+ },
+ "harness|mmlu_machine_learning|5": {
+ "acc": 0.38392857142857145,
+ "acc_stderr": 0.046161430750285455,
+ "acc_norm": 0.38392857142857145,
+ "acc_norm_stderr": 0.046161430750285455
+ },
+ "harness|mmlu_high_school_statistics|5": {
+ "acc": 0.4675925925925926,
+ "acc_stderr": 0.03402801581358966,
+ "acc_norm": 0.4675925925925926,
+ "acc_norm_stderr": 0.03402801581358966
+ },
+ "harness|mmlu_moral_scenarios|5": {
+ "acc": 0.21675977653631284,
+ "acc_stderr": 0.013780598486443363,
+ "acc_norm": 0.21675977653631284,
+ "acc_norm_stderr": 0.013780598486443363
+ },
+ "harness|mmlu_college_computer_science|5": {
+ "acc": 0.39,
+ "acc_stderr": 0.04902071300001975,
+ "acc_norm": 0.39,
+ "acc_norm_stderr": 0.04902071300001975
+ },
+ "harness|mmlu_high_school_computer_science|5": {
+ "acc": 0.71,
+ "acc_stderr": 0.04560480215720684,
+ "acc_norm": 0.71,
+ "acc_norm_stderr": 0.04560480215720684
+ },
+ "harness|mmlu_professional_medicine|5": {
+ "acc": 0.4411764705882353,
+ "acc_stderr": 0.0301619119307671,
+ "acc_norm": 0.4411764705882353,
+ "acc_norm_stderr": 0.0301619119307671
+ },
+ "harness|mmlu_security_studies|5": {
+ "acc": 0.6285714285714286,
+ "acc_stderr": 0.03093285879278986,
+ "acc_norm": 0.6285714285714286,
+ "acc_norm_stderr": 0.03093285879278986
+ },
+ "harness|mmlu_high_school_world_history|5": {
+ "acc": 0.70042194092827,
+ "acc_stderr": 0.029818024749753095,
+ "acc_norm": 0.70042194092827,
+ "acc_norm_stderr": 0.029818024749753095
+ },
+ "harness|mmlu_professional_law|5": {
+ "acc": 0.378748370273794,
+ "acc_stderr": 0.012389052105003741,
+ "acc_norm": 0.378748370273794,
+ "acc_norm_stderr": 0.012389052105003741
+ },
+ "harness|mmlu_high_school_us_history|5": {
+ "acc": 0.6225490196078431,
+ "acc_stderr": 0.03402272044340703,
+ "acc_norm": 0.6225490196078431,
+ "acc_norm_stderr": 0.03402272044340703
+ },
+ "harness|mmlu_high_school_european_history|5": {
+ "acc": 0.6666666666666666,
+ "acc_stderr": 0.03681050869161549,
+ "acc_norm": 0.6666666666666666,
+ "acc_norm_stderr": 0.03681050869161549
+ },
+ "harness|truthfulqa_mc|0": {
+ "mc1": 0.33659730722154224,
+ "mc1_stderr": 0.016542412809494877,
+ "mc2": 0.49995145184296846,
+ "mc2_stderr": 0.015887726098900913
+ },
+ "harness|commongen_v2|2": {
+ "acc": 0.564344746162928,
+ "acc_stderr": 0.017047415229476316,
+ "acc_norm": 0.6068476977567887,
+ "acc_norm_stderr": 0.016793262801287068
+ }
+ },
+ "versions": {
+ "all": 0,
+ "harness|arc_challenge|25": 0,
+ "harness|hellaswag|10": 0,
+ "harness|mmlu_world_religions|5": 1,
+ "harness|mmlu_management|5": 1,
+ "harness|mmlu_miscellaneous|5": 1,
+ "harness|mmlu_anatomy|5": 1,
+ "harness|mmlu_abstract_algebra|5": 1,
+ "harness|mmlu_conceptual_physics|5": 1,
+ "harness|mmlu_virology|5": 1,
+ "harness|mmlu_philosophy|5": 1,
+ "harness|mmlu_human_aging|5": 1,
+ "harness|mmlu_human_sexuality|5": 1,
+ "harness|mmlu_medical_genetics|5": 1,
+ "harness|mmlu_high_school_geography|5": 1,
+ "harness|mmlu_electrical_engineering|5": 1,
+ "harness|mmlu_college_physics|5": 1,
+ "harness|mmlu_high_school_microeconomics|5": 1,
+ "harness|mmlu_high_school_macroeconomics|5": 1,
+ "harness|mmlu_computer_security|5": 1,
+ "harness|mmlu_global_facts|5": 1,
+ "harness|mmlu_jurisprudence|5": 1,
+ "harness|mmlu_high_school_chemistry|5": 1,
+ "harness|mmlu_high_school_biology|5": 1,
+ "harness|mmlu_marketing|5": 1,
+ "harness|mmlu_clinical_knowledge|5": 1,
+ "harness|mmlu_public_relations|5": 1,
+ "harness|mmlu_high_school_mathematics|5": 1,
+ "harness|mmlu_high_school_physics|5": 1,
+ "harness|mmlu_sociology|5": 1,
+ "harness|mmlu_college_medicine|5": 1,
+ "harness|mmlu_elementary_mathematics|5": 1,
+ "harness|mmlu_college_biology|5": 1,
+ "harness|mmlu_college_chemistry|5": 1,
+ "harness|mmlu_us_foreign_policy|5": 1,
+ "harness|mmlu_moral_disputes|5": 1,
+ "harness|mmlu_logical_fallacies|5": 1,
+ "harness|mmlu_prehistory|5": 1,
+ "harness|mmlu_college_mathematics|5": 1,
+ "harness|mmlu_high_school_government_and_politics|5": 1,
+ "harness|mmlu_econometrics|5": 1,
+ "harness|mmlu_high_school_psychology|5": 1,
+ "harness|mmlu_formal_logic|5": 1,
+ "harness|mmlu_nutrition|5": 1,
+ "harness|mmlu_business_ethics|5": 1,
+ "harness|mmlu_international_law|5": 1,
+ "harness|mmlu_astronomy|5": 1,
+ "harness|mmlu_professional_psychology|5": 1,
+ "harness|mmlu_professional_accounting|5": 1,
+ "harness|mmlu_machine_learning|5": 1,
+ "harness|mmlu_high_school_statistics|5": 1,
+ "harness|mmlu_moral_scenarios|5": 1,
+ "harness|mmlu_college_computer_science|5": 1,
+ "harness|mmlu_high_school_computer_science|5": 1,
+ "harness|mmlu_professional_medicine|5": 1,
+ "harness|mmlu_security_studies|5": 1,
+ "harness|mmlu_high_school_world_history|5": 1,
+ "harness|mmlu_professional_law|5": 1,
+ "harness|mmlu_high_school_us_history|5": 1,
+ "harness|mmlu_high_school_european_history|5": 1,
+ "harness|truthfulqa_mc|0": 0,
+ "harness|commongen_v2|2": 1
+ },
+ "config_general": {
+ "model_name": "nlpai-lab/KULLM3",
+ "model_sha": "5a6bcd0fc7f240460eb6d57016f7b4060bc1f43b",
+ "model_dtype": "torch.float16",
+ "lighteval_sha": "",
+ "num_few_shot_default": 0,
+ "num_fewshot_seeds": 1,
+ "override_batch_size": 1,
+ "max_samples": null
+ }
+}
\ No newline at end of file
diff --git a/eval-results/x2bee/POLAR-14B-DPO-v1.3/result.json b/eval-results/x2bee/POLAR-14B-DPO-v1.3/result.json
new file mode 100644
index 0000000000000000000000000000000000000000..48ca4423df5ee9731f60f12785501c7159e68cf6
--- /dev/null
+++ b/eval-results/x2bee/POLAR-14B-DPO-v1.3/result.json
@@ -0,0 +1,450 @@
+{
+ "results": {
+ "daily": {
+ "daily": 4
+ },
+ "quarterly": {
+ "quarterly": 4
+ },
+ "harness|arc_challenge|25": {
+ "acc": 0.7465870307167235,
+ "acc_stderr": 0.012710896778378604,
+ "acc_norm": 0.7807167235494881,
+ "acc_norm_stderr": 0.012091245787615728
+ },
+ "harness|hellaswag|10": {
+ "acc": 0.6385182234614618,
+ "acc_stderr": 0.004794478426382617,
+ "acc_norm": 0.7561242780322645,
+ "acc_norm_stderr": 0.004285410130466119
+ },
+ "harness|mmlu_world_religions|5": {
+ "acc": 0.6900584795321637,
+ "acc_stderr": 0.035469769593931624,
+ "acc_norm": 0.6900584795321637,
+ "acc_norm_stderr": 0.035469769593931624
+ },
+ "harness|mmlu_management|5": {
+ "acc": 0.6601941747572816,
+ "acc_stderr": 0.046897659372781335,
+ "acc_norm": 0.6601941747572816,
+ "acc_norm_stderr": 0.046897659372781335
+ },
+ "harness|mmlu_miscellaneous|5": {
+ "acc": 0.6845466155810983,
+ "acc_stderr": 0.016617501738763408,
+ "acc_norm": 0.6845466155810983,
+ "acc_norm_stderr": 0.016617501738763408
+ },
+ "harness|mmlu_anatomy|5": {
+ "acc": 0.48148148148148145,
+ "acc_stderr": 0.04316378599511324,
+ "acc_norm": 0.48148148148148145,
+ "acc_norm_stderr": 0.04316378599511324
+ },
+ "harness|mmlu_abstract_algebra|5": {
+ "acc": 0.33,
+ "acc_stderr": 0.047258156262526045,
+ "acc_norm": 0.33,
+ "acc_norm_stderr": 0.047258156262526045
+ },
+ "harness|mmlu_conceptual_physics|5": {
+ "acc": 0.46808510638297873,
+ "acc_stderr": 0.03261936918467383,
+ "acc_norm": 0.46808510638297873,
+ "acc_norm_stderr": 0.03261936918467383
+ },
+ "harness|mmlu_virology|5": {
+ "acc": 0.4759036144578313,
+ "acc_stderr": 0.03887971849597264,
+ "acc_norm": 0.4759036144578313,
+ "acc_norm_stderr": 0.03887971849597264
+ },
+ "harness|mmlu_philosophy|5": {
+ "acc": 0.6334405144694534,
+ "acc_stderr": 0.02736807824397163,
+ "acc_norm": 0.6334405144694534,
+ "acc_norm_stderr": 0.02736807824397163
+ },
+ "harness|mmlu_human_aging|5": {
+ "acc": 0.6681614349775785,
+ "acc_stderr": 0.03160295143776679,
+ "acc_norm": 0.6681614349775785,
+ "acc_norm_stderr": 0.03160295143776679
+ },
+ "harness|mmlu_human_sexuality|5": {
+ "acc": 0.6030534351145038,
+ "acc_stderr": 0.04291135671009224,
+ "acc_norm": 0.6030534351145038,
+ "acc_norm_stderr": 0.04291135671009224
+ },
+ "harness|mmlu_medical_genetics|5": {
+ "acc": 0.51,
+ "acc_stderr": 0.05024183937956911,
+ "acc_norm": 0.51,
+ "acc_norm_stderr": 0.05024183937956911
+ },
+ "harness|mmlu_high_school_geography|5": {
+ "acc": 0.7222222222222222,
+ "acc_stderr": 0.03191178226713547,
+ "acc_norm": 0.7222222222222222,
+ "acc_norm_stderr": 0.03191178226713547
+ },
+ "harness|mmlu_electrical_engineering|5": {
+ "acc": 0.47586206896551725,
+ "acc_stderr": 0.041618085035015295,
+ "acc_norm": 0.47586206896551725,
+ "acc_norm_stderr": 0.041618085035015295
+ },
+ "harness|mmlu_college_physics|5": {
+ "acc": 0.2549019607843137,
+ "acc_stderr": 0.04336432707993178,
+ "acc_norm": 0.2549019607843137,
+ "acc_norm_stderr": 0.04336432707993178
+ },
+ "harness|mmlu_high_school_microeconomics|5": {
+ "acc": 0.592436974789916,
+ "acc_stderr": 0.031918633744784666,
+ "acc_norm": 0.592436974789916,
+ "acc_norm_stderr": 0.031918633744784666
+ },
+ "harness|mmlu_high_school_macroeconomics|5": {
+ "acc": 0.5948717948717949,
+ "acc_stderr": 0.024890471769938142,
+ "acc_norm": 0.5948717948717949,
+ "acc_norm_stderr": 0.024890471769938142
+ },
+ "harness|mmlu_computer_security|5": {
+ "acc": 0.66,
+ "acc_stderr": 0.04760952285695237,
+ "acc_norm": 0.66,
+ "acc_norm_stderr": 0.04760952285695237
+ },
+ "harness|mmlu_global_facts|5": {
+ "acc": 0.27,
+ "acc_stderr": 0.04461960433384739,
+ "acc_norm": 0.27,
+ "acc_norm_stderr": 0.04461960433384739
+ },
+ "harness|mmlu_jurisprudence|5": {
+ "acc": 0.6388888888888888,
+ "acc_stderr": 0.04643454608906275,
+ "acc_norm": 0.6388888888888888,
+ "acc_norm_stderr": 0.04643454608906275
+ },
+ "harness|mmlu_high_school_chemistry|5": {
+ "acc": 0.4433497536945813,
+ "acc_stderr": 0.034953345821629345,
+ "acc_norm": 0.4433497536945813,
+ "acc_norm_stderr": 0.034953345821629345
+ },
+ "harness|mmlu_high_school_biology|5": {
+ "acc": 0.5806451612903226,
+ "acc_stderr": 0.028071588901091838,
+ "acc_norm": 0.5806451612903226,
+ "acc_norm_stderr": 0.028071588901091838
+ },
+ "harness|mmlu_marketing|5": {
+ "acc": 0.811965811965812,
+ "acc_stderr": 0.025598193686652254,
+ "acc_norm": 0.811965811965812,
+ "acc_norm_stderr": 0.025598193686652254
+ },
+ "harness|mmlu_clinical_knowledge|5": {
+ "acc": 0.5169811320754717,
+ "acc_stderr": 0.030755120364119898,
+ "acc_norm": 0.5169811320754717,
+ "acc_norm_stderr": 0.030755120364119898
+ },
+ "harness|mmlu_public_relations|5": {
+ "acc": 0.5818181818181818,
+ "acc_stderr": 0.04724577405731573,
+ "acc_norm": 0.5818181818181818,
+ "acc_norm_stderr": 0.04724577405731573
+ },
+ "harness|mmlu_high_school_mathematics|5": {
+ "acc": 0.3888888888888889,
+ "acc_stderr": 0.029723278961476664,
+ "acc_norm": 0.3888888888888889,
+ "acc_norm_stderr": 0.029723278961476664
+ },
+ "harness|mmlu_high_school_physics|5": {
+ "acc": 0.3708609271523179,
+ "acc_stderr": 0.03943966699183629,
+ "acc_norm": 0.3708609271523179,
+ "acc_norm_stderr": 0.03943966699183629
+ },
+ "harness|mmlu_sociology|5": {
+ "acc": 0.6666666666666666,
+ "acc_stderr": 0.033333333333333326,
+ "acc_norm": 0.6666666666666666,
+ "acc_norm_stderr": 0.033333333333333326
+ },
+ "harness|mmlu_college_medicine|5": {
+ "acc": 0.47398843930635837,
+ "acc_stderr": 0.038073017265045125,
+ "acc_norm": 0.47398843930635837,
+ "acc_norm_stderr": 0.038073017265045125
+ },
+ "harness|mmlu_elementary_mathematics|5": {
+ "acc": 0.42328042328042326,
+ "acc_stderr": 0.025446365634406793,
+ "acc_norm": 0.42328042328042326,
+ "acc_norm_stderr": 0.025446365634406793
+ },
+ "harness|mmlu_college_biology|5": {
+ "acc": 0.5625,
+ "acc_stderr": 0.04148415739394154,
+ "acc_norm": 0.5625,
+ "acc_norm_stderr": 0.04148415739394154
+ },
+ "harness|mmlu_college_chemistry|5": {
+ "acc": 0.39,
+ "acc_stderr": 0.04902071300001975,
+ "acc_norm": 0.39,
+ "acc_norm_stderr": 0.04902071300001975
+ },
+ "harness|mmlu_us_foreign_policy|5": {
+ "acc": 0.78,
+ "acc_stderr": 0.04163331998932263,
+ "acc_norm": 0.78,
+ "acc_norm_stderr": 0.04163331998932263
+ },
+ "harness|mmlu_moral_disputes|5": {
+ "acc": 0.5491329479768786,
+ "acc_stderr": 0.026788811931562767,
+ "acc_norm": 0.5491329479768786,
+ "acc_norm_stderr": 0.026788811931562767
+ },
+ "harness|mmlu_logical_fallacies|5": {
+ "acc": 0.6319018404907976,
+ "acc_stderr": 0.03789213935838396,
+ "acc_norm": 0.6319018404907976,
+ "acc_norm_stderr": 0.03789213935838396
+ },
+ "harness|mmlu_prehistory|5": {
+ "acc": 0.5925925925925926,
+ "acc_stderr": 0.02733954664066273,
+ "acc_norm": 0.5925925925925926,
+ "acc_norm_stderr": 0.02733954664066273
+ },
+ "harness|mmlu_college_mathematics|5": {
+ "acc": 0.4,
+ "acc_stderr": 0.049236596391733084,
+ "acc_norm": 0.4,
+ "acc_norm_stderr": 0.049236596391733084
+ },
+ "harness|mmlu_high_school_government_and_politics|5": {
+ "acc": 0.7668393782383419,
+ "acc_stderr": 0.03051611137147601,
+ "acc_norm": 0.7668393782383419,
+ "acc_norm_stderr": 0.03051611137147601
+ },
+ "harness|mmlu_econometrics|5": {
+ "acc": 0.4473684210526316,
+ "acc_stderr": 0.046774730044912,
+ "acc_norm": 0.4473684210526316,
+ "acc_norm_stderr": 0.046774730044912
+ },
+ "harness|mmlu_high_school_psychology|5": {
+ "acc": 0.726605504587156,
+ "acc_stderr": 0.01910929984609827,
+ "acc_norm": 0.726605504587156,
+ "acc_norm_stderr": 0.01910929984609827
+ },
+ "harness|mmlu_formal_logic|5": {
+ "acc": 0.3968253968253968,
+ "acc_stderr": 0.04375888492727061,
+ "acc_norm": 0.3968253968253968,
+ "acc_norm_stderr": 0.04375888492727061
+ },
+ "harness|mmlu_nutrition|5": {
+ "acc": 0.6078431372549019,
+ "acc_stderr": 0.027956046165424516,
+ "acc_norm": 0.6078431372549019,
+ "acc_norm_stderr": 0.027956046165424516
+ },
+ "harness|mmlu_business_ethics|5": {
+ "acc": 0.55,
+ "acc_stderr": 0.05,
+ "acc_norm": 0.55,
+ "acc_norm_stderr": 0.05
+ },
+ "harness|mmlu_international_law|5": {
+ "acc": 0.6942148760330579,
+ "acc_stderr": 0.04205953933884122,
+ "acc_norm": 0.6942148760330579,
+ "acc_norm_stderr": 0.04205953933884122
+ },
+ "harness|mmlu_astronomy|5": {
+ "acc": 0.618421052631579,
+ "acc_stderr": 0.03953173377749194,
+ "acc_norm": 0.618421052631579,
+ "acc_norm_stderr": 0.03953173377749194
+ },
+ "harness|mmlu_professional_psychology|5": {
+ "acc": 0.5669934640522876,
+ "acc_stderr": 0.02004544247332422,
+ "acc_norm": 0.5669934640522876,
+ "acc_norm_stderr": 0.02004544247332422
+ },
+ "harness|mmlu_professional_accounting|5": {
+ "acc": 0.4219858156028369,
+ "acc_stderr": 0.029462189233370586,
+ "acc_norm": 0.4219858156028369,
+ "acc_norm_stderr": 0.029462189233370586
+ },
+ "harness|mmlu_machine_learning|5": {
+ "acc": 0.5089285714285714,
+ "acc_stderr": 0.04745033255489123,
+ "acc_norm": 0.5089285714285714,
+ "acc_norm_stderr": 0.04745033255489123
+ },
+ "harness|mmlu_high_school_statistics|5": {
+ "acc": 0.4351851851851852,
+ "acc_stderr": 0.03381200005643526,
+ "acc_norm": 0.4351851851851852,
+ "acc_norm_stderr": 0.03381200005643526
+ },
+ "harness|mmlu_moral_scenarios|5": {
+ "acc": 0.3787709497206704,
+ "acc_stderr": 0.016223533510365117,
+ "acc_norm": 0.3787709497206704,
+ "acc_norm_stderr": 0.016223533510365117
+ },
+ "harness|mmlu_college_computer_science|5": {
+ "acc": 0.47,
+ "acc_stderr": 0.05016135580465919,
+ "acc_norm": 0.47,
+ "acc_norm_stderr": 0.05016135580465919
+ },
+ "harness|mmlu_high_school_computer_science|5": {
+ "acc": 0.66,
+ "acc_stderr": 0.04760952285695238,
+ "acc_norm": 0.66,
+ "acc_norm_stderr": 0.04760952285695238
+ },
+ "harness|mmlu_professional_medicine|5": {
+ "acc": 0.48161764705882354,
+ "acc_stderr": 0.03035230339535196,
+ "acc_norm": 0.48161764705882354,
+ "acc_norm_stderr": 0.03035230339535196
+ },
+ "harness|mmlu_security_studies|5": {
+ "acc": 0.6448979591836734,
+ "acc_stderr": 0.030635655150387634,
+ "acc_norm": 0.6448979591836734,
+ "acc_norm_stderr": 0.030635655150387634
+ },
+ "harness|mmlu_high_school_world_history|5": {
+ "acc": 0.729957805907173,
+ "acc_stderr": 0.028900721906293426,
+ "acc_norm": 0.729957805907173,
+ "acc_norm_stderr": 0.028900721906293426
+ },
+ "harness|mmlu_professional_law|5": {
+ "acc": 0.41460234680573665,
+ "acc_stderr": 0.012582597058908284,
+ "acc_norm": 0.41460234680573665,
+ "acc_norm_stderr": 0.012582597058908284
+ },
+ "harness|mmlu_high_school_us_history|5": {
+ "acc": 0.6421568627450981,
+ "acc_stderr": 0.03364487286088298,
+ "acc_norm": 0.6421568627450981,
+ "acc_norm_stderr": 0.03364487286088298
+ },
+ "harness|mmlu_high_school_european_history|5": {
+ "acc": 0.6181818181818182,
+ "acc_stderr": 0.03793713171165635,
+ "acc_norm": 0.6181818181818182,
+ "acc_norm_stderr": 0.03793713171165635
+ },
+ "harness|truthfulqa_mc|0": {
+ "mc1": 0.6328029375764994,
+ "mc1_stderr": 0.01687480500145318,
+ "mc2": 0.7522925779273922,
+ "mc2_stderr": 0.014568927682929578
+ },
+ "harness|commongen_v2|2": {
+ "acc": 0.45218417945690675,
+ "acc_stderr": 0.017111567130916785,
+ "acc_norm": 0.45454545454545453,
+ "acc_norm_stderr": 0.017119172208061504
+ }
+ },
+ "versions": {
+ "all": 0,
+ "harness|arc_challenge|25": 0,
+ "harness|hellaswag|10": 0,
+ "harness|mmlu_world_religions|5": 1,
+ "harness|mmlu_management|5": 1,
+ "harness|mmlu_miscellaneous|5": 1,
+ "harness|mmlu_anatomy|5": 1,
+ "harness|mmlu_abstract_algebra|5": 1,
+ "harness|mmlu_conceptual_physics|5": 1,
+ "harness|mmlu_virology|5": 1,
+ "harness|mmlu_philosophy|5": 1,
+ "harness|mmlu_human_aging|5": 1,
+ "harness|mmlu_human_sexuality|5": 1,
+ "harness|mmlu_medical_genetics|5": 1,
+ "harness|mmlu_high_school_geography|5": 1,
+ "harness|mmlu_electrical_engineering|5": 1,
+ "harness|mmlu_college_physics|5": 1,
+ "harness|mmlu_high_school_microeconomics|5": 1,
+ "harness|mmlu_high_school_macroeconomics|5": 1,
+ "harness|mmlu_computer_security|5": 1,
+ "harness|mmlu_global_facts|5": 1,
+ "harness|mmlu_jurisprudence|5": 1,
+ "harness|mmlu_high_school_chemistry|5": 1,
+ "harness|mmlu_high_school_biology|5": 1,
+ "harness|mmlu_marketing|5": 1,
+ "harness|mmlu_clinical_knowledge|5": 1,
+ "harness|mmlu_public_relations|5": 1,
+ "harness|mmlu_high_school_mathematics|5": 1,
+ "harness|mmlu_high_school_physics|5": 1,
+ "harness|mmlu_sociology|5": 1,
+ "harness|mmlu_college_medicine|5": 1,
+ "harness|mmlu_elementary_mathematics|5": 1,
+ "harness|mmlu_college_biology|5": 1,
+ "harness|mmlu_college_chemistry|5": 1,
+ "harness|mmlu_us_foreign_policy|5": 1,
+ "harness|mmlu_moral_disputes|5": 1,
+ "harness|mmlu_logical_fallacies|5": 1,
+ "harness|mmlu_prehistory|5": 1,
+ "harness|mmlu_college_mathematics|5": 1,
+ "harness|mmlu_high_school_government_and_politics|5": 1,
+ "harness|mmlu_econometrics|5": 1,
+ "harness|mmlu_high_school_psychology|5": 1,
+ "harness|mmlu_formal_logic|5": 1,
+ "harness|mmlu_nutrition|5": 1,
+ "harness|mmlu_business_ethics|5": 1,
+ "harness|mmlu_international_law|5": 1,
+ "harness|mmlu_astronomy|5": 1,
+ "harness|mmlu_professional_psychology|5": 1,
+ "harness|mmlu_professional_accounting|5": 1,
+ "harness|mmlu_machine_learning|5": 1,
+ "harness|mmlu_high_school_statistics|5": 1,
+ "harness|mmlu_moral_scenarios|5": 1,
+ "harness|mmlu_college_computer_science|5": 1,
+ "harness|mmlu_high_school_computer_science|5": 1,
+ "harness|mmlu_professional_medicine|5": 1,
+ "harness|mmlu_security_studies|5": 1,
+ "harness|mmlu_high_school_world_history|5": 1,
+ "harness|mmlu_professional_law|5": 1,
+ "harness|mmlu_high_school_us_history|5": 1,
+ "harness|mmlu_high_school_european_history|5": 1,
+ "harness|truthfulqa_mc|0": 0,
+ "harness|commongen_v2|2": 1
+ },
+ "config_general": {
+ "model_name": "x2bee/POLAR-14B-DPO-v1.3",
+ "model_sha": "337edbed4c86db2da27e3b0e07086134f8d27a09",
+ "model_dtype": "torch.float16",
+ "lighteval_sha": "",
+ "num_few_shot_default": 0,
+ "num_fewshot_seeds": 1,
+ "override_batch_size": 1,
+ "max_samples": null
+ }
+}
\ No newline at end of file
diff --git a/eval-results/x2bee/POLAR-14B-DPO-v1.4/result.json b/eval-results/x2bee/POLAR-14B-DPO-v1.4/result.json
new file mode 100644
index 0000000000000000000000000000000000000000..578117675d4a1e9aaa0b0ef30e18ebbfc8fd720d
--- /dev/null
+++ b/eval-results/x2bee/POLAR-14B-DPO-v1.4/result.json
@@ -0,0 +1,450 @@
+{
+ "results": {
+ "daily": {
+ "daily": 7
+ },
+ "quarterly": {
+ "quarterly": 7
+ },
+ "harness|arc_challenge|25": {
+ "acc": 0.7363481228668942,
+ "acc_stderr": 0.012875929151297058,
+ "acc_norm": 0.7491467576791809,
+ "acc_norm_stderr": 0.012668198621315433
+ },
+ "harness|hellaswag|10": {
+ "acc": 0.7228639713204541,
+ "acc_stderr": 0.004466695023677848,
+ "acc_norm": 0.7422824138617805,
+ "acc_norm_stderr": 0.004364838000335614
+ },
+ "harness|mmlu_world_religions|5": {
+ "acc": 0.6140350877192983,
+ "acc_stderr": 0.03733756969066164,
+ "acc_norm": 0.6140350877192983,
+ "acc_norm_stderr": 0.03733756969066164
+ },
+ "harness|mmlu_management|5": {
+ "acc": 0.6893203883495146,
+ "acc_stderr": 0.045821241601615506,
+ "acc_norm": 0.6893203883495146,
+ "acc_norm_stderr": 0.045821241601615506
+ },
+ "harness|mmlu_miscellaneous|5": {
+ "acc": 0.6526181353767561,
+ "acc_stderr": 0.017026671748655728,
+ "acc_norm": 0.6526181353767561,
+ "acc_norm_stderr": 0.017026671748655728
+ },
+ "harness|mmlu_anatomy|5": {
+ "acc": 0.5037037037037037,
+ "acc_stderr": 0.043192236258113324,
+ "acc_norm": 0.5037037037037037,
+ "acc_norm_stderr": 0.043192236258113324
+ },
+ "harness|mmlu_abstract_algebra|5": {
+ "acc": 0.37,
+ "acc_stderr": 0.048523658709391,
+ "acc_norm": 0.37,
+ "acc_norm_stderr": 0.048523658709391
+ },
+ "harness|mmlu_conceptual_physics|5": {
+ "acc": 0.451063829787234,
+ "acc_stderr": 0.032529096196131965,
+ "acc_norm": 0.451063829787234,
+ "acc_norm_stderr": 0.032529096196131965
+ },
+ "harness|mmlu_virology|5": {
+ "acc": 0.4939759036144578,
+ "acc_stderr": 0.03892212195333045,
+ "acc_norm": 0.4939759036144578,
+ "acc_norm_stderr": 0.03892212195333045
+ },
+ "harness|mmlu_philosophy|5": {
+ "acc": 0.5852090032154341,
+ "acc_stderr": 0.02798268045975956,
+ "acc_norm": 0.5852090032154341,
+ "acc_norm_stderr": 0.02798268045975956
+ },
+ "harness|mmlu_human_aging|5": {
+ "acc": 0.6412556053811659,
+ "acc_stderr": 0.032190792004199956,
+ "acc_norm": 0.6412556053811659,
+ "acc_norm_stderr": 0.032190792004199956
+ },
+ "harness|mmlu_human_sexuality|5": {
+ "acc": 0.5954198473282443,
+ "acc_stderr": 0.043046937953806645,
+ "acc_norm": 0.5954198473282443,
+ "acc_norm_stderr": 0.043046937953806645
+ },
+ "harness|mmlu_medical_genetics|5": {
+ "acc": 0.47,
+ "acc_stderr": 0.05016135580465919,
+ "acc_norm": 0.47,
+ "acc_norm_stderr": 0.05016135580465919
+ },
+ "harness|mmlu_high_school_geography|5": {
+ "acc": 0.6616161616161617,
+ "acc_stderr": 0.033711241426263014,
+ "acc_norm": 0.6616161616161617,
+ "acc_norm_stderr": 0.033711241426263014
+ },
+ "harness|mmlu_electrical_engineering|5": {
+ "acc": 0.4827586206896552,
+ "acc_stderr": 0.041641887201693775,
+ "acc_norm": 0.4827586206896552,
+ "acc_norm_stderr": 0.041641887201693775
+ },
+ "harness|mmlu_college_physics|5": {
+ "acc": 0.2549019607843137,
+ "acc_stderr": 0.04336432707993178,
+ "acc_norm": 0.2549019607843137,
+ "acc_norm_stderr": 0.04336432707993178
+ },
+ "harness|mmlu_high_school_microeconomics|5": {
+ "acc": 0.5882352941176471,
+ "acc_stderr": 0.031968769891957786,
+ "acc_norm": 0.5882352941176471,
+ "acc_norm_stderr": 0.031968769891957786
+ },
+ "harness|mmlu_high_school_macroeconomics|5": {
+ "acc": 0.6025641025641025,
+ "acc_stderr": 0.024811920017903836,
+ "acc_norm": 0.6025641025641025,
+ "acc_norm_stderr": 0.024811920017903836
+ },
+ "harness|mmlu_computer_security|5": {
+ "acc": 0.66,
+ "acc_stderr": 0.04760952285695237,
+ "acc_norm": 0.66,
+ "acc_norm_stderr": 0.04760952285695237
+ },
+ "harness|mmlu_global_facts|5": {
+ "acc": 0.35,
+ "acc_stderr": 0.047937248544110196,
+ "acc_norm": 0.35,
+ "acc_norm_stderr": 0.047937248544110196
+ },
+ "harness|mmlu_jurisprudence|5": {
+ "acc": 0.5925925925925926,
+ "acc_stderr": 0.04750077341199984,
+ "acc_norm": 0.5925925925925926,
+ "acc_norm_stderr": 0.04750077341199984
+ },
+ "harness|mmlu_high_school_chemistry|5": {
+ "acc": 0.43842364532019706,
+ "acc_stderr": 0.03491207857486518,
+ "acc_norm": 0.43842364532019706,
+ "acc_norm_stderr": 0.03491207857486518
+ },
+ "harness|mmlu_high_school_biology|5": {
+ "acc": 0.567741935483871,
+ "acc_stderr": 0.028181739720019413,
+ "acc_norm": 0.567741935483871,
+ "acc_norm_stderr": 0.028181739720019413
+ },
+ "harness|mmlu_marketing|5": {
+ "acc": 0.7948717948717948,
+ "acc_stderr": 0.026453508054040356,
+ "acc_norm": 0.7948717948717948,
+ "acc_norm_stderr": 0.026453508054040356
+ },
+ "harness|mmlu_clinical_knowledge|5": {
+ "acc": 0.5169811320754717,
+ "acc_stderr": 0.030755120364119905,
+ "acc_norm": 0.5169811320754717,
+ "acc_norm_stderr": 0.030755120364119905
+ },
+ "harness|mmlu_public_relations|5": {
+ "acc": 0.5727272727272728,
+ "acc_stderr": 0.047381987035454834,
+ "acc_norm": 0.5727272727272728,
+ "acc_norm_stderr": 0.047381987035454834
+ },
+ "harness|mmlu_high_school_mathematics|5": {
+ "acc": 0.3962962962962963,
+ "acc_stderr": 0.029822619458533997,
+ "acc_norm": 0.3962962962962963,
+ "acc_norm_stderr": 0.029822619458533997
+ },
+ "harness|mmlu_high_school_physics|5": {
+ "acc": 0.3708609271523179,
+ "acc_stderr": 0.03943966699183629,
+ "acc_norm": 0.3708609271523179,
+ "acc_norm_stderr": 0.03943966699183629
+ },
+ "harness|mmlu_sociology|5": {
+ "acc": 0.6766169154228856,
+ "acc_stderr": 0.03307615947979035,
+ "acc_norm": 0.6766169154228856,
+ "acc_norm_stderr": 0.03307615947979035
+ },
+ "harness|mmlu_college_medicine|5": {
+ "acc": 0.49710982658959535,
+ "acc_stderr": 0.038124005659748335,
+ "acc_norm": 0.49710982658959535,
+ "acc_norm_stderr": 0.038124005659748335
+ },
+ "harness|mmlu_elementary_mathematics|5": {
+ "acc": 0.42592592592592593,
+ "acc_stderr": 0.02546714904546955,
+ "acc_norm": 0.42592592592592593,
+ "acc_norm_stderr": 0.02546714904546955
+ },
+ "harness|mmlu_college_biology|5": {
+ "acc": 0.5555555555555556,
+ "acc_stderr": 0.04155319955593146,
+ "acc_norm": 0.5555555555555556,
+ "acc_norm_stderr": 0.04155319955593146
+ },
+ "harness|mmlu_college_chemistry|5": {
+ "acc": 0.4,
+ "acc_stderr": 0.04923659639173309,
+ "acc_norm": 0.4,
+ "acc_norm_stderr": 0.04923659639173309
+ },
+ "harness|mmlu_us_foreign_policy|5": {
+ "acc": 0.73,
+ "acc_stderr": 0.044619604333847394,
+ "acc_norm": 0.73,
+ "acc_norm_stderr": 0.044619604333847394
+ },
+ "harness|mmlu_moral_disputes|5": {
+ "acc": 0.5549132947976878,
+ "acc_stderr": 0.02675625512966377,
+ "acc_norm": 0.5549132947976878,
+ "acc_norm_stderr": 0.02675625512966377
+ },
+ "harness|mmlu_logical_fallacies|5": {
+ "acc": 0.588957055214724,
+ "acc_stderr": 0.038656978537853624,
+ "acc_norm": 0.588957055214724,
+ "acc_norm_stderr": 0.038656978537853624
+ },
+ "harness|mmlu_prehistory|5": {
+ "acc": 0.5771604938271605,
+ "acc_stderr": 0.027487472980871595,
+ "acc_norm": 0.5771604938271605,
+ "acc_norm_stderr": 0.027487472980871595
+ },
+ "harness|mmlu_college_mathematics|5": {
+ "acc": 0.4,
+ "acc_stderr": 0.049236596391733084,
+ "acc_norm": 0.4,
+ "acc_norm_stderr": 0.049236596391733084
+ },
+ "harness|mmlu_high_school_government_and_politics|5": {
+ "acc": 0.7305699481865285,
+ "acc_stderr": 0.032018671228777947,
+ "acc_norm": 0.7305699481865285,
+ "acc_norm_stderr": 0.032018671228777947
+ },
+ "harness|mmlu_econometrics|5": {
+ "acc": 0.42105263157894735,
+ "acc_stderr": 0.046446020912223177,
+ "acc_norm": 0.42105263157894735,
+ "acc_norm_stderr": 0.046446020912223177
+ },
+ "harness|mmlu_high_school_psychology|5": {
+ "acc": 0.7064220183486238,
+ "acc_stderr": 0.019525151122639663,
+ "acc_norm": 0.7064220183486238,
+ "acc_norm_stderr": 0.019525151122639663
+ },
+ "harness|mmlu_formal_logic|5": {
+ "acc": 0.3968253968253968,
+ "acc_stderr": 0.04375888492727061,
+ "acc_norm": 0.3968253968253968,
+ "acc_norm_stderr": 0.04375888492727061
+ },
+ "harness|mmlu_nutrition|5": {
+ "acc": 0.545751633986928,
+ "acc_stderr": 0.02850980780262659,
+ "acc_norm": 0.545751633986928,
+ "acc_norm_stderr": 0.02850980780262659
+ },
+ "harness|mmlu_business_ethics|5": {
+ "acc": 0.55,
+ "acc_stderr": 0.05000000000000001,
+ "acc_norm": 0.55,
+ "acc_norm_stderr": 0.05000000000000001
+ },
+ "harness|mmlu_international_law|5": {
+ "acc": 0.6859504132231405,
+ "acc_stderr": 0.04236964753041019,
+ "acc_norm": 0.6859504132231405,
+ "acc_norm_stderr": 0.04236964753041019
+ },
+ "harness|mmlu_astronomy|5": {
+ "acc": 0.6052631578947368,
+ "acc_stderr": 0.039777499346220734,
+ "acc_norm": 0.6052631578947368,
+ "acc_norm_stderr": 0.039777499346220734
+ },
+ "harness|mmlu_professional_psychology|5": {
+ "acc": 0.5392156862745098,
+ "acc_stderr": 0.02016552331390791,
+ "acc_norm": 0.5392156862745098,
+ "acc_norm_stderr": 0.02016552331390791
+ },
+ "harness|mmlu_professional_accounting|5": {
+ "acc": 0.35815602836879434,
+ "acc_stderr": 0.02860208586275942,
+ "acc_norm": 0.35815602836879434,
+ "acc_norm_stderr": 0.02860208586275942
+ },
+ "harness|mmlu_machine_learning|5": {
+ "acc": 0.4107142857142857,
+ "acc_stderr": 0.04669510663875192,
+ "acc_norm": 0.4107142857142857,
+ "acc_norm_stderr": 0.04669510663875192
+ },
+ "harness|mmlu_high_school_statistics|5": {
+ "acc": 0.44907407407407407,
+ "acc_stderr": 0.03392238405321617,
+ "acc_norm": 0.44907407407407407,
+ "acc_norm_stderr": 0.03392238405321617
+ },
+ "harness|mmlu_moral_scenarios|5": {
+ "acc": 0.3452513966480447,
+ "acc_stderr": 0.015901432608930354,
+ "acc_norm": 0.3452513966480447,
+ "acc_norm_stderr": 0.015901432608930354
+ },
+ "harness|mmlu_college_computer_science|5": {
+ "acc": 0.43,
+ "acc_stderr": 0.049756985195624284,
+ "acc_norm": 0.43,
+ "acc_norm_stderr": 0.049756985195624284
+ },
+ "harness|mmlu_high_school_computer_science|5": {
+ "acc": 0.66,
+ "acc_stderr": 0.04760952285695238,
+ "acc_norm": 0.66,
+ "acc_norm_stderr": 0.04760952285695238
+ },
+ "harness|mmlu_professional_medicine|5": {
+ "acc": 0.45588235294117646,
+ "acc_stderr": 0.030254372573976694,
+ "acc_norm": 0.45588235294117646,
+ "acc_norm_stderr": 0.030254372573976694
+ },
+ "harness|mmlu_security_studies|5": {
+ "acc": 0.6204081632653061,
+ "acc_stderr": 0.031067211262872457,
+ "acc_norm": 0.6204081632653061,
+ "acc_norm_stderr": 0.031067211262872457
+ },
+ "harness|mmlu_high_school_world_history|5": {
+ "acc": 0.6582278481012658,
+ "acc_stderr": 0.030874537537553617,
+ "acc_norm": 0.6582278481012658,
+ "acc_norm_stderr": 0.030874537537553617
+ },
+ "harness|mmlu_professional_law|5": {
+ "acc": 0.4152542372881356,
+ "acc_stderr": 0.012585471793400667,
+ "acc_norm": 0.4152542372881356,
+ "acc_norm_stderr": 0.012585471793400667
+ },
+ "harness|mmlu_high_school_us_history|5": {
+ "acc": 0.5343137254901961,
+ "acc_stderr": 0.03501038327635896,
+ "acc_norm": 0.5343137254901961,
+ "acc_norm_stderr": 0.03501038327635896
+ },
+ "harness|mmlu_high_school_european_history|5": {
+ "acc": 0.5454545454545454,
+ "acc_stderr": 0.038881769216741004,
+ "acc_norm": 0.5454545454545454,
+ "acc_norm_stderr": 0.038881769216741004
+ },
+ "harness|truthfulqa_mc|0": {
+ "mc1": 0.4663402692778458,
+ "mc1_stderr": 0.01746379386716811,
+ "mc2": NaN,
+ "mc2_stderr": NaN
+ },
+ "harness|commongen_v2|2": {
+ "acc": 0.44037780401416765,
+ "acc_stderr": 0.01706769977431298,
+ "acc_norm": 0.44510035419126326,
+ "acc_norm_stderr": 0.01708641743100547
+ }
+ },
+ "versions": {
+ "all": 0,
+ "harness|arc_challenge|25": 0,
+ "harness|hellaswag|10": 0,
+ "harness|mmlu_world_religions|5": 1,
+ "harness|mmlu_management|5": 1,
+ "harness|mmlu_miscellaneous|5": 1,
+ "harness|mmlu_anatomy|5": 1,
+ "harness|mmlu_abstract_algebra|5": 1,
+ "harness|mmlu_conceptual_physics|5": 1,
+ "harness|mmlu_virology|5": 1,
+ "harness|mmlu_philosophy|5": 1,
+ "harness|mmlu_human_aging|5": 1,
+ "harness|mmlu_human_sexuality|5": 1,
+ "harness|mmlu_medical_genetics|5": 1,
+ "harness|mmlu_high_school_geography|5": 1,
+ "harness|mmlu_electrical_engineering|5": 1,
+ "harness|mmlu_college_physics|5": 1,
+ "harness|mmlu_high_school_microeconomics|5": 1,
+ "harness|mmlu_high_school_macroeconomics|5": 1,
+ "harness|mmlu_computer_security|5": 1,
+ "harness|mmlu_global_facts|5": 1,
+ "harness|mmlu_jurisprudence|5": 1,
+ "harness|mmlu_high_school_chemistry|5": 1,
+ "harness|mmlu_high_school_biology|5": 1,
+ "harness|mmlu_marketing|5": 1,
+ "harness|mmlu_clinical_knowledge|5": 1,
+ "harness|mmlu_public_relations|5": 1,
+ "harness|mmlu_high_school_mathematics|5": 1,
+ "harness|mmlu_high_school_physics|5": 1,
+ "harness|mmlu_sociology|5": 1,
+ "harness|mmlu_college_medicine|5": 1,
+ "harness|mmlu_elementary_mathematics|5": 1,
+ "harness|mmlu_college_biology|5": 1,
+ "harness|mmlu_college_chemistry|5": 1,
+ "harness|mmlu_us_foreign_policy|5": 1,
+ "harness|mmlu_moral_disputes|5": 1,
+ "harness|mmlu_logical_fallacies|5": 1,
+ "harness|mmlu_prehistory|5": 1,
+ "harness|mmlu_college_mathematics|5": 1,
+ "harness|mmlu_high_school_government_and_politics|5": 1,
+ "harness|mmlu_econometrics|5": 1,
+ "harness|mmlu_high_school_psychology|5": 1,
+ "harness|mmlu_formal_logic|5": 1,
+ "harness|mmlu_nutrition|5": 1,
+ "harness|mmlu_business_ethics|5": 1,
+ "harness|mmlu_international_law|5": 1,
+ "harness|mmlu_astronomy|5": 1,
+ "harness|mmlu_professional_psychology|5": 1,
+ "harness|mmlu_professional_accounting|5": 1,
+ "harness|mmlu_machine_learning|5": 1,
+ "harness|mmlu_high_school_statistics|5": 1,
+ "harness|mmlu_moral_scenarios|5": 1,
+ "harness|mmlu_college_computer_science|5": 1,
+ "harness|mmlu_high_school_computer_science|5": 1,
+ "harness|mmlu_professional_medicine|5": 1,
+ "harness|mmlu_security_studies|5": 1,
+ "harness|mmlu_high_school_world_history|5": 1,
+ "harness|mmlu_professional_law|5": 1,
+ "harness|mmlu_high_school_us_history|5": 1,
+ "harness|mmlu_high_school_european_history|5": 1,
+ "harness|truthfulqa_mc|0": 0,
+ "harness|commongen_v2|2": 1
+ },
+ "config_general": {
+ "model_name": "x2bee/POLAR-14B-DPO-v1.4",
+ "model_sha": "a6e64075fafaa3d5e393ff89c3cb26f9615e6de9",
+ "model_dtype": "torch.float16",
+ "lighteval_sha": "",
+ "num_few_shot_default": 0,
+ "num_fewshot_seeds": 1,
+ "override_batch_size": 1,
+ "max_samples": null
+ }
+}
\ No newline at end of file
diff --git a/eval-results/x2bee/POLAR-14B-HES-DPO-v1.5/result.json b/eval-results/x2bee/POLAR-14B-HES-DPO-v1.5/result.json
new file mode 100644
index 0000000000000000000000000000000000000000..8347643b533d653103e30f952f4381c7265e6bee
--- /dev/null
+++ b/eval-results/x2bee/POLAR-14B-HES-DPO-v1.5/result.json
@@ -0,0 +1,450 @@
+{
+ "results": {
+ "daily": {
+ "daily": 5
+ },
+ "quarterly": {
+ "quarterly": 5
+ },
+ "harness|arc_challenge|25": {
+ "acc": 0.6638225255972696,
+ "acc_stderr": 0.013804855026205756,
+ "acc_norm": 0.7278156996587031,
+ "acc_norm_stderr": 0.013006600406423709
+ },
+ "harness|hellaswag|10": {
+ "acc": 0.45648277235610435,
+ "acc_stderr": 0.004970846697552306,
+ "acc_norm": 0.6349332802230632,
+ "acc_norm_stderr": 0.004804649197163697
+ },
+ "harness|mmlu_world_religions|5": {
+ "acc": 0.7309941520467836,
+ "acc_stderr": 0.0340105262010409,
+ "acc_norm": 0.7309941520467836,
+ "acc_norm_stderr": 0.0340105262010409
+ },
+ "harness|mmlu_management|5": {
+ "acc": 0.7766990291262136,
+ "acc_stderr": 0.04123553189891431,
+ "acc_norm": 0.7766990291262136,
+ "acc_norm_stderr": 0.04123553189891431
+ },
+ "harness|mmlu_miscellaneous|5": {
+ "acc": 0.7343550446998723,
+ "acc_stderr": 0.01579430248788872,
+ "acc_norm": 0.7343550446998723,
+ "acc_norm_stderr": 0.01579430248788872
+ },
+ "harness|mmlu_anatomy|5": {
+ "acc": 0.45185185185185184,
+ "acc_stderr": 0.04299268905480863,
+ "acc_norm": 0.45185185185185184,
+ "acc_norm_stderr": 0.04299268905480863
+ },
+ "harness|mmlu_abstract_algebra|5": {
+ "acc": 0.35,
+ "acc_stderr": 0.04793724854411019,
+ "acc_norm": 0.35,
+ "acc_norm_stderr": 0.04793724854411019
+ },
+ "harness|mmlu_conceptual_physics|5": {
+ "acc": 0.5276595744680851,
+ "acc_stderr": 0.03263597118409769,
+ "acc_norm": 0.5276595744680851,
+ "acc_norm_stderr": 0.03263597118409769
+ },
+ "harness|mmlu_virology|5": {
+ "acc": 0.4759036144578313,
+ "acc_stderr": 0.03887971849597264,
+ "acc_norm": 0.4759036144578313,
+ "acc_norm_stderr": 0.03887971849597264
+ },
+ "harness|mmlu_philosophy|5": {
+ "acc": 0.6559485530546624,
+ "acc_stderr": 0.026981478043648043,
+ "acc_norm": 0.6559485530546624,
+ "acc_norm_stderr": 0.026981478043648043
+ },
+ "harness|mmlu_human_aging|5": {
+ "acc": 0.6412556053811659,
+ "acc_stderr": 0.032190792004199956,
+ "acc_norm": 0.6412556053811659,
+ "acc_norm_stderr": 0.032190792004199956
+ },
+ "harness|mmlu_human_sexuality|5": {
+ "acc": 0.648854961832061,
+ "acc_stderr": 0.04186445163013751,
+ "acc_norm": 0.648854961832061,
+ "acc_norm_stderr": 0.04186445163013751
+ },
+ "harness|mmlu_medical_genetics|5": {
+ "acc": 0.54,
+ "acc_stderr": 0.05009082659620333,
+ "acc_norm": 0.54,
+ "acc_norm_stderr": 0.05009082659620333
+ },
+ "harness|mmlu_high_school_geography|5": {
+ "acc": 0.7777777777777778,
+ "acc_stderr": 0.029620227874790465,
+ "acc_norm": 0.7777777777777778,
+ "acc_norm_stderr": 0.029620227874790465
+ },
+ "harness|mmlu_electrical_engineering|5": {
+ "acc": 0.5103448275862069,
+ "acc_stderr": 0.04165774775728762,
+ "acc_norm": 0.5103448275862069,
+ "acc_norm_stderr": 0.04165774775728762
+ },
+ "harness|mmlu_college_physics|5": {
+ "acc": 0.3627450980392157,
+ "acc_stderr": 0.04784060704105655,
+ "acc_norm": 0.3627450980392157,
+ "acc_norm_stderr": 0.04784060704105655
+ },
+ "harness|mmlu_high_school_microeconomics|5": {
+ "acc": 0.6680672268907563,
+ "acc_stderr": 0.03058869701378364,
+ "acc_norm": 0.6680672268907563,
+ "acc_norm_stderr": 0.03058869701378364
+ },
+ "harness|mmlu_high_school_macroeconomics|5": {
+ "acc": 0.6384615384615384,
+ "acc_stderr": 0.024359581465397,
+ "acc_norm": 0.6384615384615384,
+ "acc_norm_stderr": 0.024359581465397
+ },
+ "harness|mmlu_computer_security|5": {
+ "acc": 0.65,
+ "acc_stderr": 0.04793724854411021,
+ "acc_norm": 0.65,
+ "acc_norm_stderr": 0.04793724854411021
+ },
+ "harness|mmlu_global_facts|5": {
+ "acc": 0.37,
+ "acc_stderr": 0.04852365870939099,
+ "acc_norm": 0.37,
+ "acc_norm_stderr": 0.04852365870939099
+ },
+ "harness|mmlu_jurisprudence|5": {
+ "acc": 0.6851851851851852,
+ "acc_stderr": 0.04489931073591312,
+ "acc_norm": 0.6851851851851852,
+ "acc_norm_stderr": 0.04489931073591312
+ },
+ "harness|mmlu_high_school_chemistry|5": {
+ "acc": 0.46798029556650245,
+ "acc_stderr": 0.035107665979592154,
+ "acc_norm": 0.46798029556650245,
+ "acc_norm_stderr": 0.035107665979592154
+ },
+ "harness|mmlu_high_school_biology|5": {
+ "acc": 0.6548387096774193,
+ "acc_stderr": 0.02704574657353432,
+ "acc_norm": 0.6548387096774193,
+ "acc_norm_stderr": 0.02704574657353432
+ },
+ "harness|mmlu_marketing|5": {
+ "acc": 0.8162393162393162,
+ "acc_stderr": 0.025372139671722933,
+ "acc_norm": 0.8162393162393162,
+ "acc_norm_stderr": 0.025372139671722933
+ },
+ "harness|mmlu_clinical_knowledge|5": {
+ "acc": 0.5773584905660377,
+ "acc_stderr": 0.03040233144576954,
+ "acc_norm": 0.5773584905660377,
+ "acc_norm_stderr": 0.03040233144576954
+ },
+ "harness|mmlu_public_relations|5": {
+ "acc": 0.6454545454545455,
+ "acc_stderr": 0.045820048415054174,
+ "acc_norm": 0.6454545454545455,
+ "acc_norm_stderr": 0.045820048415054174
+ },
+ "harness|mmlu_high_school_mathematics|5": {
+ "acc": 0.4074074074074074,
+ "acc_stderr": 0.029958249250082118,
+ "acc_norm": 0.4074074074074074,
+ "acc_norm_stderr": 0.029958249250082118
+ },
+ "harness|mmlu_high_school_physics|5": {
+ "acc": 0.3509933774834437,
+ "acc_stderr": 0.03896981964257375,
+ "acc_norm": 0.3509933774834437,
+ "acc_norm_stderr": 0.03896981964257375
+ },
+ "harness|mmlu_sociology|5": {
+ "acc": 0.7263681592039801,
+ "acc_stderr": 0.03152439186555404,
+ "acc_norm": 0.7263681592039801,
+ "acc_norm_stderr": 0.03152439186555404
+ },
+ "harness|mmlu_college_medicine|5": {
+ "acc": 0.5375722543352601,
+ "acc_stderr": 0.0380168510452446,
+ "acc_norm": 0.5375722543352601,
+ "acc_norm_stderr": 0.0380168510452446
+ },
+ "harness|mmlu_elementary_mathematics|5": {
+ "acc": 0.4365079365079365,
+ "acc_stderr": 0.025542846817400496,
+ "acc_norm": 0.4365079365079365,
+ "acc_norm_stderr": 0.025542846817400496
+ },
+ "harness|mmlu_college_biology|5": {
+ "acc": 0.5694444444444444,
+ "acc_stderr": 0.04140685639111503,
+ "acc_norm": 0.5694444444444444,
+ "acc_norm_stderr": 0.04140685639111503
+ },
+ "harness|mmlu_college_chemistry|5": {
+ "acc": 0.43,
+ "acc_stderr": 0.049756985195624284,
+ "acc_norm": 0.43,
+ "acc_norm_stderr": 0.049756985195624284
+ },
+ "harness|mmlu_us_foreign_policy|5": {
+ "acc": 0.78,
+ "acc_stderr": 0.04163331998932263,
+ "acc_norm": 0.78,
+ "acc_norm_stderr": 0.04163331998932263
+ },
+ "harness|mmlu_moral_disputes|5": {
+ "acc": 0.6098265895953757,
+ "acc_stderr": 0.026261677607806642,
+ "acc_norm": 0.6098265895953757,
+ "acc_norm_stderr": 0.026261677607806642
+ },
+ "harness|mmlu_logical_fallacies|5": {
+ "acc": 0.656441717791411,
+ "acc_stderr": 0.03731133519673893,
+ "acc_norm": 0.656441717791411,
+ "acc_norm_stderr": 0.03731133519673893
+ },
+ "harness|mmlu_prehistory|5": {
+ "acc": 0.6574074074074074,
+ "acc_stderr": 0.02640614597362568,
+ "acc_norm": 0.6574074074074074,
+ "acc_norm_stderr": 0.02640614597362568
+ },
+ "harness|mmlu_college_mathematics|5": {
+ "acc": 0.38,
+ "acc_stderr": 0.04878317312145632,
+ "acc_norm": 0.38,
+ "acc_norm_stderr": 0.04878317312145632
+ },
+ "harness|mmlu_high_school_government_and_politics|5": {
+ "acc": 0.7668393782383419,
+ "acc_stderr": 0.03051611137147601,
+ "acc_norm": 0.7668393782383419,
+ "acc_norm_stderr": 0.03051611137147601
+ },
+ "harness|mmlu_econometrics|5": {
+ "acc": 0.45614035087719296,
+ "acc_stderr": 0.046854730419077895,
+ "acc_norm": 0.45614035087719296,
+ "acc_norm_stderr": 0.046854730419077895
+ },
+ "harness|mmlu_high_school_psychology|5": {
+ "acc": 0.7853211009174312,
+ "acc_stderr": 0.017604304149256494,
+ "acc_norm": 0.7853211009174312,
+ "acc_norm_stderr": 0.017604304149256494
+ },
+ "harness|mmlu_formal_logic|5": {
+ "acc": 0.4523809523809524,
+ "acc_stderr": 0.044518079590553275,
+ "acc_norm": 0.4523809523809524,
+ "acc_norm_stderr": 0.044518079590553275
+ },
+ "harness|mmlu_nutrition|5": {
+ "acc": 0.6405228758169934,
+ "acc_stderr": 0.027475969910660952,
+ "acc_norm": 0.6405228758169934,
+ "acc_norm_stderr": 0.027475969910660952
+ },
+ "harness|mmlu_business_ethics|5": {
+ "acc": 0.66,
+ "acc_stderr": 0.04760952285695237,
+ "acc_norm": 0.66,
+ "acc_norm_stderr": 0.04760952285695237
+ },
+ "harness|mmlu_international_law|5": {
+ "acc": 0.7933884297520661,
+ "acc_stderr": 0.03695980128098824,
+ "acc_norm": 0.7933884297520661,
+ "acc_norm_stderr": 0.03695980128098824
+ },
+ "harness|mmlu_astronomy|5": {
+ "acc": 0.6842105263157895,
+ "acc_stderr": 0.0378272898086547,
+ "acc_norm": 0.6842105263157895,
+ "acc_norm_stderr": 0.0378272898086547
+ },
+ "harness|mmlu_professional_psychology|5": {
+ "acc": 0.5964052287581699,
+ "acc_stderr": 0.019848280168401164,
+ "acc_norm": 0.5964052287581699,
+ "acc_norm_stderr": 0.019848280168401164
+ },
+ "harness|mmlu_professional_accounting|5": {
+ "acc": 0.4397163120567376,
+ "acc_stderr": 0.02960991207559411,
+ "acc_norm": 0.4397163120567376,
+ "acc_norm_stderr": 0.02960991207559411
+ },
+ "harness|mmlu_machine_learning|5": {
+ "acc": 0.39285714285714285,
+ "acc_stderr": 0.04635550135609976,
+ "acc_norm": 0.39285714285714285,
+ "acc_norm_stderr": 0.04635550135609976
+ },
+ "harness|mmlu_high_school_statistics|5": {
+ "acc": 0.5787037037037037,
+ "acc_stderr": 0.03367462138896078,
+ "acc_norm": 0.5787037037037037,
+ "acc_norm_stderr": 0.03367462138896078
+ },
+ "harness|mmlu_moral_scenarios|5": {
+ "acc": 0.264804469273743,
+ "acc_stderr": 0.01475690648326066,
+ "acc_norm": 0.264804469273743,
+ "acc_norm_stderr": 0.01475690648326066
+ },
+ "harness|mmlu_college_computer_science|5": {
+ "acc": 0.52,
+ "acc_stderr": 0.050211673156867795,
+ "acc_norm": 0.52,
+ "acc_norm_stderr": 0.050211673156867795
+ },
+ "harness|mmlu_high_school_computer_science|5": {
+ "acc": 0.7,
+ "acc_stderr": 0.04605661864718381,
+ "acc_norm": 0.7,
+ "acc_norm_stderr": 0.04605661864718381
+ },
+ "harness|mmlu_professional_medicine|5": {
+ "acc": 0.5588235294117647,
+ "acc_stderr": 0.03016191193076711,
+ "acc_norm": 0.5588235294117647,
+ "acc_norm_stderr": 0.03016191193076711
+ },
+ "harness|mmlu_security_studies|5": {
+ "acc": 0.6448979591836734,
+ "acc_stderr": 0.030635655150387634,
+ "acc_norm": 0.6448979591836734,
+ "acc_norm_stderr": 0.030635655150387634
+ },
+ "harness|mmlu_high_school_world_history|5": {
+ "acc": 0.7426160337552743,
+ "acc_stderr": 0.028458820991460302,
+ "acc_norm": 0.7426160337552743,
+ "acc_norm_stderr": 0.028458820991460302
+ },
+ "harness|mmlu_professional_law|5": {
+ "acc": 0.44654498044328556,
+ "acc_stderr": 0.012697046024399661,
+ "acc_norm": 0.44654498044328556,
+ "acc_norm_stderr": 0.012697046024399661
+ },
+ "harness|mmlu_high_school_us_history|5": {
+ "acc": 0.6225490196078431,
+ "acc_stderr": 0.03402272044340703,
+ "acc_norm": 0.6225490196078431,
+ "acc_norm_stderr": 0.03402272044340703
+ },
+ "harness|mmlu_high_school_european_history|5": {
+ "acc": 0.6303030303030303,
+ "acc_stderr": 0.03769430314512569,
+ "acc_norm": 0.6303030303030303,
+ "acc_norm_stderr": 0.03769430314512569
+ },
+ "harness|truthfulqa_mc|0": {
+ "mc1": 0.6634026927784578,
+ "mc1_stderr": 0.0165424128094949,
+ "mc2": 0.7515104740134964,
+ "mc2_stderr": 0.014200593490054807
+ },
+ "harness|commongen_v2|2": {
+ "acc": 0.5147579693034239,
+ "acc_stderr": 0.01718286443499856,
+ "acc_norm": 0.526564344746163,
+ "acc_norm_stderr": 0.017166075717577747
+ }
+ },
+ "versions": {
+ "all": 0,
+ "harness|arc_challenge|25": 0,
+ "harness|hellaswag|10": 0,
+ "harness|mmlu_world_religions|5": 1,
+ "harness|mmlu_management|5": 1,
+ "harness|mmlu_miscellaneous|5": 1,
+ "harness|mmlu_anatomy|5": 1,
+ "harness|mmlu_abstract_algebra|5": 1,
+ "harness|mmlu_conceptual_physics|5": 1,
+ "harness|mmlu_virology|5": 1,
+ "harness|mmlu_philosophy|5": 1,
+ "harness|mmlu_human_aging|5": 1,
+ "harness|mmlu_human_sexuality|5": 1,
+ "harness|mmlu_medical_genetics|5": 1,
+ "harness|mmlu_high_school_geography|5": 1,
+ "harness|mmlu_electrical_engineering|5": 1,
+ "harness|mmlu_college_physics|5": 1,
+ "harness|mmlu_high_school_microeconomics|5": 1,
+ "harness|mmlu_high_school_macroeconomics|5": 1,
+ "harness|mmlu_computer_security|5": 1,
+ "harness|mmlu_global_facts|5": 1,
+ "harness|mmlu_jurisprudence|5": 1,
+ "harness|mmlu_high_school_chemistry|5": 1,
+ "harness|mmlu_high_school_biology|5": 1,
+ "harness|mmlu_marketing|5": 1,
+ "harness|mmlu_clinical_knowledge|5": 1,
+ "harness|mmlu_public_relations|5": 1,
+ "harness|mmlu_high_school_mathematics|5": 1,
+ "harness|mmlu_high_school_physics|5": 1,
+ "harness|mmlu_sociology|5": 1,
+ "harness|mmlu_college_medicine|5": 1,
+ "harness|mmlu_elementary_mathematics|5": 1,
+ "harness|mmlu_college_biology|5": 1,
+ "harness|mmlu_college_chemistry|5": 1,
+ "harness|mmlu_us_foreign_policy|5": 1,
+ "harness|mmlu_moral_disputes|5": 1,
+ "harness|mmlu_logical_fallacies|5": 1,
+ "harness|mmlu_prehistory|5": 1,
+ "harness|mmlu_college_mathematics|5": 1,
+ "harness|mmlu_high_school_government_and_politics|5": 1,
+ "harness|mmlu_econometrics|5": 1,
+ "harness|mmlu_high_school_psychology|5": 1,
+ "harness|mmlu_formal_logic|5": 1,
+ "harness|mmlu_nutrition|5": 1,
+ "harness|mmlu_business_ethics|5": 1,
+ "harness|mmlu_international_law|5": 1,
+ "harness|mmlu_astronomy|5": 1,
+ "harness|mmlu_professional_psychology|5": 1,
+ "harness|mmlu_professional_accounting|5": 1,
+ "harness|mmlu_machine_learning|5": 1,
+ "harness|mmlu_high_school_statistics|5": 1,
+ "harness|mmlu_moral_scenarios|5": 1,
+ "harness|mmlu_college_computer_science|5": 1,
+ "harness|mmlu_high_school_computer_science|5": 1,
+ "harness|mmlu_professional_medicine|5": 1,
+ "harness|mmlu_security_studies|5": 1,
+ "harness|mmlu_high_school_world_history|5": 1,
+ "harness|mmlu_professional_law|5": 1,
+ "harness|mmlu_high_school_us_history|5": 1,
+ "harness|mmlu_high_school_european_history|5": 1,
+ "harness|truthfulqa_mc|0": 0,
+ "harness|commongen_v2|2": 1
+ },
+ "config_general": {
+ "model_name": "x2bee/POLAR-14B-HES-DPO-v1.5",
+ "model_sha": "f0bc8e2566ba28c8232d7c690098e634ea894e8d",
+ "model_dtype": "torch.float16",
+ "lighteval_sha": "",
+ "num_few_shot_default": 0,
+ "num_fewshot_seeds": 1,
+ "override_batch_size": 1,
+ "max_samples": null
+ }
+}
\ No newline at end of file
diff --git a/eval-results/x2bee/POLAR-14B-SON-SFT-v0.1/result.json b/eval-results/x2bee/POLAR-14B-SON-SFT-v0.1/result.json
new file mode 100644
index 0000000000000000000000000000000000000000..79c8abce9a4e3db7fb8fa93c24ae76400cea8f9c
--- /dev/null
+++ b/eval-results/x2bee/POLAR-14B-SON-SFT-v0.1/result.json
@@ -0,0 +1,450 @@
+{
+ "results": {
+ "daily": {
+ "daily": 3
+ },
+ "quarterly": {
+ "quarterly": 3
+ },
+ "harness|arc_challenge|25": {
+ "acc": 0.6646757679180887,
+ "acc_stderr": 0.013796182947785564,
+ "acc_norm": 0.7244027303754266,
+ "acc_norm_stderr": 0.01305716965576184
+ },
+ "harness|hellaswag|10": {
+ "acc": 0.46036646086436966,
+ "acc_stderr": 0.004974080638364276,
+ "acc_norm": 0.6195976897032464,
+ "acc_norm_stderr": 0.004844935327599196
+ },
+ "harness|mmlu_world_religions|5": {
+ "acc": 0.7602339181286549,
+ "acc_stderr": 0.03274485211946956,
+ "acc_norm": 0.7602339181286549,
+ "acc_norm_stderr": 0.03274485211946956
+ },
+ "harness|mmlu_management|5": {
+ "acc": 0.7766990291262136,
+ "acc_stderr": 0.04123553189891431,
+ "acc_norm": 0.7766990291262136,
+ "acc_norm_stderr": 0.04123553189891431
+ },
+ "harness|mmlu_miscellaneous|5": {
+ "acc": 0.7381864623243933,
+ "acc_stderr": 0.01572083867844526,
+ "acc_norm": 0.7381864623243933,
+ "acc_norm_stderr": 0.01572083867844526
+ },
+ "harness|mmlu_anatomy|5": {
+ "acc": 0.5037037037037037,
+ "acc_stderr": 0.04319223625811331,
+ "acc_norm": 0.5037037037037037,
+ "acc_norm_stderr": 0.04319223625811331
+ },
+ "harness|mmlu_abstract_algebra|5": {
+ "acc": 0.35,
+ "acc_stderr": 0.04793724854411019,
+ "acc_norm": 0.35,
+ "acc_norm_stderr": 0.04793724854411019
+ },
+ "harness|mmlu_conceptual_physics|5": {
+ "acc": 0.5404255319148936,
+ "acc_stderr": 0.032579014820998335,
+ "acc_norm": 0.5404255319148936,
+ "acc_norm_stderr": 0.032579014820998335
+ },
+ "harness|mmlu_virology|5": {
+ "acc": 0.5180722891566265,
+ "acc_stderr": 0.038899512528272166,
+ "acc_norm": 0.5180722891566265,
+ "acc_norm_stderr": 0.038899512528272166
+ },
+ "harness|mmlu_philosophy|5": {
+ "acc": 0.6559485530546624,
+ "acc_stderr": 0.026981478043648043,
+ "acc_norm": 0.6559485530546624,
+ "acc_norm_stderr": 0.026981478043648043
+ },
+ "harness|mmlu_human_aging|5": {
+ "acc": 0.6591928251121076,
+ "acc_stderr": 0.0318114974705536,
+ "acc_norm": 0.6591928251121076,
+ "acc_norm_stderr": 0.0318114974705536
+ },
+ "harness|mmlu_human_sexuality|5": {
+ "acc": 0.6564885496183206,
+ "acc_stderr": 0.041649760719448786,
+ "acc_norm": 0.6564885496183206,
+ "acc_norm_stderr": 0.041649760719448786
+ },
+ "harness|mmlu_medical_genetics|5": {
+ "acc": 0.52,
+ "acc_stderr": 0.050211673156867795,
+ "acc_norm": 0.52,
+ "acc_norm_stderr": 0.050211673156867795
+ },
+ "harness|mmlu_high_school_geography|5": {
+ "acc": 0.7575757575757576,
+ "acc_stderr": 0.030532892233932036,
+ "acc_norm": 0.7575757575757576,
+ "acc_norm_stderr": 0.030532892233932036
+ },
+ "harness|mmlu_electrical_engineering|5": {
+ "acc": 0.5586206896551724,
+ "acc_stderr": 0.04137931034482757,
+ "acc_norm": 0.5586206896551724,
+ "acc_norm_stderr": 0.04137931034482757
+ },
+ "harness|mmlu_college_physics|5": {
+ "acc": 0.3137254901960784,
+ "acc_stderr": 0.04617034827006717,
+ "acc_norm": 0.3137254901960784,
+ "acc_norm_stderr": 0.04617034827006717
+ },
+ "harness|mmlu_high_school_microeconomics|5": {
+ "acc": 0.6512605042016807,
+ "acc_stderr": 0.03095663632856655,
+ "acc_norm": 0.6512605042016807,
+ "acc_norm_stderr": 0.03095663632856655
+ },
+ "harness|mmlu_high_school_macroeconomics|5": {
+ "acc": 0.6230769230769231,
+ "acc_stderr": 0.024570975364225995,
+ "acc_norm": 0.6230769230769231,
+ "acc_norm_stderr": 0.024570975364225995
+ },
+ "harness|mmlu_computer_security|5": {
+ "acc": 0.73,
+ "acc_stderr": 0.04461960433384739,
+ "acc_norm": 0.73,
+ "acc_norm_stderr": 0.04461960433384739
+ },
+ "harness|mmlu_global_facts|5": {
+ "acc": 0.32,
+ "acc_stderr": 0.04688261722621505,
+ "acc_norm": 0.32,
+ "acc_norm_stderr": 0.04688261722621505
+ },
+ "harness|mmlu_jurisprudence|5": {
+ "acc": 0.7037037037037037,
+ "acc_stderr": 0.04414343666854933,
+ "acc_norm": 0.7037037037037037,
+ "acc_norm_stderr": 0.04414343666854933
+ },
+ "harness|mmlu_high_school_chemistry|5": {
+ "acc": 0.4630541871921182,
+ "acc_stderr": 0.035083705204426656,
+ "acc_norm": 0.4630541871921182,
+ "acc_norm_stderr": 0.035083705204426656
+ },
+ "harness|mmlu_high_school_biology|5": {
+ "acc": 0.603225806451613,
+ "acc_stderr": 0.027831231605767944,
+ "acc_norm": 0.603225806451613,
+ "acc_norm_stderr": 0.027831231605767944
+ },
+ "harness|mmlu_marketing|5": {
+ "acc": 0.8205128205128205,
+ "acc_stderr": 0.025140935950335435,
+ "acc_norm": 0.8205128205128205,
+ "acc_norm_stderr": 0.025140935950335435
+ },
+ "harness|mmlu_clinical_knowledge|5": {
+ "acc": 0.5962264150943396,
+ "acc_stderr": 0.03019761160019795,
+ "acc_norm": 0.5962264150943396,
+ "acc_norm_stderr": 0.03019761160019795
+ },
+ "harness|mmlu_public_relations|5": {
+ "acc": 0.6181818181818182,
+ "acc_stderr": 0.046534298079135075,
+ "acc_norm": 0.6181818181818182,
+ "acc_norm_stderr": 0.046534298079135075
+ },
+ "harness|mmlu_high_school_mathematics|5": {
+ "acc": 0.37407407407407406,
+ "acc_stderr": 0.029502861128955293,
+ "acc_norm": 0.37407407407407406,
+ "acc_norm_stderr": 0.029502861128955293
+ },
+ "harness|mmlu_high_school_physics|5": {
+ "acc": 0.32450331125827814,
+ "acc_stderr": 0.038227469376587525,
+ "acc_norm": 0.32450331125827814,
+ "acc_norm_stderr": 0.038227469376587525
+ },
+ "harness|mmlu_sociology|5": {
+ "acc": 0.7164179104477612,
+ "acc_stderr": 0.03187187537919796,
+ "acc_norm": 0.7164179104477612,
+ "acc_norm_stderr": 0.03187187537919796
+ },
+ "harness|mmlu_college_medicine|5": {
+ "acc": 0.5375722543352601,
+ "acc_stderr": 0.03801685104524458,
+ "acc_norm": 0.5375722543352601,
+ "acc_norm_stderr": 0.03801685104524458
+ },
+ "harness|mmlu_elementary_mathematics|5": {
+ "acc": 0.42857142857142855,
+ "acc_stderr": 0.025487187147859372,
+ "acc_norm": 0.42857142857142855,
+ "acc_norm_stderr": 0.025487187147859372
+ },
+ "harness|mmlu_college_biology|5": {
+ "acc": 0.5902777777777778,
+ "acc_stderr": 0.04112490974670787,
+ "acc_norm": 0.5902777777777778,
+ "acc_norm_stderr": 0.04112490974670787
+ },
+ "harness|mmlu_college_chemistry|5": {
+ "acc": 0.45,
+ "acc_stderr": 0.049999999999999996,
+ "acc_norm": 0.45,
+ "acc_norm_stderr": 0.049999999999999996
+ },
+ "harness|mmlu_us_foreign_policy|5": {
+ "acc": 0.78,
+ "acc_stderr": 0.04163331998932263,
+ "acc_norm": 0.78,
+ "acc_norm_stderr": 0.04163331998932263
+ },
+ "harness|mmlu_moral_disputes|5": {
+ "acc": 0.6184971098265896,
+ "acc_stderr": 0.026152198619726803,
+ "acc_norm": 0.6184971098265896,
+ "acc_norm_stderr": 0.026152198619726803
+ },
+ "harness|mmlu_logical_fallacies|5": {
+ "acc": 0.6441717791411042,
+ "acc_stderr": 0.03761521380046734,
+ "acc_norm": 0.6441717791411042,
+ "acc_norm_stderr": 0.03761521380046734
+ },
+ "harness|mmlu_prehistory|5": {
+ "acc": 0.6944444444444444,
+ "acc_stderr": 0.025630824975621365,
+ "acc_norm": 0.6944444444444444,
+ "acc_norm_stderr": 0.025630824975621365
+ },
+ "harness|mmlu_college_mathematics|5": {
+ "acc": 0.43,
+ "acc_stderr": 0.049756985195624284,
+ "acc_norm": 0.43,
+ "acc_norm_stderr": 0.049756985195624284
+ },
+ "harness|mmlu_high_school_government_and_politics|5": {
+ "acc": 0.7927461139896373,
+ "acc_stderr": 0.029252823291803638,
+ "acc_norm": 0.7927461139896373,
+ "acc_norm_stderr": 0.029252823291803638
+ },
+ "harness|mmlu_econometrics|5": {
+ "acc": 0.43859649122807015,
+ "acc_stderr": 0.04668000738510455,
+ "acc_norm": 0.43859649122807015,
+ "acc_norm_stderr": 0.04668000738510455
+ },
+ "harness|mmlu_high_school_psychology|5": {
+ "acc": 0.7853211009174312,
+ "acc_stderr": 0.017604304149256494,
+ "acc_norm": 0.7853211009174312,
+ "acc_norm_stderr": 0.017604304149256494
+ },
+ "harness|mmlu_formal_logic|5": {
+ "acc": 0.3968253968253968,
+ "acc_stderr": 0.04375888492727062,
+ "acc_norm": 0.3968253968253968,
+ "acc_norm_stderr": 0.04375888492727062
+ },
+ "harness|mmlu_nutrition|5": {
+ "acc": 0.6437908496732027,
+ "acc_stderr": 0.027420477662629245,
+ "acc_norm": 0.6437908496732027,
+ "acc_norm_stderr": 0.027420477662629245
+ },
+ "harness|mmlu_business_ethics|5": {
+ "acc": 0.63,
+ "acc_stderr": 0.04852365870939099,
+ "acc_norm": 0.63,
+ "acc_norm_stderr": 0.04852365870939099
+ },
+ "harness|mmlu_international_law|5": {
+ "acc": 0.7603305785123967,
+ "acc_stderr": 0.03896878985070415,
+ "acc_norm": 0.7603305785123967,
+ "acc_norm_stderr": 0.03896878985070415
+ },
+ "harness|mmlu_astronomy|5": {
+ "acc": 0.625,
+ "acc_stderr": 0.039397364351956274,
+ "acc_norm": 0.625,
+ "acc_norm_stderr": 0.039397364351956274
+ },
+ "harness|mmlu_professional_psychology|5": {
+ "acc": 0.619281045751634,
+ "acc_stderr": 0.019643801557924806,
+ "acc_norm": 0.619281045751634,
+ "acc_norm_stderr": 0.019643801557924806
+ },
+ "harness|mmlu_professional_accounting|5": {
+ "acc": 0.45390070921985815,
+ "acc_stderr": 0.029700453247291467,
+ "acc_norm": 0.45390070921985815,
+ "acc_norm_stderr": 0.029700453247291467
+ },
+ "harness|mmlu_machine_learning|5": {
+ "acc": 0.41964285714285715,
+ "acc_stderr": 0.04684099321077106,
+ "acc_norm": 0.41964285714285715,
+ "acc_norm_stderr": 0.04684099321077106
+ },
+ "harness|mmlu_high_school_statistics|5": {
+ "acc": 0.5555555555555556,
+ "acc_stderr": 0.03388857118502326,
+ "acc_norm": 0.5555555555555556,
+ "acc_norm_stderr": 0.03388857118502326
+ },
+ "harness|mmlu_moral_scenarios|5": {
+ "acc": 0.3575418994413408,
+ "acc_stderr": 0.016029394474894893,
+ "acc_norm": 0.3575418994413408,
+ "acc_norm_stderr": 0.016029394474894893
+ },
+ "harness|mmlu_college_computer_science|5": {
+ "acc": 0.52,
+ "acc_stderr": 0.050211673156867795,
+ "acc_norm": 0.52,
+ "acc_norm_stderr": 0.050211673156867795
+ },
+ "harness|mmlu_high_school_computer_science|5": {
+ "acc": 0.75,
+ "acc_stderr": 0.04351941398892446,
+ "acc_norm": 0.75,
+ "acc_norm_stderr": 0.04351941398892446
+ },
+ "harness|mmlu_professional_medicine|5": {
+ "acc": 0.5735294117647058,
+ "acc_stderr": 0.03004261583271486,
+ "acc_norm": 0.5735294117647058,
+ "acc_norm_stderr": 0.03004261583271486
+ },
+ "harness|mmlu_security_studies|5": {
+ "acc": 0.6816326530612244,
+ "acc_stderr": 0.02982253379398204,
+ "acc_norm": 0.6816326530612244,
+ "acc_norm_stderr": 0.02982253379398204
+ },
+ "harness|mmlu_high_school_world_history|5": {
+ "acc": 0.7468354430379747,
+ "acc_stderr": 0.028304657943035293,
+ "acc_norm": 0.7468354430379747,
+ "acc_norm_stderr": 0.028304657943035293
+ },
+ "harness|mmlu_professional_law|5": {
+ "acc": 0.455019556714472,
+ "acc_stderr": 0.012718456618701789,
+ "acc_norm": 0.455019556714472,
+ "acc_norm_stderr": 0.012718456618701789
+ },
+ "harness|mmlu_high_school_us_history|5": {
+ "acc": 0.6666666666666666,
+ "acc_stderr": 0.033086111132364364,
+ "acc_norm": 0.6666666666666666,
+ "acc_norm_stderr": 0.033086111132364364
+ },
+ "harness|mmlu_high_school_european_history|5": {
+ "acc": 0.6484848484848484,
+ "acc_stderr": 0.037282069986826503,
+ "acc_norm": 0.6484848484848484,
+ "acc_norm_stderr": 0.037282069986826503
+ },
+ "harness|truthfulqa_mc|0": {
+ "mc1": 0.605875152998776,
+ "mc1_stderr": 0.017106588140700332,
+ "mc2": 0.7254831072808595,
+ "mc2_stderr": 0.014162522228042162
+ },
+ "harness|commongen_v2|2": {
+ "acc": 0.5926800472255017,
+ "acc_stderr": 0.01689245669519127,
+ "acc_norm": 0.6269185360094451,
+ "acc_norm_stderr": 0.016627318275137453
+ }
+ },
+ "versions": {
+ "all": 0,
+ "harness|arc_challenge|25": 0,
+ "harness|hellaswag|10": 0,
+ "harness|mmlu_world_religions|5": 1,
+ "harness|mmlu_management|5": 1,
+ "harness|mmlu_miscellaneous|5": 1,
+ "harness|mmlu_anatomy|5": 1,
+ "harness|mmlu_abstract_algebra|5": 1,
+ "harness|mmlu_conceptual_physics|5": 1,
+ "harness|mmlu_virology|5": 1,
+ "harness|mmlu_philosophy|5": 1,
+ "harness|mmlu_human_aging|5": 1,
+ "harness|mmlu_human_sexuality|5": 1,
+ "harness|mmlu_medical_genetics|5": 1,
+ "harness|mmlu_high_school_geography|5": 1,
+ "harness|mmlu_electrical_engineering|5": 1,
+ "harness|mmlu_college_physics|5": 1,
+ "harness|mmlu_high_school_microeconomics|5": 1,
+ "harness|mmlu_high_school_macroeconomics|5": 1,
+ "harness|mmlu_computer_security|5": 1,
+ "harness|mmlu_global_facts|5": 1,
+ "harness|mmlu_jurisprudence|5": 1,
+ "harness|mmlu_high_school_chemistry|5": 1,
+ "harness|mmlu_high_school_biology|5": 1,
+ "harness|mmlu_marketing|5": 1,
+ "harness|mmlu_clinical_knowledge|5": 1,
+ "harness|mmlu_public_relations|5": 1,
+ "harness|mmlu_high_school_mathematics|5": 1,
+ "harness|mmlu_high_school_physics|5": 1,
+ "harness|mmlu_sociology|5": 1,
+ "harness|mmlu_college_medicine|5": 1,
+ "harness|mmlu_elementary_mathematics|5": 1,
+ "harness|mmlu_college_biology|5": 1,
+ "harness|mmlu_college_chemistry|5": 1,
+ "harness|mmlu_us_foreign_policy|5": 1,
+ "harness|mmlu_moral_disputes|5": 1,
+ "harness|mmlu_logical_fallacies|5": 1,
+ "harness|mmlu_prehistory|5": 1,
+ "harness|mmlu_college_mathematics|5": 1,
+ "harness|mmlu_high_school_government_and_politics|5": 1,
+ "harness|mmlu_econometrics|5": 1,
+ "harness|mmlu_high_school_psychology|5": 1,
+ "harness|mmlu_formal_logic|5": 1,
+ "harness|mmlu_nutrition|5": 1,
+ "harness|mmlu_business_ethics|5": 1,
+ "harness|mmlu_international_law|5": 1,
+ "harness|mmlu_astronomy|5": 1,
+ "harness|mmlu_professional_psychology|5": 1,
+ "harness|mmlu_professional_accounting|5": 1,
+ "harness|mmlu_machine_learning|5": 1,
+ "harness|mmlu_high_school_statistics|5": 1,
+ "harness|mmlu_moral_scenarios|5": 1,
+ "harness|mmlu_college_computer_science|5": 1,
+ "harness|mmlu_high_school_computer_science|5": 1,
+ "harness|mmlu_professional_medicine|5": 1,
+ "harness|mmlu_security_studies|5": 1,
+ "harness|mmlu_high_school_world_history|5": 1,
+ "harness|mmlu_professional_law|5": 1,
+ "harness|mmlu_high_school_us_history|5": 1,
+ "harness|mmlu_high_school_european_history|5": 1,
+ "harness|truthfulqa_mc|0": 0,
+ "harness|commongen_v2|2": 1
+ },
+ "config_general": {
+ "model_name": "x2bee/POLAR-14B-SON-SFT-v0.1",
+ "model_sha": "01286a13088332c1eda4279b5bcfa7a0a33e145f",
+ "model_dtype": "torch.float16",
+ "lighteval_sha": "",
+ "num_few_shot_default": 0,
+ "num_fewshot_seeds": 1,
+ "override_batch_size": 1,
+ "max_samples": null
+ }
+}
\ No newline at end of file
diff --git a/eval-results/x2bee/POLAR-14B-v0.2/result.json b/eval-results/x2bee/POLAR-14B-v0.2/result.json
new file mode 100644
index 0000000000000000000000000000000000000000..63fae0f070c9ffe9477f498fe19733803d29c81e
--- /dev/null
+++ b/eval-results/x2bee/POLAR-14B-v0.2/result.json
@@ -0,0 +1,450 @@
+{
+ "results": {
+ "daily": {
+ "daily": 2
+ },
+ "quarterly": {
+ "quarterly": 2
+ },
+ "harness|arc_challenge|25": {
+ "acc": 0.7465870307167235,
+ "acc_stderr": 0.012710896778378602,
+ "acc_norm": 0.7687713310580204,
+ "acc_norm_stderr": 0.012320858834772264
+ },
+ "harness|hellaswag|10": {
+ "acc": 0.681736705835491,
+ "acc_stderr": 0.004648503177353952,
+ "acc_norm": 0.7999402509460267,
+ "acc_norm_stderr": 0.003992272261659531
+ },
+ "harness|mmlu_world_religions|5": {
+ "acc": 0.6549707602339181,
+ "acc_stderr": 0.036459813773888065,
+ "acc_norm": 0.6549707602339181,
+ "acc_norm_stderr": 0.036459813773888065
+ },
+ "harness|mmlu_management|5": {
+ "acc": 0.7378640776699029,
+ "acc_stderr": 0.043546310772605956,
+ "acc_norm": 0.7378640776699029,
+ "acc_norm_stderr": 0.043546310772605956
+ },
+ "harness|mmlu_miscellaneous|5": {
+ "acc": 0.6922094508301405,
+ "acc_stderr": 0.016506045045155633,
+ "acc_norm": 0.6922094508301405,
+ "acc_norm_stderr": 0.016506045045155633
+ },
+ "harness|mmlu_anatomy|5": {
+ "acc": 0.4666666666666667,
+ "acc_stderr": 0.043097329010363554,
+ "acc_norm": 0.4666666666666667,
+ "acc_norm_stderr": 0.043097329010363554
+ },
+ "harness|mmlu_abstract_algebra|5": {
+ "acc": 0.35,
+ "acc_stderr": 0.047937248544110196,
+ "acc_norm": 0.35,
+ "acc_norm_stderr": 0.047937248544110196
+ },
+ "harness|mmlu_conceptual_physics|5": {
+ "acc": 0.4595744680851064,
+ "acc_stderr": 0.03257901482099836,
+ "acc_norm": 0.4595744680851064,
+ "acc_norm_stderr": 0.03257901482099836
+ },
+ "harness|mmlu_virology|5": {
+ "acc": 0.4879518072289157,
+ "acc_stderr": 0.03891364495835821,
+ "acc_norm": 0.4879518072289157,
+ "acc_norm_stderr": 0.03891364495835821
+ },
+ "harness|mmlu_philosophy|5": {
+ "acc": 0.6045016077170418,
+ "acc_stderr": 0.027770918531427834,
+ "acc_norm": 0.6045016077170418,
+ "acc_norm_stderr": 0.027770918531427834
+ },
+ "harness|mmlu_human_aging|5": {
+ "acc": 0.6233183856502242,
+ "acc_stderr": 0.03252113489929188,
+ "acc_norm": 0.6233183856502242,
+ "acc_norm_stderr": 0.03252113489929188
+ },
+ "harness|mmlu_human_sexuality|5": {
+ "acc": 0.6412213740458015,
+ "acc_stderr": 0.04206739313864908,
+ "acc_norm": 0.6412213740458015,
+ "acc_norm_stderr": 0.04206739313864908
+ },
+ "harness|mmlu_medical_genetics|5": {
+ "acc": 0.51,
+ "acc_stderr": 0.05024183937956911,
+ "acc_norm": 0.51,
+ "acc_norm_stderr": 0.05024183937956911
+ },
+ "harness|mmlu_high_school_geography|5": {
+ "acc": 0.7222222222222222,
+ "acc_stderr": 0.03191178226713547,
+ "acc_norm": 0.7222222222222222,
+ "acc_norm_stderr": 0.03191178226713547
+ },
+ "harness|mmlu_electrical_engineering|5": {
+ "acc": 0.5241379310344828,
+ "acc_stderr": 0.0416180850350153,
+ "acc_norm": 0.5241379310344828,
+ "acc_norm_stderr": 0.0416180850350153
+ },
+ "harness|mmlu_college_physics|5": {
+ "acc": 0.3235294117647059,
+ "acc_stderr": 0.046550104113196177,
+ "acc_norm": 0.3235294117647059,
+ "acc_norm_stderr": 0.046550104113196177
+ },
+ "harness|mmlu_high_school_microeconomics|5": {
+ "acc": 0.6764705882352942,
+ "acc_stderr": 0.030388353551886793,
+ "acc_norm": 0.6764705882352942,
+ "acc_norm_stderr": 0.030388353551886793
+ },
+ "harness|mmlu_high_school_macroeconomics|5": {
+ "acc": 0.6384615384615384,
+ "acc_stderr": 0.024359581465397,
+ "acc_norm": 0.6384615384615384,
+ "acc_norm_stderr": 0.024359581465397
+ },
+ "harness|mmlu_computer_security|5": {
+ "acc": 0.65,
+ "acc_stderr": 0.0479372485441102,
+ "acc_norm": 0.65,
+ "acc_norm_stderr": 0.0479372485441102
+ },
+ "harness|mmlu_global_facts|5": {
+ "acc": 0.31,
+ "acc_stderr": 0.04648231987117316,
+ "acc_norm": 0.31,
+ "acc_norm_stderr": 0.04648231987117316
+ },
+ "harness|mmlu_jurisprudence|5": {
+ "acc": 0.6296296296296297,
+ "acc_stderr": 0.04668408033024931,
+ "acc_norm": 0.6296296296296297,
+ "acc_norm_stderr": 0.04668408033024931
+ },
+ "harness|mmlu_high_school_chemistry|5": {
+ "acc": 0.4729064039408867,
+ "acc_stderr": 0.03512819077876105,
+ "acc_norm": 0.4729064039408867,
+ "acc_norm_stderr": 0.03512819077876105
+ },
+ "harness|mmlu_high_school_biology|5": {
+ "acc": 0.5709677419354838,
+ "acc_stderr": 0.028156036538233193,
+ "acc_norm": 0.5709677419354838,
+ "acc_norm_stderr": 0.028156036538233193
+ },
+ "harness|mmlu_marketing|5": {
+ "acc": 0.8034188034188035,
+ "acc_stderr": 0.026035386098951292,
+ "acc_norm": 0.8034188034188035,
+ "acc_norm_stderr": 0.026035386098951292
+ },
+ "harness|mmlu_clinical_knowledge|5": {
+ "acc": 0.5547169811320755,
+ "acc_stderr": 0.030588052974270655,
+ "acc_norm": 0.5547169811320755,
+ "acc_norm_stderr": 0.030588052974270655
+ },
+ "harness|mmlu_public_relations|5": {
+ "acc": 0.6363636363636364,
+ "acc_stderr": 0.04607582090719976,
+ "acc_norm": 0.6363636363636364,
+ "acc_norm_stderr": 0.04607582090719976
+ },
+ "harness|mmlu_high_school_mathematics|5": {
+ "acc": 0.3592592592592593,
+ "acc_stderr": 0.029252905927251976,
+ "acc_norm": 0.3592592592592593,
+ "acc_norm_stderr": 0.029252905927251976
+ },
+ "harness|mmlu_high_school_physics|5": {
+ "acc": 0.3576158940397351,
+ "acc_stderr": 0.03913453431177258,
+ "acc_norm": 0.3576158940397351,
+ "acc_norm_stderr": 0.03913453431177258
+ },
+ "harness|mmlu_sociology|5": {
+ "acc": 0.6268656716417911,
+ "acc_stderr": 0.034198326081760065,
+ "acc_norm": 0.6268656716417911,
+ "acc_norm_stderr": 0.034198326081760065
+ },
+ "harness|mmlu_college_medicine|5": {
+ "acc": 0.48554913294797686,
+ "acc_stderr": 0.03810871630454764,
+ "acc_norm": 0.48554913294797686,
+ "acc_norm_stderr": 0.03810871630454764
+ },
+ "harness|mmlu_elementary_mathematics|5": {
+ "acc": 0.4497354497354497,
+ "acc_stderr": 0.025620857042936648,
+ "acc_norm": 0.4497354497354497,
+ "acc_norm_stderr": 0.025620857042936648
+ },
+ "harness|mmlu_college_biology|5": {
+ "acc": 0.6041666666666666,
+ "acc_stderr": 0.04089465449325582,
+ "acc_norm": 0.6041666666666666,
+ "acc_norm_stderr": 0.04089465449325582
+ },
+ "harness|mmlu_college_chemistry|5": {
+ "acc": 0.32,
+ "acc_stderr": 0.046882617226215034,
+ "acc_norm": 0.32,
+ "acc_norm_stderr": 0.046882617226215034
+ },
+ "harness|mmlu_us_foreign_policy|5": {
+ "acc": 0.71,
+ "acc_stderr": 0.045604802157206824,
+ "acc_norm": 0.71,
+ "acc_norm_stderr": 0.045604802157206824
+ },
+ "harness|mmlu_moral_disputes|5": {
+ "acc": 0.5664739884393064,
+ "acc_stderr": 0.026680134761679217,
+ "acc_norm": 0.5664739884393064,
+ "acc_norm_stderr": 0.026680134761679217
+ },
+ "harness|mmlu_logical_fallacies|5": {
+ "acc": 0.6196319018404908,
+ "acc_stderr": 0.038142698932618374,
+ "acc_norm": 0.6196319018404908,
+ "acc_norm_stderr": 0.038142698932618374
+ },
+ "harness|mmlu_prehistory|5": {
+ "acc": 0.6574074074074074,
+ "acc_stderr": 0.026406145973625686,
+ "acc_norm": 0.6574074074074074,
+ "acc_norm_stderr": 0.026406145973625686
+ },
+ "harness|mmlu_college_mathematics|5": {
+ "acc": 0.37,
+ "acc_stderr": 0.04852365870939098,
+ "acc_norm": 0.37,
+ "acc_norm_stderr": 0.04852365870939098
+ },
+ "harness|mmlu_high_school_government_and_politics|5": {
+ "acc": 0.7616580310880829,
+ "acc_stderr": 0.030748905363909895,
+ "acc_norm": 0.7616580310880829,
+ "acc_norm_stderr": 0.030748905363909895
+ },
+ "harness|mmlu_econometrics|5": {
+ "acc": 0.5,
+ "acc_stderr": 0.047036043419179864,
+ "acc_norm": 0.5,
+ "acc_norm_stderr": 0.047036043419179864
+ },
+ "harness|mmlu_high_school_psychology|5": {
+ "acc": 0.7211009174311926,
+ "acc_stderr": 0.01922746887646353,
+ "acc_norm": 0.7211009174311926,
+ "acc_norm_stderr": 0.01922746887646353
+ },
+ "harness|mmlu_formal_logic|5": {
+ "acc": 0.42857142857142855,
+ "acc_stderr": 0.0442626668137991,
+ "acc_norm": 0.42857142857142855,
+ "acc_norm_stderr": 0.0442626668137991
+ },
+ "harness|mmlu_nutrition|5": {
+ "acc": 0.5816993464052288,
+ "acc_stderr": 0.0282451340243873,
+ "acc_norm": 0.5816993464052288,
+ "acc_norm_stderr": 0.0282451340243873
+ },
+ "harness|mmlu_business_ethics|5": {
+ "acc": 0.73,
+ "acc_stderr": 0.044619604333847394,
+ "acc_norm": 0.73,
+ "acc_norm_stderr": 0.044619604333847394
+ },
+ "harness|mmlu_international_law|5": {
+ "acc": 0.7107438016528925,
+ "acc_stderr": 0.041391127276354626,
+ "acc_norm": 0.7107438016528925,
+ "acc_norm_stderr": 0.041391127276354626
+ },
+ "harness|mmlu_astronomy|5": {
+ "acc": 0.6513157894736842,
+ "acc_stderr": 0.038781398887976104,
+ "acc_norm": 0.6513157894736842,
+ "acc_norm_stderr": 0.038781398887976104
+ },
+ "harness|mmlu_professional_psychology|5": {
+ "acc": 0.5686274509803921,
+ "acc_stderr": 0.020036393768352624,
+ "acc_norm": 0.5686274509803921,
+ "acc_norm_stderr": 0.020036393768352624
+ },
+ "harness|mmlu_professional_accounting|5": {
+ "acc": 0.45390070921985815,
+ "acc_stderr": 0.029700453247291477,
+ "acc_norm": 0.45390070921985815,
+ "acc_norm_stderr": 0.029700453247291477
+ },
+ "harness|mmlu_machine_learning|5": {
+ "acc": 0.4642857142857143,
+ "acc_stderr": 0.04733667890053756,
+ "acc_norm": 0.4642857142857143,
+ "acc_norm_stderr": 0.04733667890053756
+ },
+ "harness|mmlu_high_school_statistics|5": {
+ "acc": 0.5092592592592593,
+ "acc_stderr": 0.034093869469927006,
+ "acc_norm": 0.5092592592592593,
+ "acc_norm_stderr": 0.034093869469927006
+ },
+ "harness|mmlu_moral_scenarios|5": {
+ "acc": 0.37206703910614525,
+ "acc_stderr": 0.016165847583563295,
+ "acc_norm": 0.37206703910614525,
+ "acc_norm_stderr": 0.016165847583563295
+ },
+ "harness|mmlu_college_computer_science|5": {
+ "acc": 0.43,
+ "acc_stderr": 0.049756985195624284,
+ "acc_norm": 0.43,
+ "acc_norm_stderr": 0.049756985195624284
+ },
+ "harness|mmlu_high_school_computer_science|5": {
+ "acc": 0.71,
+ "acc_stderr": 0.045604802157206845,
+ "acc_norm": 0.71,
+ "acc_norm_stderr": 0.045604802157206845
+ },
+ "harness|mmlu_professional_medicine|5": {
+ "acc": 0.5404411764705882,
+ "acc_stderr": 0.030273325077345755,
+ "acc_norm": 0.5404411764705882,
+ "acc_norm_stderr": 0.030273325077345755
+ },
+ "harness|mmlu_security_studies|5": {
+ "acc": 0.6122448979591837,
+ "acc_stderr": 0.03119223072679566,
+ "acc_norm": 0.6122448979591837,
+ "acc_norm_stderr": 0.03119223072679566
+ },
+ "harness|mmlu_high_school_world_history|5": {
+ "acc": 0.7257383966244726,
+ "acc_stderr": 0.029041333510598025,
+ "acc_norm": 0.7257383966244726,
+ "acc_norm_stderr": 0.029041333510598025
+ },
+ "harness|mmlu_professional_law|5": {
+ "acc": 0.4641460234680574,
+ "acc_stderr": 0.01273736131873058,
+ "acc_norm": 0.4641460234680574,
+ "acc_norm_stderr": 0.01273736131873058
+ },
+ "harness|mmlu_high_school_us_history|5": {
+ "acc": 0.6568627450980392,
+ "acc_stderr": 0.03332139944668086,
+ "acc_norm": 0.6568627450980392,
+ "acc_norm_stderr": 0.03332139944668086
+ },
+ "harness|mmlu_high_school_european_history|5": {
+ "acc": 0.6,
+ "acc_stderr": 0.03825460278380025,
+ "acc_norm": 0.6,
+ "acc_norm_stderr": 0.03825460278380025
+ },
+ "harness|truthfulqa_mc|0": {
+ "mc1": 0.7246022031823746,
+ "mc1_stderr": 0.01563813566777552,
+ "mc2": 0.8107575910195236,
+ "mc2_stderr": 0.013335029489665237
+ },
+ "harness|commongen_v2|2": {
+ "acc": 0.525383707201889,
+ "acc_stderr": 0.017168187201429253,
+ "acc_norm": 0.5442739079102715,
+ "acc_norm_stderr": 0.017122829143292655
+ }
+ },
+ "versions": {
+ "all": 0,
+ "harness|arc_challenge|25": 0,
+ "harness|hellaswag|10": 0,
+ "harness|mmlu_world_religions|5": 1,
+ "harness|mmlu_management|5": 1,
+ "harness|mmlu_miscellaneous|5": 1,
+ "harness|mmlu_anatomy|5": 1,
+ "harness|mmlu_abstract_algebra|5": 1,
+ "harness|mmlu_conceptual_physics|5": 1,
+ "harness|mmlu_virology|5": 1,
+ "harness|mmlu_philosophy|5": 1,
+ "harness|mmlu_human_aging|5": 1,
+ "harness|mmlu_human_sexuality|5": 1,
+ "harness|mmlu_medical_genetics|5": 1,
+ "harness|mmlu_high_school_geography|5": 1,
+ "harness|mmlu_electrical_engineering|5": 1,
+ "harness|mmlu_college_physics|5": 1,
+ "harness|mmlu_high_school_microeconomics|5": 1,
+ "harness|mmlu_high_school_macroeconomics|5": 1,
+ "harness|mmlu_computer_security|5": 1,
+ "harness|mmlu_global_facts|5": 1,
+ "harness|mmlu_jurisprudence|5": 1,
+ "harness|mmlu_high_school_chemistry|5": 1,
+ "harness|mmlu_high_school_biology|5": 1,
+ "harness|mmlu_marketing|5": 1,
+ "harness|mmlu_clinical_knowledge|5": 1,
+ "harness|mmlu_public_relations|5": 1,
+ "harness|mmlu_high_school_mathematics|5": 1,
+ "harness|mmlu_high_school_physics|5": 1,
+ "harness|mmlu_sociology|5": 1,
+ "harness|mmlu_college_medicine|5": 1,
+ "harness|mmlu_elementary_mathematics|5": 1,
+ "harness|mmlu_college_biology|5": 1,
+ "harness|mmlu_college_chemistry|5": 1,
+ "harness|mmlu_us_foreign_policy|5": 1,
+ "harness|mmlu_moral_disputes|5": 1,
+ "harness|mmlu_logical_fallacies|5": 1,
+ "harness|mmlu_prehistory|5": 1,
+ "harness|mmlu_college_mathematics|5": 1,
+ "harness|mmlu_high_school_government_and_politics|5": 1,
+ "harness|mmlu_econometrics|5": 1,
+ "harness|mmlu_high_school_psychology|5": 1,
+ "harness|mmlu_formal_logic|5": 1,
+ "harness|mmlu_nutrition|5": 1,
+ "harness|mmlu_business_ethics|5": 1,
+ "harness|mmlu_international_law|5": 1,
+ "harness|mmlu_astronomy|5": 1,
+ "harness|mmlu_professional_psychology|5": 1,
+ "harness|mmlu_professional_accounting|5": 1,
+ "harness|mmlu_machine_learning|5": 1,
+ "harness|mmlu_high_school_statistics|5": 1,
+ "harness|mmlu_moral_scenarios|5": 1,
+ "harness|mmlu_college_computer_science|5": 1,
+ "harness|mmlu_high_school_computer_science|5": 1,
+ "harness|mmlu_professional_medicine|5": 1,
+ "harness|mmlu_security_studies|5": 1,
+ "harness|mmlu_high_school_world_history|5": 1,
+ "harness|mmlu_professional_law|5": 1,
+ "harness|mmlu_high_school_us_history|5": 1,
+ "harness|mmlu_high_school_european_history|5": 1,
+ "harness|truthfulqa_mc|0": 0,
+ "harness|commongen_v2|2": 1
+ },
+ "config_general": {
+ "model_name": "x2bee/POLAR-14B-v0.2",
+ "model_sha": "8d905623a3972e11260420130039c62e115cbbaa",
+ "model_dtype": "torch.float16",
+ "lighteval_sha": "",
+ "num_few_shot_default": 0,
+ "num_fewshot_seeds": 1,
+ "override_batch_size": 1,
+ "max_samples": null
+ }
+}
\ No newline at end of file
diff --git a/eval-results/x2bee/POLAR-14B-v0.5/result.json b/eval-results/x2bee/POLAR-14B-v0.5/result.json
new file mode 100644
index 0000000000000000000000000000000000000000..8cd609e2bdf51cc5b4d666ff5ee4db16a30aed52
--- /dev/null
+++ b/eval-results/x2bee/POLAR-14B-v0.5/result.json
@@ -0,0 +1,450 @@
+{
+ "results": {
+ "daily": {
+ "daily": 1
+ },
+ "quarterly": {
+ "quarterly": 1
+ },
+ "harness|arc_challenge|25": {
+ "acc": 0.75,
+ "acc_stderr": 0.012653835621466646,
+ "acc_norm": 0.7798634812286689,
+ "acc_norm_stderr": 0.012108124883460988
+ },
+ "harness|hellaswag|10": {
+ "acc": 0.6500697072296355,
+ "acc_stderr": 0.004759729267943182,
+ "acc_norm": 0.775542720573591,
+ "acc_norm_stderr": 0.004163717220873764
+ },
+ "harness|mmlu_world_religions|5": {
+ "acc": 0.6374269005847953,
+ "acc_stderr": 0.036871306155620606,
+ "acc_norm": 0.6374269005847953,
+ "acc_norm_stderr": 0.036871306155620606
+ },
+ "harness|mmlu_management|5": {
+ "acc": 0.7087378640776699,
+ "acc_stderr": 0.044986763205729224,
+ "acc_norm": 0.7087378640776699,
+ "acc_norm_stderr": 0.044986763205729224
+ },
+ "harness|mmlu_miscellaneous|5": {
+ "acc": 0.6730523627075351,
+ "acc_stderr": 0.016774908180131484,
+ "acc_norm": 0.6730523627075351,
+ "acc_norm_stderr": 0.016774908180131484
+ },
+ "harness|mmlu_anatomy|5": {
+ "acc": 0.45185185185185184,
+ "acc_stderr": 0.04299268905480864,
+ "acc_norm": 0.45185185185185184,
+ "acc_norm_stderr": 0.04299268905480864
+ },
+ "harness|mmlu_abstract_algebra|5": {
+ "acc": 0.36,
+ "acc_stderr": 0.048241815132442176,
+ "acc_norm": 0.36,
+ "acc_norm_stderr": 0.048241815132442176
+ },
+ "harness|mmlu_conceptual_physics|5": {
+ "acc": 0.4723404255319149,
+ "acc_stderr": 0.03263597118409769,
+ "acc_norm": 0.4723404255319149,
+ "acc_norm_stderr": 0.03263597118409769
+ },
+ "harness|mmlu_virology|5": {
+ "acc": 0.46987951807228917,
+ "acc_stderr": 0.03885425420866766,
+ "acc_norm": 0.46987951807228917,
+ "acc_norm_stderr": 0.03885425420866766
+ },
+ "harness|mmlu_philosophy|5": {
+ "acc": 0.594855305466238,
+ "acc_stderr": 0.027882383791325963,
+ "acc_norm": 0.594855305466238,
+ "acc_norm_stderr": 0.027882383791325963
+ },
+ "harness|mmlu_human_aging|5": {
+ "acc": 0.6412556053811659,
+ "acc_stderr": 0.032190792004199956,
+ "acc_norm": 0.6412556053811659,
+ "acc_norm_stderr": 0.032190792004199956
+ },
+ "harness|mmlu_human_sexuality|5": {
+ "acc": 0.5954198473282443,
+ "acc_stderr": 0.043046937953806645,
+ "acc_norm": 0.5954198473282443,
+ "acc_norm_stderr": 0.043046937953806645
+ },
+ "harness|mmlu_medical_genetics|5": {
+ "acc": 0.5,
+ "acc_stderr": 0.050251890762960605,
+ "acc_norm": 0.5,
+ "acc_norm_stderr": 0.050251890762960605
+ },
+ "harness|mmlu_high_school_geography|5": {
+ "acc": 0.7272727272727273,
+ "acc_stderr": 0.03173071239071724,
+ "acc_norm": 0.7272727272727273,
+ "acc_norm_stderr": 0.03173071239071724
+ },
+ "harness|mmlu_electrical_engineering|5": {
+ "acc": 0.503448275862069,
+ "acc_stderr": 0.0416656757710158,
+ "acc_norm": 0.503448275862069,
+ "acc_norm_stderr": 0.0416656757710158
+ },
+ "harness|mmlu_college_physics|5": {
+ "acc": 0.3431372549019608,
+ "acc_stderr": 0.04724007352383888,
+ "acc_norm": 0.3431372549019608,
+ "acc_norm_stderr": 0.04724007352383888
+ },
+ "harness|mmlu_high_school_microeconomics|5": {
+ "acc": 0.6596638655462185,
+ "acc_stderr": 0.03077805742293167,
+ "acc_norm": 0.6596638655462185,
+ "acc_norm_stderr": 0.03077805742293167
+ },
+ "harness|mmlu_high_school_macroeconomics|5": {
+ "acc": 0.6102564102564103,
+ "acc_stderr": 0.024726967886647078,
+ "acc_norm": 0.6102564102564103,
+ "acc_norm_stderr": 0.024726967886647078
+ },
+ "harness|mmlu_computer_security|5": {
+ "acc": 0.67,
+ "acc_stderr": 0.047258156262526094,
+ "acc_norm": 0.67,
+ "acc_norm_stderr": 0.047258156262526094
+ },
+ "harness|mmlu_global_facts|5": {
+ "acc": 0.33,
+ "acc_stderr": 0.047258156262526045,
+ "acc_norm": 0.33,
+ "acc_norm_stderr": 0.047258156262526045
+ },
+ "harness|mmlu_jurisprudence|5": {
+ "acc": 0.6481481481481481,
+ "acc_stderr": 0.04616631111801714,
+ "acc_norm": 0.6481481481481481,
+ "acc_norm_stderr": 0.04616631111801714
+ },
+ "harness|mmlu_high_school_chemistry|5": {
+ "acc": 0.4729064039408867,
+ "acc_stderr": 0.03512819077876105,
+ "acc_norm": 0.4729064039408867,
+ "acc_norm_stderr": 0.03512819077876105
+ },
+ "harness|mmlu_high_school_biology|5": {
+ "acc": 0.5709677419354838,
+ "acc_stderr": 0.028156036538233193,
+ "acc_norm": 0.5709677419354838,
+ "acc_norm_stderr": 0.028156036538233193
+ },
+ "harness|mmlu_marketing|5": {
+ "acc": 0.7735042735042735,
+ "acc_stderr": 0.027421007295392943,
+ "acc_norm": 0.7735042735042735,
+ "acc_norm_stderr": 0.027421007295392943
+ },
+ "harness|mmlu_clinical_knowledge|5": {
+ "acc": 0.5660377358490566,
+ "acc_stderr": 0.030503292013342596,
+ "acc_norm": 0.5660377358490566,
+ "acc_norm_stderr": 0.030503292013342596
+ },
+ "harness|mmlu_public_relations|5": {
+ "acc": 0.6272727272727273,
+ "acc_stderr": 0.04631381319425465,
+ "acc_norm": 0.6272727272727273,
+ "acc_norm_stderr": 0.04631381319425465
+ },
+ "harness|mmlu_high_school_mathematics|5": {
+ "acc": 0.3333333333333333,
+ "acc_stderr": 0.0287420409039485,
+ "acc_norm": 0.3333333333333333,
+ "acc_norm_stderr": 0.0287420409039485
+ },
+ "harness|mmlu_high_school_physics|5": {
+ "acc": 0.39072847682119205,
+ "acc_stderr": 0.039837983066598075,
+ "acc_norm": 0.39072847682119205,
+ "acc_norm_stderr": 0.039837983066598075
+ },
+ "harness|mmlu_sociology|5": {
+ "acc": 0.6417910447761194,
+ "acc_stderr": 0.03390393042268814,
+ "acc_norm": 0.6417910447761194,
+ "acc_norm_stderr": 0.03390393042268814
+ },
+ "harness|mmlu_college_medicine|5": {
+ "acc": 0.5028901734104047,
+ "acc_stderr": 0.038124005659748335,
+ "acc_norm": 0.5028901734104047,
+ "acc_norm_stderr": 0.038124005659748335
+ },
+ "harness|mmlu_elementary_mathematics|5": {
+ "acc": 0.42857142857142855,
+ "acc_stderr": 0.025487187147859372,
+ "acc_norm": 0.42857142857142855,
+ "acc_norm_stderr": 0.025487187147859372
+ },
+ "harness|mmlu_college_biology|5": {
+ "acc": 0.6180555555555556,
+ "acc_stderr": 0.040629907841466674,
+ "acc_norm": 0.6180555555555556,
+ "acc_norm_stderr": 0.040629907841466674
+ },
+ "harness|mmlu_college_chemistry|5": {
+ "acc": 0.3,
+ "acc_stderr": 0.046056618647183814,
+ "acc_norm": 0.3,
+ "acc_norm_stderr": 0.046056618647183814
+ },
+ "harness|mmlu_us_foreign_policy|5": {
+ "acc": 0.72,
+ "acc_stderr": 0.04512608598542127,
+ "acc_norm": 0.72,
+ "acc_norm_stderr": 0.04512608598542127
+ },
+ "harness|mmlu_moral_disputes|5": {
+ "acc": 0.5809248554913294,
+ "acc_stderr": 0.026564178111422622,
+ "acc_norm": 0.5809248554913294,
+ "acc_norm_stderr": 0.026564178111422622
+ },
+ "harness|mmlu_logical_fallacies|5": {
+ "acc": 0.6257668711656442,
+ "acc_stderr": 0.03802068102899615,
+ "acc_norm": 0.6257668711656442,
+ "acc_norm_stderr": 0.03802068102899615
+ },
+ "harness|mmlu_prehistory|5": {
+ "acc": 0.5987654320987654,
+ "acc_stderr": 0.027272582849839803,
+ "acc_norm": 0.5987654320987654,
+ "acc_norm_stderr": 0.027272582849839803
+ },
+ "harness|mmlu_college_mathematics|5": {
+ "acc": 0.34,
+ "acc_stderr": 0.04760952285695235,
+ "acc_norm": 0.34,
+ "acc_norm_stderr": 0.04760952285695235
+ },
+ "harness|mmlu_high_school_government_and_politics|5": {
+ "acc": 0.7512953367875648,
+ "acc_stderr": 0.031195840877700304,
+ "acc_norm": 0.7512953367875648,
+ "acc_norm_stderr": 0.031195840877700304
+ },
+ "harness|mmlu_econometrics|5": {
+ "acc": 0.47368421052631576,
+ "acc_stderr": 0.046970851366478626,
+ "acc_norm": 0.47368421052631576,
+ "acc_norm_stderr": 0.046970851366478626
+ },
+ "harness|mmlu_high_school_psychology|5": {
+ "acc": 0.7229357798165138,
+ "acc_stderr": 0.019188482590169538,
+ "acc_norm": 0.7229357798165138,
+ "acc_norm_stderr": 0.019188482590169538
+ },
+ "harness|mmlu_formal_logic|5": {
+ "acc": 0.4523809523809524,
+ "acc_stderr": 0.044518079590553275,
+ "acc_norm": 0.4523809523809524,
+ "acc_norm_stderr": 0.044518079590553275
+ },
+ "harness|mmlu_nutrition|5": {
+ "acc": 0.5718954248366013,
+ "acc_stderr": 0.028332397483664278,
+ "acc_norm": 0.5718954248366013,
+ "acc_norm_stderr": 0.028332397483664278
+ },
+ "harness|mmlu_business_ethics|5": {
+ "acc": 0.68,
+ "acc_stderr": 0.04688261722621504,
+ "acc_norm": 0.68,
+ "acc_norm_stderr": 0.04688261722621504
+ },
+ "harness|mmlu_international_law|5": {
+ "acc": 0.7520661157024794,
+ "acc_stderr": 0.039418975265163025,
+ "acc_norm": 0.7520661157024794,
+ "acc_norm_stderr": 0.039418975265163025
+ },
+ "harness|mmlu_astronomy|5": {
+ "acc": 0.618421052631579,
+ "acc_stderr": 0.03953173377749194,
+ "acc_norm": 0.618421052631579,
+ "acc_norm_stderr": 0.03953173377749194
+ },
+ "harness|mmlu_professional_psychology|5": {
+ "acc": 0.5408496732026143,
+ "acc_stderr": 0.020160213617222516,
+ "acc_norm": 0.5408496732026143,
+ "acc_norm_stderr": 0.020160213617222516
+ },
+ "harness|mmlu_professional_accounting|5": {
+ "acc": 0.45390070921985815,
+ "acc_stderr": 0.029700453247291463,
+ "acc_norm": 0.45390070921985815,
+ "acc_norm_stderr": 0.029700453247291463
+ },
+ "harness|mmlu_machine_learning|5": {
+ "acc": 0.44642857142857145,
+ "acc_stderr": 0.04718471485219588,
+ "acc_norm": 0.44642857142857145,
+ "acc_norm_stderr": 0.04718471485219588
+ },
+ "harness|mmlu_high_school_statistics|5": {
+ "acc": 0.5416666666666666,
+ "acc_stderr": 0.03398110890294636,
+ "acc_norm": 0.5416666666666666,
+ "acc_norm_stderr": 0.03398110890294636
+ },
+ "harness|mmlu_moral_scenarios|5": {
+ "acc": 0.35195530726256985,
+ "acc_stderr": 0.01597266852368907,
+ "acc_norm": 0.35195530726256985,
+ "acc_norm_stderr": 0.01597266852368907
+ },
+ "harness|mmlu_college_computer_science|5": {
+ "acc": 0.44,
+ "acc_stderr": 0.0498887651569859,
+ "acc_norm": 0.44,
+ "acc_norm_stderr": 0.0498887651569859
+ },
+ "harness|mmlu_high_school_computer_science|5": {
+ "acc": 0.68,
+ "acc_stderr": 0.04688261722621503,
+ "acc_norm": 0.68,
+ "acc_norm_stderr": 0.04688261722621503
+ },
+ "harness|mmlu_professional_medicine|5": {
+ "acc": 0.5147058823529411,
+ "acc_stderr": 0.03035969707904612,
+ "acc_norm": 0.5147058823529411,
+ "acc_norm_stderr": 0.03035969707904612
+ },
+ "harness|mmlu_security_studies|5": {
+ "acc": 0.6122448979591837,
+ "acc_stderr": 0.031192230726795656,
+ "acc_norm": 0.6122448979591837,
+ "acc_norm_stderr": 0.031192230726795656
+ },
+ "harness|mmlu_high_school_world_history|5": {
+ "acc": 0.7215189873417721,
+ "acc_stderr": 0.029178682304842538,
+ "acc_norm": 0.7215189873417721,
+ "acc_norm_stderr": 0.029178682304842538
+ },
+ "harness|mmlu_professional_law|5": {
+ "acc": 0.4634941329856584,
+ "acc_stderr": 0.012736153390214963,
+ "acc_norm": 0.4634941329856584,
+ "acc_norm_stderr": 0.012736153390214963
+ },
+ "harness|mmlu_high_school_us_history|5": {
+ "acc": 0.6568627450980392,
+ "acc_stderr": 0.03332139944668086,
+ "acc_norm": 0.6568627450980392,
+ "acc_norm_stderr": 0.03332139944668086
+ },
+ "harness|mmlu_high_school_european_history|5": {
+ "acc": 0.5818181818181818,
+ "acc_stderr": 0.03851716319398393,
+ "acc_norm": 0.5818181818181818,
+ "acc_norm_stderr": 0.03851716319398393
+ },
+ "harness|truthfulqa_mc|0": {
+ "mc1": 0.7833537331701347,
+ "mc1_stderr": 0.014421468452506978,
+ "mc2": 0.8572574997405501,
+ "mc2_stderr": 0.01200311225898601
+ },
+ "harness|commongen_v2|2": {
+ "acc": 0.5159386068476978,
+ "acc_stderr": 0.017181617837190195,
+ "acc_norm": 0.5301062573789846,
+ "acc_norm_stderr": 0.01715916359017022
+ }
+ },
+ "versions": {
+ "all": 0,
+ "harness|arc_challenge|25": 0,
+ "harness|hellaswag|10": 0,
+ "harness|mmlu_world_religions|5": 1,
+ "harness|mmlu_management|5": 1,
+ "harness|mmlu_miscellaneous|5": 1,
+ "harness|mmlu_anatomy|5": 1,
+ "harness|mmlu_abstract_algebra|5": 1,
+ "harness|mmlu_conceptual_physics|5": 1,
+ "harness|mmlu_virology|5": 1,
+ "harness|mmlu_philosophy|5": 1,
+ "harness|mmlu_human_aging|5": 1,
+ "harness|mmlu_human_sexuality|5": 1,
+ "harness|mmlu_medical_genetics|5": 1,
+ "harness|mmlu_high_school_geography|5": 1,
+ "harness|mmlu_electrical_engineering|5": 1,
+ "harness|mmlu_college_physics|5": 1,
+ "harness|mmlu_high_school_microeconomics|5": 1,
+ "harness|mmlu_high_school_macroeconomics|5": 1,
+ "harness|mmlu_computer_security|5": 1,
+ "harness|mmlu_global_facts|5": 1,
+ "harness|mmlu_jurisprudence|5": 1,
+ "harness|mmlu_high_school_chemistry|5": 1,
+ "harness|mmlu_high_school_biology|5": 1,
+ "harness|mmlu_marketing|5": 1,
+ "harness|mmlu_clinical_knowledge|5": 1,
+ "harness|mmlu_public_relations|5": 1,
+ "harness|mmlu_high_school_mathematics|5": 1,
+ "harness|mmlu_high_school_physics|5": 1,
+ "harness|mmlu_sociology|5": 1,
+ "harness|mmlu_college_medicine|5": 1,
+ "harness|mmlu_elementary_mathematics|5": 1,
+ "harness|mmlu_college_biology|5": 1,
+ "harness|mmlu_college_chemistry|5": 1,
+ "harness|mmlu_us_foreign_policy|5": 1,
+ "harness|mmlu_moral_disputes|5": 1,
+ "harness|mmlu_logical_fallacies|5": 1,
+ "harness|mmlu_prehistory|5": 1,
+ "harness|mmlu_college_mathematics|5": 1,
+ "harness|mmlu_high_school_government_and_politics|5": 1,
+ "harness|mmlu_econometrics|5": 1,
+ "harness|mmlu_high_school_psychology|5": 1,
+ "harness|mmlu_formal_logic|5": 1,
+ "harness|mmlu_nutrition|5": 1,
+ "harness|mmlu_business_ethics|5": 1,
+ "harness|mmlu_international_law|5": 1,
+ "harness|mmlu_astronomy|5": 1,
+ "harness|mmlu_professional_psychology|5": 1,
+ "harness|mmlu_professional_accounting|5": 1,
+ "harness|mmlu_machine_learning|5": 1,
+ "harness|mmlu_high_school_statistics|5": 1,
+ "harness|mmlu_moral_scenarios|5": 1,
+ "harness|mmlu_college_computer_science|5": 1,
+ "harness|mmlu_high_school_computer_science|5": 1,
+ "harness|mmlu_professional_medicine|5": 1,
+ "harness|mmlu_security_studies|5": 1,
+ "harness|mmlu_high_school_world_history|5": 1,
+ "harness|mmlu_professional_law|5": 1,
+ "harness|mmlu_high_school_us_history|5": 1,
+ "harness|mmlu_high_school_european_history|5": 1,
+ "harness|truthfulqa_mc|0": 0,
+ "harness|commongen_v2|2": 1
+ },
+ "config_general": {
+ "model_name": "x2bee/POLAR-14B-v0.5",
+ "model_sha": "74a1ef65a8d650e5358be229def31688738d8c6a",
+ "model_dtype": "torch.float16",
+ "lighteval_sha": "",
+ "num_few_shot_default": 0,
+ "num_fewshot_seeds": 1,
+ "override_batch_size": 1,
+ "max_samples": null
+ }
+}
\ No newline at end of file
diff --git a/requirements.txt b/requirements.txt
new file mode 100755
index 0000000000000000000000000000000000000000..ed4ea4ad1c28eca4cafb4fc1925a92446a8d9efc
--- /dev/null
+++ b/requirements.txt
@@ -0,0 +1,19 @@
+APScheduler==3.10.1
+black==23.11.0
+click==8.1.3
+datasets==2.14.5
+gradio==4.19.2
+gradio_client==0.10.1
+huggingface-hub>=0.18.0
+matplotlib==3.7.1
+numpy==1.24.2
+pandas==2.0.0
+plotly==5.14.1
+python-dateutil==2.8.2
+requests==2.28.2
+sentencepiece
+tqdm==4.65.0
+transformers==4.38.2
+tokenizers>=0.15.0
+gradio-space-ci @ git+https://huggingface.co/spaces/Wauplin/gradio-space-ci@0.1.2 # CI !!!
+torch
\ No newline at end of file
diff --git a/scripts/create_request_file.py b/scripts/create_request_file.py
new file mode 100755
index 0000000000000000000000000000000000000000..56edc72e65a479d05d4254e1a90faa214dd9cb05
--- /dev/null
+++ b/scripts/create_request_file.py
@@ -0,0 +1,107 @@
+import json
+import os
+import pprint
+import re
+from datetime import datetime, timezone
+
+import click
+from colorama import Fore
+from huggingface_hub import HfApi, snapshot_download
+
+EVAL_REQUESTS_PATH = "eval-queue"
+QUEUE_REPO = "open-ko-llm-leaderboard/requests"
+
+precisions = ("float16", "bfloat16", "8bit (LLM.int8)", "4bit (QLoRA / FP4)", "GPTQ")
+model_types = ("pretrained", "fine-tuned", "RL-tuned", "instruction-tuned")
+weight_types = ("Original", "Delta", "Adapter")
+
+
+def get_model_size(model_info, precision: str):
+ size_pattern = re.compile(r"(\d+\.)?\d+(b|m)")
+ try:
+ model_size = round(model_info.safetensors["total"] / 1e9, 3)
+ except (AttributeError, TypeError):
+ try:
+ size_match = re.search(size_pattern, model_info.modelId.lower())
+ model_size = size_match.group(0)
+ model_size = round(float(model_size[:-1]) if model_size[-1] == "b" else float(model_size[:-1]) / 1e3, 3)
+ except AttributeError:
+ return 0 # Unknown model sizes are indicated as 0, see NUMERIC_INTERVALS in app.py
+
+ size_factor = 8 if (precision == "GPTQ" or "gptq" in model_info.modelId.lower()) else 1
+ model_size = size_factor * model_size
+ return model_size
+
+
+def main():
+ api = HfApi()
+ current_time = datetime.now(timezone.utc).strftime("%Y-%m-%dT%H:%M:%SZ")
+ snapshot_download(repo_id=QUEUE_REPO, revision="main", local_dir=EVAL_REQUESTS_PATH, repo_type="dataset")
+
+ model_name = click.prompt("Enter model name")
+ revision = click.prompt("Enter revision", default="main")
+ precision = click.prompt("Enter precision", default="float16", type=click.Choice(precisions))
+ model_type = click.prompt("Enter model type", type=click.Choice(model_types))
+ weight_type = click.prompt("Enter weight type", default="Original", type=click.Choice(weight_types))
+ base_model = click.prompt("Enter base model", default="")
+ status = click.prompt("Enter status", default="FINISHED")
+
+ try:
+ model_info = api.model_info(repo_id=model_name, revision=revision)
+ except Exception as e:
+ print(f"{Fore.RED}Could not find model info for {model_name} on the Hub\n{e}{Fore.RESET}")
+ return 1
+
+ model_size = get_model_size(model_info=model_info, precision=precision)
+
+ try:
+ license = model_info.cardData["license"]
+ except Exception:
+ license = "?"
+
+ eval_entry = {
+ "model": model_name,
+ "base_model": base_model,
+ "revision": revision,
+ "private": False,
+ "precision": precision,
+ "weight_type": weight_type,
+ "status": status,
+ "submitted_time": current_time,
+ "model_type": model_type,
+ "likes": model_info.likes,
+ "params": model_size,
+ "license": license,
+ }
+
+ user_name = ""
+ model_path = model_name
+ if "/" in model_name:
+ user_name = model_name.split("/")[0]
+ model_path = model_name.split("/")[1]
+
+ pprint.pprint(eval_entry)
+
+ if click.confirm("Do you want to continue? This request file will be pushed to the hub"):
+ click.echo("continuing...")
+
+ out_dir = f"{EVAL_REQUESTS_PATH}/{user_name}"
+ os.makedirs(out_dir, exist_ok=True)
+ out_path = f"{out_dir}/{model_path}_eval_request_{False}_{precision}_{weight_type}.json"
+
+ with open(out_path, "w") as f:
+ f.write(json.dumps(eval_entry))
+
+ api.upload_file(
+ path_or_fileobj=out_path,
+ path_in_repo=out_path.split(f"{EVAL_REQUESTS_PATH}/")[1],
+ repo_id=QUEUE_REPO,
+ repo_type="dataset",
+ commit_message=f"Add {model_name} to eval queue",
+ )
+ else:
+ click.echo("aborting...")
+
+
+if __name__ == "__main__":
+ main()
diff --git a/scripts/update_request_files.py b/scripts/update_request_files.py
new file mode 100755
index 0000000000000000000000000000000000000000..5a15079c799a8a592dda0197e4d53e26f7b67a75
--- /dev/null
+++ b/scripts/update_request_files.py
@@ -0,0 +1,82 @@
+import json
+import os
+import glob
+import pprint
+import re
+from datetime import datetime, timezone
+
+import click
+from colorama import Fore
+from huggingface_hub import HfApi, snapshot_download
+from huggingface_hub.hf_api import ModelInfo
+
+API = HfApi()
+
+
+def get_model_size(model_info: ModelInfo, precision: str):
+ size_pattern = re.compile(r"(\d+\.)?\d+(b|m)")
+ try:
+ model_size = round(model_info.safetensors["total"] / 1e9, 3)
+ except (AttributeError, TypeError ):
+ try:
+ size_match = re.search(size_pattern, model_info.modelId.split("/")[-1].lower())
+ model_size = size_match.group(0)
+ model_size = round(float(model_size[:-1]) if model_size[-1] == "b" else float(model_size[:-1]) / 1e3, 3)
+ except AttributeError:
+ return 0 # Unknown model sizes are indicated as 0, see NUMERIC_INTERVALS in app.py
+
+ size_factor = 8 if (precision == "GPTQ" or "gptq" in model_info.modelId.split("/")[-1].lower()) else 1
+ model_size = size_factor * model_size
+ return model_size
+
+
+def update_request_files(requests_path):
+ request_files = os.path.join(
+ requests_path, "*/*.json"
+ )
+ request_files = glob.glob(request_files)
+
+ request_files = sorted(request_files, reverse=True)
+ for tmp_request_file in request_files:
+ with open(tmp_request_file, "r") as f:
+ req_content = json.load(f)
+ new_req_content = add_model_info(req_content)
+
+ # if new content is different, update the file
+ if new_req_content != req_content:
+ with open(tmp_request_file, "w") as f:
+ f.write(json.dumps(new_req_content, indent=4))
+
+def add_model_info(entry):
+
+ model = entry["model"]
+ revision = entry["revision"]
+
+ try:
+ model_info = API.model_info(repo_id=model, revision=revision)
+ except Exception:
+ print(f"Could not get model information for {model} revision {revision}")
+ return entry
+
+ new_entry = entry.copy()
+
+ model_size = get_model_size(model_info=model_info, precision='float16')
+ new_entry["params"] = model_size
+
+ new_entry["likes"] = model_info.likes
+
+ # Were the model card and license filled?
+ try:
+ license = model_info.cardData["license"]
+ new_entry["license"] = license
+ except Exception:
+ print(f"No license for {model} revision {revision}")
+
+ print(json.dumps(new_entry, indent=4))
+ return new_entry
+
+
+if __name__ == "__main__":
+ # update_request_files("/Users/sean/workspace/leaderboard/leaderboard-test-requests")
+ update_request_files("/Volumes/Data-case-sensitive/requests")
+
\ No newline at end of file
diff --git a/src/__pycache__/envs.cpython-310.pyc b/src/__pycache__/envs.cpython-310.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..f25811fd5a115e2825ce938713c0932d7f7219cf
Binary files /dev/null and b/src/__pycache__/envs.cpython-310.pyc differ
diff --git a/src/__pycache__/populate.cpython-310.pyc b/src/__pycache__/populate.cpython-310.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..3b759c2ce988d45325e00cb9bcabc4b03722d7a3
Binary files /dev/null and b/src/__pycache__/populate.cpython-310.pyc differ
diff --git a/src/display/__pycache__/about.cpython-310.pyc b/src/display/__pycache__/about.cpython-310.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..7b1487ef8761d7c5372b926ff561ee50e2597545
Binary files /dev/null and b/src/display/__pycache__/about.cpython-310.pyc differ
diff --git a/src/display/__pycache__/css_html_js.cpython-310.pyc b/src/display/__pycache__/css_html_js.cpython-310.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..1545b916cd62f47006792d16b980b4bdebdbcedd
Binary files /dev/null and b/src/display/__pycache__/css_html_js.cpython-310.pyc differ
diff --git a/src/display/__pycache__/formatting.cpython-310.pyc b/src/display/__pycache__/formatting.cpython-310.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..cee5c180af060ff169f5bf7e37a2ff38b587a185
Binary files /dev/null and b/src/display/__pycache__/formatting.cpython-310.pyc differ
diff --git a/src/display/__pycache__/utils.cpython-310.pyc b/src/display/__pycache__/utils.cpython-310.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..7eb97b3a373d80b7475d24d488453623ac502d19
Binary files /dev/null and b/src/display/__pycache__/utils.cpython-310.pyc differ
diff --git a/src/display/about.py b/src/display/about.py
new file mode 100755
index 0000000000000000000000000000000000000000..072a9ebc699e8dc73ca6d314791b31b1d2e2598c
--- /dev/null
+++ b/src/display/about.py
@@ -0,0 +1,84 @@
+from src.display.utils import ModelType
+
+
+TITLE = """
"""
+
+INTRODUCTION_TEXT = f"""
+Welcome to the Self-Improving Leaderboard (SIL) - A Revolutionary Platform for Evaluating Large Language Models
+The SIL offers a dynamic approach to assessing and ranking open-source LLMs and chatbots. Our innovative system continuously updates test datasets and recalculates rankings daily, ensuring evaluations reflect the rapid evolution of language processing capabilities.
+Key Features:
+• Daily-refreshed test datasets
+• Adaptive ranking system
+• Real-world language processing challenges
+• Comprehensive model performance insights
+Explore our cutting-edge evaluation process, gain deep insights into model capabilities, and see how different LLMs compare in this ever-changing landscape.
+Ready to participate? Submit your model for evaluation on the 'Submit' page and join the forefront of LLM advancement. For a detailed look at our methodology, visit the 'About' page.
+The SIL is proudly developed and maintained by [Your Organization/Team Name]. Together, let's push the boundaries of language AI!
+"""
+
+LLM_BENCHMARKS_TEXT = f"""
+# How it works
+🔄 The Self-Improving Leaderboard (SIL) operates on a dynamic evaluation system that continuously evolves to reflect real-world language processing challenges. Here's an overview of our process:
+Daily Dataset Refresh
+Our system generates new test data daily from diverse, reputable sources.
+Advanced Large Language Models (LLMs) are utilized to synthesize additional relevant content.
+The dataset is divided into two sections:
+A primary dataset maintaining the integrity of sourced data
+A noise-injected dataset simulating real-world data complexities
+Model Evaluation
+Participating models are rigorously evaluated against the refreshed dataset every 24 hours.
+We employ a comprehensive set of metrics aligned with industry-standard benchmarks.
+Our evaluation framework is built on the Eleuther AI Language Model Evaluation Harness, ensuring a robust and consistent assessment.
+Ranking System
+Model rankings are updated daily based on their performance across various tasks.
+The leaderboard reflects not only the latest scores but also tracks consistency and adaptability over time.
+Quarterly Comprehensive Evaluation
+Every three months, we conduct an in-depth analysis of model performance.
+This evaluation considers long-term trends, adaptability to evolving data, and overall efficacy.
+Special recognition (e.g., medals or badges) may be awarded based on sustained excellence.
+By continuously refreshing our test data and evaluation criteria, SIL aims to provide a more accurate representation of model performance in real-world scenarios, driving innovation in the field of Natural Language Processing.
+
+## Icons
+{ModelType.PT.to_str(" : ")} model
+{ModelType.IFT.to_str(" : ")} model
+{ModelType.RL.to_str(" : ")} model
+If there is no icon, it indicates that there is insufficient information about the model.
+Please provide information about the model through an issue! 🤩
+
+## Details and Logs
+- Detailed numerical results in the `results` dataset: https://huggingface.co/datasets/junkim100/SIL_results
+- Community queries and running status in the `requests` dataset: https://huggingface.co/datasets/junkim100/SIL_requests
+"""
+
+EVALUATION_QUEUE_TEXT = f"""
+# Evaluation Queue for the 🔄 Self-Improving Leaderboard
+
+##
+
+### 1️⃣ Make sure you can load your model and tokenizer using AutoClasses
+```python
+from transformers import AutoConfig, AutoModel, AutoTokenizer
+config = AutoConfig.from_pretrained("your model name", revision=revision)
+model = AutoModel.from_pretrained("your model name", revision=revision)
+tokenizer = AutoTokenizer.from_pretrained("your model name", revision=revision)
+```
+
+If this step fails, follow the error messages to debug your model before submitting it. It's likely your model has been improperly uploaded.
+
+⚠️ Make sure your model is public!
+
+⚠️ Maker sure your model runs with [Eleuther AI Language Model Evaluation Harness](https://github.com/EleutherAI/lm-evaluation-harness)
+
+
+### 2️⃣ Convert your model weights to [safetensors](https://huggingface.co/docs/safetensors/index)
+It's a new format for storing weights which is safer and faster to load and use. It will also allow us to add the number of parameters of your model to the `Extended Viewer`!
+
+### 3️⃣ Make sure your model has an open license!
+We'd love for as many people as possible to know they can use your model
+
+### 4️⃣ Fill up your model card
+When we add extra information about models to the leaderboard, it will be automatically taken from the model card
+
+## In case of model failure
+If your model is displayed in the `FAILED` category, its execution stopped. Make sure you have followed the above steps first. If everything is done, check you can launch the EleutherAIHarness on your model locally, using the above command without modifications (you can add `--limit` to limit the number of examples per task).
+"""
\ No newline at end of file
diff --git a/src/display/css_html_js.py b/src/display/css_html_js.py
new file mode 100755
index 0000000000000000000000000000000000000000..b812611116a6e5c963994fc138a38a04a011bdce
--- /dev/null
+++ b/src/display/css_html_js.py
@@ -0,0 +1,84 @@
+custom_css = """
+/* Hides the final AutoEvalColumn */
+#llm-benchmark-tab-table table td:last-child,
+#llm-benchmark-tab-table table th:last-child {
+ display: none;
+}
+
+/* Limit the width of the first AutoEvalColumn so that names don't expand too much */
+table td:first-child,
+table th:first-child {
+ max-width: 400px;
+ overflow: auto;
+ white-space: nowrap;
+}
+
+/* Full width space */
+.gradio-container {
+ max-width: 95%!important;
+}
+
+/* Text style and margins */
+.markdown-text {
+ font-size: 16px !important;
+}
+
+#models-to-add-text {
+ font-size: 18px !important;
+}
+
+#search-bar-table-box > div:first-child {
+ background: none;
+ border: none;
+}
+
+#search-bar {
+ padding: 0px;
+}
+
+.tab-buttons button {
+ font-size: 20px;
+}
+
+/* Filters style */
+#filter_type{
+ border: 0;
+ padding-left: 0;
+ padding-top: 0;
+}
+#filter_type label {
+ display: flex;
+}
+#filter_type label > span{
+ margin-top: var(--spacing-lg);
+ margin-right: 0.5em;
+}
+#filter_type label > .wrap{
+ width: 103px;
+}
+#filter_type label > .wrap .wrap-inner{
+ padding: 2px;
+}
+#filter_type label > .wrap .wrap-inner input{
+ width: 1px
+}
+#filter-columns-type{
+ border:0;
+ padding:0.5;
+}
+#filter-columns-size{
+ border:0;
+ padding:0.5;
+}
+#box-filter > .form{
+ border: 0
+}
+"""
+
+get_window_url_params = """
+ function(url_params) {
+ const params = new URLSearchParams(window.location.search);
+ url_params = Object.fromEntries(params);
+ return url_params;
+ }
+ """
diff --git a/src/display/formatting.py b/src/display/formatting.py
new file mode 100755
index 0000000000000000000000000000000000000000..9b0b52ec315cc9725738adec26332b05a9826bc9
--- /dev/null
+++ b/src/display/formatting.py
@@ -0,0 +1,40 @@
+import os
+from datetime import datetime, timezone
+
+from huggingface_hub import HfApi
+from huggingface_hub.hf_api import ModelInfo
+
+
+API = HfApi()
+
+def model_hyperlink(link, model_name):
+ return f'{model_name}'
+
+
+def make_clickable_model(model_name):
+ link = f"https://huggingface.co/{model_name}"
+
+ details_model_name = model_name.replace("/", "__")
+ details_link = f"https://huggingface.co/datasets/open-ko-llm-leaderboard/details_{details_model_name}"
+
+ return model_hyperlink(link, model_name) + " " + model_hyperlink(details_link, "📑")
+
+
+def styled_error(error):
+ return f"{error}
"
+
+
+def styled_warning(warn):
+ return f"{warn}
"
+
+
+def styled_message(message):
+ return f"{message}
"
+
+
+def has_no_nan_values(df, columns):
+ return df[columns].notna().all(axis=1)
+
+
+def has_nan_values(df, columns):
+ return df[columns].isna().any(axis=1)
diff --git a/src/display/utils.py b/src/display/utils.py
new file mode 100755
index 0000000000000000000000000000000000000000..166001f47d46eb20fda526c0ccf2346db7226588
--- /dev/null
+++ b/src/display/utils.py
@@ -0,0 +1,164 @@
+from dataclasses import dataclass, make_dataclass
+from enum import Enum
+
+import pandas as pd
+
+def fields(raw_class):
+ return [v for k, v in raw_class.__dict__.items() if k[:2] != "__" and k[-2:] != "__"]
+
+
+@dataclass
+class Task:
+ benchmark: str
+ metric: str
+ col_name: str
+
+class Tasks(Enum):
+ arc = Task("arc_challenge", "acc_norm", "ARC")
+ hellaswag = Task("hellaswag", "acc_norm", "HellaSwag")
+ mmlu = Task("mmlu", "acc", "MMLU")
+ truthfulqa = Task("truthfulqa_mc", "mc2", "TruthfulQA")
+ # winogrande = Task("winogrande", "acc_norm", "Winogrande")
+ # gsm8k = Task("gsm8k", "acc_norm", "GSM8k")
+ commongen_v2 = Task("commongen_v2", "acc_norm", "CommonGen V2")
+ # eqBench = Task("eq_bench", "acc_norm", "EQ Bench")
+ # instFollow = Task("inst_follow", "acc_norm", "InstFollow")
+ # harmlessness = Task("harmlessness", "acc_norm", "Harmlessness")
+ # helpfulness = Task("helpfulness", "acc_norm", "Helpfulness")
+
+class Ranks(Enum):
+ daily = Task("daily", "daily", "Daily Rank")
+ quarterly = Task("quarterly", "quarterly", "Quarterly Rank")
+
+
+# These classes are for user facing column names,
+# to avoid having to change them all around the code
+# when a modif is needed
+@dataclass
+class ColumnContent:
+ name: str
+ type: str
+ displayed_by_default: bool
+ hidden: bool = False
+ never_hidden: bool = False
+ dummy: bool = False
+
+auto_eval_column_dict = []
+# Init
+auto_eval_column_dict.append(["model_type_symbol", ColumnContent, ColumnContent("T", "str", True, never_hidden=True)])
+auto_eval_column_dict.append(["model", ColumnContent, ColumnContent("Model", "markdown", True, never_hidden=True)])
+# Ranks
+auto_eval_column_dict.append(["daily", ColumnContent, ColumnContent("Daily Rank", "number", True)])
+auto_eval_column_dict.append(["quarterly", ColumnContent, ColumnContent("Quarterly Rank", "number", True)])
+# Scores
+auto_eval_column_dict.append(["average", ColumnContent, ColumnContent("Average ⬆️", "number", True)])
+for task in Tasks:
+ auto_eval_column_dict.append([task.name, ColumnContent, ColumnContent(task.value.col_name, "number", True)])
+# Model information
+auto_eval_column_dict.append(["model_type", ColumnContent, ColumnContent("Type", "str", False)])
+auto_eval_column_dict.append(["architecture", ColumnContent, ColumnContent("Architecture", "str", False)])
+auto_eval_column_dict.append(["weight_type", ColumnContent, ColumnContent("Weight type", "str", False, True)])
+auto_eval_column_dict.append(["precision", ColumnContent, ColumnContent("Precision", "str", False)])
+auto_eval_column_dict.append(["merged", ColumnContent, ColumnContent("Merged", "bool", False)])
+auto_eval_column_dict.append(["license", ColumnContent, ColumnContent("Hub License", "str", False)])
+auto_eval_column_dict.append(["params", ColumnContent, ColumnContent("#Params (B)", "number", False)])
+auto_eval_column_dict.append(["likes", ColumnContent, ColumnContent("Hub ❤️", "number", False)])
+auto_eval_column_dict.append(["still_on_hub", ColumnContent, ColumnContent("Available on the hub", "bool", False)])
+auto_eval_column_dict.append(["revision", ColumnContent, ColumnContent("Model sha", "str", False, False)])
+auto_eval_column_dict.append(["flagged", ColumnContent, ColumnContent("Flagged", "bool", False, False)])
+# Dummy column for the search bar (hidden by the custom CSS)
+auto_eval_column_dict.append(["dummy", ColumnContent, ColumnContent("model_name_for_query", "str", False, dummy=True)])
+
+# We use make dataclass to dynamically fill the scores from Tasks
+AutoEvalColumn = make_dataclass("AutoEvalColumn", auto_eval_column_dict, frozen=True)
+
+
+@dataclass(frozen=True)
+class EvalQueueColumn: # Queue column
+ model = ColumnContent("model", "markdown", True)
+ revision = ColumnContent("revision", "str", True)
+ private = ColumnContent("private", "bool", True)
+ precision = ColumnContent("precision", "str", True)
+ weight_type = ColumnContent("weight_type", "str", "Original")
+ status = ColumnContent("status", "str", True)
+
+# Define the human baselines
+human_baseline_row = {
+ AutoEvalColumn.model.name: "Human performance
",
+}
+
+@dataclass
+class ModelDetails:
+ name: str
+ symbol: str = "" # emoji, only for the model type
+
+
+class ModelType(Enum):
+ PT = ModelDetails(name="pretrained", symbol="🟢")
+ # FT = ModelDetails(name="fine-tuned", symbol="🔶")
+ IFT = ModelDetails(name="instruction-tuned", symbol="⭕")
+ RL = ModelDetails(name="RL-tuned", symbol="🟦")
+ Unknown = ModelDetails(name="", symbol="?")
+
+ def to_str(self, separator=" "):
+ return f"{self.value.symbol}{separator}{self.value.name}"
+
+ @staticmethod
+ def from_str(type):
+ # if "fine-tuned" in type or "🔶" in type:
+ # return ModelType.FT
+ if "pretrained" in type or "🟢" in type:
+ return ModelType.PT
+ if "RL-tuned" in type or "🟦" in type:
+ return ModelType.RL
+ if "instruction-tuned" in type or "⭕" in type:
+ return ModelType.IFT
+ return ModelType.Unknown
+
+class WeightType(Enum):
+ Adapter = ModelDetails("Adapter")
+ Original = ModelDetails("Original")
+ Delta = ModelDetails("Delta")
+
+class Precision(Enum):
+ float16 = ModelDetails("float16")
+ # bfloat16 = ModelDetails("bfloat16")
+ # qt_8bit = ModelDetails("8bit")
+ # qt_4bit = ModelDetails("4bit")
+ # qt_GPTQ = ModelDetails("GPTQ")
+ Unknown = ModelDetails("?")
+
+ def from_str(precision):
+ if precision in ["torch.float16", "float16"]:
+ return Precision.float16
+ if precision in ["8bit"]:
+ return Precision.qt_8bit
+ if precision in ["4bit"]:
+ return Precision.qt_4bit
+ if precision in ["GPTQ", "None"]:
+ return Precision.qt_GPTQ
+ return Precision.Unknown
+
+
+
+
+# Column selection
+COLS = [c.name for c in fields(AutoEvalColumn) if not c.hidden]
+TYPES = [c.type for c in fields(AutoEvalColumn) if not c.hidden]
+COLS_LITE = [c.name for c in fields(AutoEvalColumn) if c.displayed_by_default and not c.hidden]
+TYPES_LITE = [c.type for c in fields(AutoEvalColumn) if c.displayed_by_default and not c.hidden]
+
+EVAL_COLS = [c.name for c in fields(EvalQueueColumn)]
+EVAL_TYPES = [c.type for c in fields(EvalQueueColumn)]
+
+BENCHMARK_COLS = [t.value.col_name for t in Tasks]
+
+NUMERIC_INTERVALS = {
+ "Unknown": pd.Interval(-1, 0, closed="right"),
+ "0~3B": pd.Interval(0, 3, closed="right"),
+ "3~7B": pd.Interval(3, 7.3, closed="right"),
+ "7~13B": pd.Interval(7.3, 13, closed="right"),
+ "13~35B": pd.Interval(13, 35, closed="right"),
+ "35~60B": pd.Interval(35, 60, closed="right"),
+ "60B+": pd.Interval(60, 10000, closed="right"),
+}
diff --git a/src/envs.py b/src/envs.py
new file mode 100755
index 0000000000000000000000000000000000000000..a3f4f73977ab2e88b4c99c4aaeebb73aa9f210f4
--- /dev/null
+++ b/src/envs.py
@@ -0,0 +1,32 @@
+import os
+
+from huggingface_hub import HfApi
+
+# clone / pull the lmeh eval data
+H4_TOKEN = os.environ.get("H4_TOKEN", None)
+
+REPO_ID = "junkim100/self-improving-leaderboard"
+QUEUE_REPO = "junkim100/SIL_requests"
+RESULTS_REPO = "junkim100/SIL_results"
+
+PRIVATE_QUEUE_REPO = "open-ko-llm-leaderboard/private-requests"
+PRIVATE_RESULTS_REPO = "open-ko-llm-leaderboard/private-results"
+
+IS_PUBLIC = bool(os.environ.get("IS_PUBLIC", True))
+
+CACHE_PATH=os.getenv("HF_HOME", ".")
+
+EVAL_REQUESTS_PATH = os.path.join(CACHE_PATH, "eval-queue")
+EVAL_RESULTS_PATH = os.path.join(CACHE_PATH, "eval-results")
+
+EVAL_REQUESTS_PATH_PRIVATE = "eval-queue-private"
+EVAL_RESULTS_PATH_PRIVATE = "eval-results-private"
+
+PATH_TO_COLLECTION = "open-ko-llm-leaderboard/ko-llm-leaderboard-best-models-659c7e45a481ceea4c883506"
+
+# Rate limit variables
+RATE_LIMIT_PERIOD = 7
+RATE_LIMIT_QUOTA = 5
+HAS_HIGHER_RATE_LIMIT = []
+
+API = HfApi(token=H4_TOKEN)
diff --git a/src/leaderboard/__pycache__/filter_models.cpython-310.pyc b/src/leaderboard/__pycache__/filter_models.cpython-310.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..7500ee03c42db8f02506351ae84b64e5e93508b0
Binary files /dev/null and b/src/leaderboard/__pycache__/filter_models.cpython-310.pyc differ
diff --git a/src/leaderboard/__pycache__/read_evals.cpython-310.pyc b/src/leaderboard/__pycache__/read_evals.cpython-310.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..ca2d5dba4b960f6b5a8718f33138bc6f99aa980a
Binary files /dev/null and b/src/leaderboard/__pycache__/read_evals.cpython-310.pyc differ
diff --git a/src/leaderboard/filter_models.py b/src/leaderboard/filter_models.py
new file mode 100755
index 0000000000000000000000000000000000000000..95ea9e5214ef507a41e512052ec4a27dae7dc9a4
--- /dev/null
+++ b/src/leaderboard/filter_models.py
@@ -0,0 +1,51 @@
+from src.display.formatting import model_hyperlink
+from src.display.utils import AutoEvalColumn
+
+# Models which have been flagged by users as being problematic for a reason or another
+# (Model name to forum discussion link)
+FLAGGED_MODELS = {
+ "merged": "https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/510",
+ "TeamUNIVA/Komodo_7B_v0.1.0": "https://huggingface.co/spaces/upstage/open-ko-llm-leaderboard/discussions/44",
+}
+
+# Models which have been requested by orgs to not be submitted on the leaderboard
+DO_NOT_SUBMIT_MODELS = [
+]
+
+
+def flag_models(leaderboard_data: list[dict]):
+ for model_data in leaderboard_data:
+ # Merges are flagged automatically
+ if model_data[AutoEvalColumn.flagged.name] == True:
+ flag_key = "merged"
+ else:
+ flag_key = model_data["model_name_for_query"]
+
+ if flag_key in FLAGGED_MODELS:
+ issue_num = FLAGGED_MODELS[flag_key].split("/")[-1]
+ issue_link = model_hyperlink(
+ FLAGGED_MODELS[flag_key],
+ f"See discussion #{issue_num}",
+ )
+ model_data[
+ AutoEvalColumn.model.name
+ ] = f"{model_data[AutoEvalColumn.model.name]} has been flagged! {issue_link}"
+ model_data[AutoEvalColumn.flagged.name] = True
+ else:
+ model_data[AutoEvalColumn.flagged.name] = False
+
+
+def remove_forbidden_models(leaderboard_data: list[dict]):
+ indices_to_remove = []
+ for ix, model in enumerate(leaderboard_data):
+ if model["model_name_for_query"] in DO_NOT_SUBMIT_MODELS:
+ indices_to_remove.append(ix)
+
+ for ix in reversed(indices_to_remove):
+ leaderboard_data.pop(ix)
+ return leaderboard_data
+
+
+def filter_models(leaderboard_data: list[dict]):
+ leaderboard_data = remove_forbidden_models(leaderboard_data)
+ flag_models(leaderboard_data)
diff --git a/src/leaderboard/read_evals.py b/src/leaderboard/read_evals.py
new file mode 100755
index 0000000000000000000000000000000000000000..bdba3a3d90c40cfc647d55becadd6ebc568c3aad
--- /dev/null
+++ b/src/leaderboard/read_evals.py
@@ -0,0 +1,272 @@
+import glob
+import json
+import math
+import os
+from dataclasses import dataclass
+
+import dateutil
+import numpy as np
+
+from huggingface_hub import ModelCard
+
+from src.display.formatting import make_clickable_model
+from src.display.utils import AutoEvalColumn, ModelType, Tasks, Ranks, Precision, WeightType
+from src.submission.check_validity import is_model_on_hub, check_model_card
+
+
+@dataclass
+class EvalResult:
+ # Also see src.display.utils.AutoEvalColumn for what will be displayed.
+ eval_name: str # org_model_precision (uid)
+ full_model: str # org/model (path on hub)
+ org: str
+ model: str
+ revision: str # commit hash, "" if main
+ results: dict
+ precision: Precision = Precision.Unknown
+ model_type: ModelType = ModelType.Unknown # Pretrained, fine tuned, ...
+ weight_type: WeightType = WeightType.Original # Original or Adapter
+ architecture: str = "Unknown" # From config file
+ license: str = "?"
+ likes: int = 0
+ num_params: int = 0
+ date: str = "" # submission date of request file
+ still_on_hub: bool = False
+ is_merge: bool = False
+ flagged: bool = False
+
+ @classmethod
+ def init_from_json_file(self, json_filepath):
+ """Inits the result from the specific model result file"""
+ with open(json_filepath) as fp:
+ data = json.load(fp)
+
+ # We manage the legacy config format
+ config = data.get("config", data.get("config_general", None))
+
+ # Precision
+ precision = Precision.from_str(config.get("model_dtype"))
+
+ # Get model and org
+ org_and_model = config.get("model_name", config.get("model_args", None))
+ org_and_model = org_and_model.split("/", 1)
+
+ if len(org_and_model) == 1:
+ org = None
+ model = org_and_model[0]
+ result_key = f"{model}_{precision.value.name}"
+ else:
+ org = org_and_model[0]
+ model = org_and_model[1]
+ result_key = f"{org}_{model}_{precision.value.name}"
+ full_model = "/".join(org_and_model)
+
+ still_on_hub, error, model_config = is_model_on_hub(
+ full_model, config.get("model_sha", "main"), trust_remote_code=True, test_tokenizer=False
+ )
+ architecture = "?"
+ if model_config is not None:
+ architectures = getattr(model_config, "architectures", None)
+ if architectures:
+ architecture = ";".join(architectures)
+
+ # If the model doesn't have a model card or a license, we consider it's deleted
+ if still_on_hub:
+ try:
+ if check_model_card(full_model)[0] is False:
+ still_on_hub = False
+ except Exception:
+ still_on_hub = False
+
+ # Check if the model is a merge
+ is_merge_from_metadata = False
+ flagged = False
+ if still_on_hub:
+ model_card = ModelCard.load(full_model)
+
+ if model_card.data.tags:
+ is_merge_from_metadata = "merge" in model_card.data.tags
+ merge_keywords = ["mergekit", "merged model", "merge model", "merging", "Carbon"]
+ # If the model is a merge but not saying it in the metadata, we flag it
+ is_merge_from_model_card = any(keyword in model_card.text.lower() for keyword in merge_keywords)
+ flagged = is_merge_from_model_card and not is_merge_from_metadata
+
+
+ # Extract results available in this file (some results are split in several files)
+ results = {}
+ for rank in Ranks:
+ rank = rank.value
+ if rank.benchmark in data["results"]:
+ results[rank.benchmark] = data["results"][rank.benchmark][rank.metric]
+ for task in Tasks:
+ task = task.value
+
+ # Some truthfulQA values are NaNs
+ if task.benchmark == "truthfulqa:mc" and "harness|truthfulqa:mc|0" in data["results"]:
+ if math.isnan(float(data["results"]["harness|truthfulqa:mc|0"][task.metric])):
+ results[task.benchmark] = 0.0
+ continue
+
+ # New tasks have been added, we need to skip them if not exists
+ if task.benchmark in ["winogrande", "gsm8k", "eq_bench", "inst_follow", "harmlessness", "helpfulness"]:
+ accs = np.array([v.get(task.metric, None) for k, v in data["results"].items() if task.benchmark in k])
+ if accs.size == 0 or any([acc is None for acc in accs]):
+ results[task.benchmark] = 0.0
+ continue
+
+ # We average all scores of a given metric (mostly for mmlu)
+ accs = np.array([v.get(task.metric, None) for k, v in data["results"].items() if task.benchmark in k])
+ if accs.size == 0 or any([acc is None for acc in accs]):
+ continue
+
+ mean_acc = np.mean(accs) * 100.0
+ results[task.benchmark] = mean_acc
+
+ return self(
+ eval_name=result_key,
+ full_model=full_model,
+ org=org,
+ model=model,
+ results=results,
+ precision=precision,
+ revision= config.get("model_sha", ""),
+ still_on_hub=still_on_hub,
+ architecture=architecture,
+ is_merge=is_merge_from_metadata,
+ flagged=flagged,
+ )
+
+ def update_with_request_file(self, requests_path):
+ """Finds the relevant request file for the current model and updates info with it"""
+ request_file = get_request_file_for_model(requests_path, self.full_model, self.precision.value.name)
+
+ try:
+ with open(request_file, "r") as f:
+ request = json.load(f)
+ self.model_type = ModelType.from_str(request.get("model_type", ""))
+ self.weight_type = WeightType[request.get("weight_type", "Original")]
+ self.license = request.get("license", "?")
+ self.likes = request.get("likes", 0)
+ self.num_params = request.get("params", 0)
+ self.date = request.get("submitted_time", "")
+ except Exception:
+ print(f"Could not find request file for {self.org}/{self.model}")
+
+ def to_dict(self):
+ """Converts the Eval Result to a dict compatible with our dataframe display"""
+
+ # Skip the new tasks for now
+ # TODO: safely remove this code when the task results are all added
+ skip_avg_len = 0
+ # if self.results['winogrande'] == 0.0:
+ # skip_avg_len += 1
+ # if self.results['gsm8k'] == 0.0:
+ # skip_avg_len += 1
+ # if self.results['eq_bench'] == 0.0:
+ # skip_avg_len += 1
+ # if self.results['inst_follow'] == 0.0:
+ # skip_avg_len += 1
+ # if self.results['harmlessness'] == 0.0:
+ # skip_avg_len += 1
+ # if self.results['helpfulness'] == 0.0:
+ # skip_avg_len += 1
+
+ average = sum([v for v in self.results.values() if v is not None]) / (len(Tasks) - skip_avg_len)
+
+ data_dict = {
+ "eval_name": self.eval_name, # not a column, just a save name,
+ AutoEvalColumn.precision.name: self.precision.value.name,
+ AutoEvalColumn.model_type.name: self.model_type.value.name,
+ AutoEvalColumn.merged.name: self.is_merge,
+ AutoEvalColumn.model_type_symbol.name: self.model_type.value.symbol, # + "🥦" if self.is_merge,
+ AutoEvalColumn.weight_type.name: self.weight_type.value.name,
+ AutoEvalColumn.architecture.name: self.architecture,
+ AutoEvalColumn.model.name: make_clickable_model(self.full_model),
+ AutoEvalColumn.dummy.name: self.full_model,
+ AutoEvalColumn.revision.name: self.revision,
+ AutoEvalColumn.average.name: average,
+ AutoEvalColumn.license.name: self.license,
+ AutoEvalColumn.likes.name: self.likes,
+ AutoEvalColumn.params.name: self.num_params,
+ AutoEvalColumn.still_on_hub.name: self.still_on_hub,
+ AutoEvalColumn.flagged.name: self.flagged
+ }
+
+ AllColumns = []
+ for task in Tasks:
+ AllColumns.append(task.value)
+ for rank in Ranks:
+ AllColumns.append(rank.value)
+
+ for a in AllColumns:
+ if a.benchmark in ["daily", "quarterly"]:
+ data_dict[a.col_name] = self.results[a.benchmark]
+ print(a.benchmark, self.results[a.benchmark], a.col_name)
+ else:
+ data_dict[a.col_name] = self.results[a.benchmark]
+
+ return data_dict
+
+
+def get_request_file_for_model(requests_path, model_name, precision):
+ """Selects the correct request file for a given model. Only keeps runs tagged as FINISHED"""
+ request_files = os.path.join(
+ requests_path,
+ f"{model_name}_eval_request_*.json",
+ )
+ request_files = glob.glob(request_files)
+
+ # Select correct request file (precision)
+ request_file = ""
+ request_files = sorted(request_files, reverse=True)
+ for tmp_request_file in request_files:
+ with open(tmp_request_file, "r") as f:
+ req_content = json.load(f)
+ if (
+ req_content["status"] in ["FINISHED"]
+ and req_content["precision"] == precision.split(".")[-1]
+ ):
+ request_file = tmp_request_file
+ return request_file
+
+
+def get_raw_eval_results(results_path: str, requests_path: str) -> list[EvalResult]:
+ """From the path of the results folder root, extract all needed info for results"""
+ model_result_filepaths = []
+
+ for root, _, files in os.walk(results_path):
+ # We should only have json files in model results
+ if len(files) == 0 or any([not f.endswith(".json") for f in files]):
+ continue
+
+ # Sort the files by date
+ try:
+ files.sort(key=lambda x: x.removesuffix(".json").removeprefix("results_")[:-7])
+ except dateutil.parser._parser.ParserError:
+ files = [files[-1]]
+
+ for file in files:
+ model_result_filepaths.append(os.path.join(root, file))
+
+ eval_results = {}
+ for model_result_filepath in model_result_filepaths:
+ # Creation of result
+ eval_result = EvalResult.init_from_json_file(model_result_filepath)
+ eval_result.update_with_request_file(requests_path)
+
+ # Store results of same eval together
+ eval_name = eval_result.eval_name
+ if eval_name in eval_results.keys():
+ eval_results[eval_name].results.update({k: v for k, v in eval_result.results.items() if v is not None})
+ else:
+ eval_results[eval_name] = eval_result
+
+ results = []
+ for v in eval_results.values():
+ try:
+ v.to_dict() # we test if the dict version is complete
+ results.append(v)
+ except KeyError: # not all eval values present
+ continue
+
+ return results
diff --git a/src/populate.py b/src/populate.py
new file mode 100755
index 0000000000000000000000000000000000000000..df32a6311e2305ecb14eef8cf2feda655a261bab
--- /dev/null
+++ b/src/populate.py
@@ -0,0 +1,70 @@
+import json
+import os
+
+import pandas as pd
+
+from src.display.formatting import has_no_nan_values, make_clickable_model
+from src.display.utils import AutoEvalColumn, EvalQueueColumn
+from src.leaderboard.filter_models import filter_models
+from src.leaderboard.read_evals import get_raw_eval_results
+
+
+def get_leaderboard_df(results_path: str, requests_path: str, cols: list, benchmark_cols: list) -> pd.DataFrame:
+ raw_data = get_raw_eval_results(results_path, requests_path)
+ all_data_json = [v.to_dict() for v in raw_data]
+ # all_data_json.append(baseline_row)
+ filter_models(all_data_json)
+
+ df = pd.DataFrame.from_records(all_data_json)
+ # df = df.sort_values(by=[AutoEvalColumn.average.name], ascending=False)
+ df = df.sort_values(by=["Daily Rank"], ascending=True)
+
+ # print(df[AutoEvalColumn.average.name])
+ try:
+ df = df[cols].round(decimals=2)
+ except:
+ pass
+
+ # filter out if any of the benchmarks have not been produced
+ try:
+ df = df[has_no_nan_values(df, benchmark_cols)]
+ except:
+ pass
+ return raw_data, df
+
+
+def get_evaluation_queue_df(save_path: str, cols: list) -> list[pd.DataFrame]:
+ entries = [entry for entry in os.listdir(save_path) if not entry.startswith(".")]
+ all_evals = []
+
+ for entry in entries:
+ if ".json" in entry:
+ file_path = os.path.join(save_path, entry)
+ with open(file_path) as fp:
+ data = json.load(fp)
+
+ data[EvalQueueColumn.model.name] = make_clickable_model(data["model"])
+ data[EvalQueueColumn.revision.name] = data.get("revision", "main")
+
+ all_evals.append(data)
+ elif ".md" not in entry:
+ # this is a folder
+ sub_entries = [e for e in os.listdir(f"{save_path}/{entry}") if not e.startswith(".")]
+ for sub_entry in sub_entries:
+ file_path = os.path.join(save_path, entry, sub_entry)
+ with open(file_path) as fp:
+ data = json.load(fp)
+
+ data[EvalQueueColumn.model.name] = make_clickable_model(data["model"])
+ data[EvalQueueColumn.revision.name] = data.get("revision", "main")
+ all_evals.append(data)
+
+ pending_list = [e for e in all_evals if e["status"] in ["PENDING", "RERUN"]]
+ running_list = [e for e in all_evals if e["status"] == "RUNNING"]
+ finished_list = [e for e in all_evals if e["status"].startswith("FINISHED") or e["status"] == "PENDING_NEW_EVAL"]
+ failed_list = [e for e in all_evals if e["status"] == "FAILED"]
+ df_pending = pd.DataFrame.from_records(pending_list, columns=cols)
+ df_running = pd.DataFrame.from_records(running_list, columns=cols)
+ df_finished = pd.DataFrame.from_records(finished_list, columns=cols)
+ df_failed = pd.DataFrame.from_records(failed_list, columns=cols)
+ return df_finished[cols], df_running[cols], df_pending[cols], df_failed[cols]
diff --git a/src/submission/__pycache__/check_validity.cpython-310.pyc b/src/submission/__pycache__/check_validity.cpython-310.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..414d55f4b552402359607afd13cb0ca8696bcc75
Binary files /dev/null and b/src/submission/__pycache__/check_validity.cpython-310.pyc differ
diff --git a/src/submission/__pycache__/submit.cpython-310.pyc b/src/submission/__pycache__/submit.cpython-310.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..c78822770039cb8f07a97c179a7fb6b2c258b748
Binary files /dev/null and b/src/submission/__pycache__/submit.cpython-310.pyc differ
diff --git a/src/submission/check_validity.py b/src/submission/check_validity.py
new file mode 100755
index 0000000000000000000000000000000000000000..aba20a605b10aa4418703cad0345a3c0091fe1b3
--- /dev/null
+++ b/src/submission/check_validity.py
@@ -0,0 +1,129 @@
+import json
+import os
+import re
+from collections import defaultdict
+from datetime import datetime, timedelta, timezone
+
+import huggingface_hub
+from huggingface_hub import ModelCard
+from huggingface_hub.hf_api import ModelInfo
+from transformers import AutoConfig, AutoTokenizer
+
+from src.envs import HAS_HIGHER_RATE_LIMIT
+
+
+# ht to @Wauplin, thank you for the snippet!
+# See https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard/discussions/317
+def check_model_card(repo_id: str) -> tuple[bool, str]:
+ # Returns operation status, and error message
+ try:
+ card = ModelCard.load(repo_id)
+ except huggingface_hub.utils.EntryNotFoundError:
+ return False, "Please add a model card to your model to explain how you trained/fine-tuned it."
+
+ # Enforce license metadata
+ if card.data.license is None:
+ if not ("license_name" in card.data and "license_link" in card.data):
+ return False, (
+ "License not found. Please add a license to your model card using the `license` metadata or a"
+ " `license_name`/`license_link` pair."
+ )
+
+ # Enforce card content
+ if len(card.text) < 200:
+ return False, "Please add a description to your model card, it is too short."
+
+ return True, ""
+
+
+def is_model_on_hub(model_name: str, revision: str, token: str = None, trust_remote_code=False, test_tokenizer=False) -> tuple[bool, str]:
+ try:
+ config = AutoConfig.from_pretrained(model_name, revision=revision, trust_remote_code=trust_remote_code, token=token) #, force_download=True)
+ if test_tokenizer:
+ try:
+ tk = AutoTokenizer.from_pretrained(model_name, revision=revision, trust_remote_code=trust_remote_code, token=token)
+ except ValueError as e:
+ return (
+ False,
+ f"uses a tokenizer which is not in a transformers release: {e}",
+ None
+ )
+ except Exception as e:
+ return (False, "'s tokenizer cannot be loaded. Is your tokenizer class in a stable transformers release, and correctly configured?", None)
+ return True, None, config
+
+ except ValueError:
+ return (
+ False,
+ "needs to be launched with `trust_remote_code=True`. For safety reason, we do not allow these models to be automatically submitted to the leaderboard.",
+ None
+ )
+
+ except Exception as e:
+ return False, "was not found on hub!", None
+
+
+def get_model_size(model_info: ModelInfo, precision: str):
+ size_pattern = re.compile(r"(\d+\.)?\d+(b|m)")
+ try:
+ model_size = round(model_info.safetensors["total"] / 1e9, 3)
+ except (AttributeError, TypeError ):
+ try:
+ size_match = re.search(size_pattern, model_info.modelId.split("/")[-1].lower())
+ model_size = size_match.group(0)
+ model_size = round(float(model_size[:-1]) if model_size[-1] == "b" else float(model_size[:-1]) / 1e3, 3)
+ except AttributeError:
+ return 0 # Unknown model sizes are indicated as 0, see NUMERIC_INTERVALS in app.py
+
+ size_factor = 8 if (precision == "GPTQ" or "gptq" in model_info.modelId.split("/")[-1].lower()) else 1
+ model_size = size_factor * model_size
+ return model_size
+
+def get_model_arch(model_info: ModelInfo):
+ return model_info.config.get("architectures", "Unknown")
+
+def user_submission_permission(org_or_user, users_to_submission_dates, rate_limit_period, rate_limit_quota):
+ if org_or_user not in users_to_submission_dates:
+ return True, ""
+ submission_dates = sorted(users_to_submission_dates[org_or_user])
+
+ time_limit = (datetime.now(timezone.utc) - timedelta(days=rate_limit_period)).strftime("%Y-%m-%dT%H:%M:%SZ")
+ submissions_after_timelimit = [d for d in submission_dates if d > time_limit]
+
+ num_models_submitted_in_period = len(submissions_after_timelimit)
+ if org_or_user in HAS_HIGHER_RATE_LIMIT:
+ rate_limit_quota = 2 * rate_limit_quota
+
+ if num_models_submitted_in_period > rate_limit_quota:
+ error_msg = f"Organisation or user `{org_or_user}`"
+ error_msg += f"already has {num_models_submitted_in_period} model requests submitted to the leaderboard "
+ error_msg += f"in the last {rate_limit_period} days.\n"
+ error_msg += (
+ "Please wait a couple of days before resubmitting, so that everybody can enjoy using the leaderboard 🤗"
+ )
+ return False, error_msg
+ return True, ""
+
+
+def already_submitted_models(requested_models_dir: str) -> set[str]:
+ depth = 1
+ file_names = []
+ users_to_submission_dates = defaultdict(list)
+
+ for root, _, files in os.walk(requested_models_dir):
+ current_depth = root.count(os.sep) - requested_models_dir.count(os.sep)
+ if current_depth == depth:
+ for file in files:
+ if not file.endswith(".json"):
+ continue
+ with open(os.path.join(root, file), "r") as f:
+ info = json.load(f)
+ file_names.append(f"{info['model']}_{info['revision']}_{info['precision']}")
+
+ # Select organisation
+ if info["model"].count("/") == 0 or "submitted_time" not in info:
+ continue
+ organisation, _ = info["model"].split("/")
+ users_to_submission_dates[organisation].append(info["submitted_time"])
+
+ return set(file_names), users_to_submission_dates
diff --git a/src/submission/submit.py b/src/submission/submit.py
new file mode 100755
index 0000000000000000000000000000000000000000..5ce81c0df846365e4580fe7f59261cfe8002d9ea
--- /dev/null
+++ b/src/submission/submit.py
@@ -0,0 +1,143 @@
+import json
+import os
+from datetime import datetime, timezone
+
+from src.display.formatting import styled_error, styled_message, styled_warning
+from src.envs import API, EVAL_REQUESTS_PATH, H4_TOKEN, QUEUE_REPO, RATE_LIMIT_PERIOD, RATE_LIMIT_QUOTA
+from src.leaderboard.filter_models import DO_NOT_SUBMIT_MODELS
+from src.submission.check_validity import (
+ already_submitted_models,
+ check_model_card,
+ get_model_size,
+ is_model_on_hub,
+ user_submission_permission,
+)
+
+REQUESTED_MODELS = None
+USERS_TO_SUBMISSION_DATES = None
+
+def add_new_eval(
+ model: str,
+ base_model: str,
+ revision: str,
+ precision: str,
+ private: bool,
+ weight_type: str,
+ model_type: str,
+):
+ global REQUESTED_MODELS
+ global USERS_TO_SUBMISSION_DATES
+ if not REQUESTED_MODELS:
+ REQUESTED_MODELS, USERS_TO_SUBMISSION_DATES = already_submitted_models(EVAL_REQUESTS_PATH)
+
+ user_name = ""
+ model_path = model
+ if "/" in model:
+ user_name = model.split("/")[0]
+ model_path = model.split("/")[1]
+
+ precision = precision.split(" ")[0]
+ current_time = datetime.now(timezone.utc).strftime("%Y-%m-%dT%H:%M:%SZ")
+
+ if model_type is None or model_type == "":
+ return styled_error("Please select a model type.")
+
+ # Upstage models are now allowed to be submitted to ensure the transparency and fairness of the leaderboard.
+ if user_name == "upstage":
+ return styled_warning("We do not conduct evaluations on Upstage models to ensure the transparency and fairness of the leaderboard. Please take this into consideration.")
+
+ # Is the user rate limited?
+ if user_name != "":
+ user_can_submit, error_msg = user_submission_permission(
+ user_name, USERS_TO_SUBMISSION_DATES, RATE_LIMIT_PERIOD, RATE_LIMIT_QUOTA
+ )
+ if not user_can_submit:
+ return styled_error(error_msg)
+
+ # Did the model authors forbid its submission to the leaderboard?
+ if model in DO_NOT_SUBMIT_MODELS or base_model in DO_NOT_SUBMIT_MODELS:
+ return styled_warning("Model authors have requested that their model be not submitted on the leaderboard.")
+
+ # Does the model actually exist?
+ if revision == "":
+ revision = "main"
+
+ # Is the model on the hub?
+ if weight_type in ["Delta", "Adapter"]:
+ base_model_on_hub, error, _ = is_model_on_hub(model_name=base_model, revision=revision, token=H4_TOKEN, test_tokenizer=True)
+ if not base_model_on_hub:
+ return styled_error(f'Base model "{base_model}" {error}')
+
+ if not weight_type == "Adapter":
+ model_on_hub, error, _ = is_model_on_hub(model_name=model, revision=revision, test_tokenizer=True, trust_remote_code=True if model == "KT-AI/midm-bitext-S-7B-inst-v2" else False)
+ if not model_on_hub:
+ return styled_error(f'Model "{model}" {error}')
+
+ # Is the model info correctly filled?
+ try:
+ model_info = API.model_info(repo_id=model, revision=revision)
+ except Exception:
+ return styled_error("Could not get your model information. Please fill it up properly.")
+
+ model_size = get_model_size(model_info=model_info, precision=precision)
+
+ # Were the model size less than 30B?
+ if float(model_size) >= 30:
+ return styled_error("Please submit a model that is less than 30B")
+
+ # Were the model card and license filled?
+ try:
+ license = model_info.cardData["license"]
+ except Exception:
+ return styled_error("Please select a license for your model")
+
+ modelcard_OK, error_msg = check_model_card(model)
+ if not modelcard_OK:
+ return styled_error(error_msg)
+
+ # Seems good, creating the eval
+ print("Adding new eval")
+
+ eval_entry = {
+ "model": model,
+ "base_model": base_model,
+ "revision": revision,
+ "private": private,
+ "precision": precision,
+ "weight_type": weight_type,
+ "status": "PENDING",
+ "submitted_time": current_time,
+ "model_type": model_type,
+ "likes": model_info.likes,
+ "params": model_size,
+ "license": license,
+ }
+
+ # Check for duplicate submission
+ if f"{model}_{revision}_{precision}" in REQUESTED_MODELS:
+ return styled_warning("This model has been already submitted.")
+
+ print("Creating eval file")
+ OUT_DIR = f"{EVAL_REQUESTS_PATH}/{user_name}"
+ os.makedirs(OUT_DIR, exist_ok=True)
+ out_path = f"{OUT_DIR}/{model_path}_eval_request_{private}_{precision}_{weight_type}.json"
+
+ with open(out_path, "w") as f:
+ f.write(json.dumps(eval_entry))
+
+ print("Uploading eval file")
+ API.upload_file(
+ path_or_fileobj=out_path,
+ path_in_repo=out_path.split("eval-queue/")[1],
+ repo_id=QUEUE_REPO,
+ repo_type="dataset",
+ commit_message=f"Update status to PENDING.",
+ )
+
+ # Remove the local file
+ os.remove(out_path)
+
+ return styled_message(
+ "Your request has been submitted to the evaluation queue!\nPlease wait for up to an hour for the model to show in the PENDING list."
+ )
+
diff --git a/src/tools/__pycache__/collections.cpython-310.pyc b/src/tools/__pycache__/collections.cpython-310.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..b4f21bd55aed9baed92a9ea8ff889d99f9cbf051
Binary files /dev/null and b/src/tools/__pycache__/collections.cpython-310.pyc differ
diff --git a/src/tools/__pycache__/plots.cpython-310.pyc b/src/tools/__pycache__/plots.cpython-310.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..e3b8ae0e3feef7362ecf64dadaac3e5f923c5e44
Binary files /dev/null and b/src/tools/__pycache__/plots.cpython-310.pyc differ
diff --git a/src/tools/collections.py b/src/tools/collections.py
new file mode 100755
index 0000000000000000000000000000000000000000..cca2e0afa2ca9d5cd2f6820f1aed841c78f6bb9a
--- /dev/null
+++ b/src/tools/collections.py
@@ -0,0 +1,82 @@
+import os
+
+import pandas as pd
+from huggingface_hub import add_collection_item, delete_collection_item, get_collection, update_collection_item
+from huggingface_hub.utils._errors import HfHubHTTPError
+from pandas import DataFrame
+
+from src.display.utils import AutoEvalColumn, ModelType
+from src.envs import H4_TOKEN, PATH_TO_COLLECTION
+
+# Specific intervals for the collections
+intervals = {
+ "0~3B": pd.Interval(0, 3, closed="right"),
+ "3~7B": pd.Interval(3, 7.3, closed="right"),
+ "7~13B": pd.Interval(7.3, 13, closed="right"),
+ "13~35B": pd.Interval(13, 35, closed="right"),
+ "35~60B": pd.Interval(35, 60, closed="right"),
+ "60B+": pd.Interval(60, 10000, closed="right"),
+}
+
+def update_collections(df: DataFrame):
+ """This function updates the Open Ko LLM Leaderboard model collection with the latest best models for
+ each size category and type.
+ """
+ collection = get_collection(collection_slug=PATH_TO_COLLECTION, token=H4_TOKEN)
+ params_column = pd.to_numeric(df[AutoEvalColumn.params.name], errors="coerce")
+
+ cur_best_models = []
+
+ ix = 0
+ for type in ModelType:
+ if type.value.name == "":
+ continue
+ for size in intervals:
+ # We filter the df to gather the relevant models
+ type_emoji = [t[0] for t in type.value.symbol]
+ filtered_df = df[df[AutoEvalColumn.model_type_symbol.name].isin(type_emoji)]
+
+ numeric_interval = pd.IntervalIndex([intervals[size]])
+ mask = params_column.apply(lambda x: any(numeric_interval.contains(x)))
+ filtered_df = filtered_df.loc[mask]
+
+ best_models = list(
+ filtered_df.sort_values(AutoEvalColumn.average.name, ascending=False)[AutoEvalColumn.dummy.name]
+ )
+ print(type.value.symbol, size, best_models[:10])
+
+ # We add them one by one to the leaderboard
+ for model in best_models:
+ ix += 1
+ cur_len_collection = len(collection.items)
+ try:
+ collection = add_collection_item(
+ PATH_TO_COLLECTION,
+ item_id=model,
+ item_type="model",
+ exists_ok=True,
+ note=f"Best {type.to_str(' ')} model of size {size} on the leaderboard today!",
+ token=H4_TOKEN,
+ )
+ if (
+ len(collection.items) > cur_len_collection
+ ): # we added an item - we make sure its position is correct
+ item_object_id = collection.items[-1].item_object_id
+ update_collection_item(
+ collection_slug=PATH_TO_COLLECTION, item_object_id=item_object_id, position=ix
+ )
+ cur_len_collection = len(collection.items)
+ cur_best_models.append(model)
+ break
+ except HfHubHTTPError:
+ continue
+
+ collection = get_collection(PATH_TO_COLLECTION, token=H4_TOKEN)
+ for item in collection.items:
+ if item.item_id not in cur_best_models:
+ try:
+ delete_collection_item(
+ collection_slug=PATH_TO_COLLECTION, item_object_id=item.item_object_id, token=H4_TOKEN
+ )
+ except HfHubHTTPError:
+ continue
diff --git a/src/tools/model_backlinks.py b/src/tools/model_backlinks.py
new file mode 100755
index 0000000000000000000000000000000000000000..497422aed078b65099708880e788f6f442c41c4f
--- /dev/null
+++ b/src/tools/model_backlinks.py
@@ -0,0 +1,3 @@
+models = [
+ "baseline",
+]
diff --git a/src/tools/plots.py b/src/tools/plots.py
new file mode 100755
index 0000000000000000000000000000000000000000..57a55c6bf60a91c2193803adcfd1f11cc87ac555
--- /dev/null
+++ b/src/tools/plots.py
@@ -0,0 +1,168 @@
+import pandas as pd
+import numpy as np
+import plotly.express as px
+from plotly.graph_objs import Figure
+
+from src.leaderboard.filter_models import FLAGGED_MODELS
+from src.display.utils import human_baseline_row as HUMAN_BASELINE, AutoEvalColumn, Tasks, Task, BENCHMARK_COLS
+from src.leaderboard.read_evals import EvalResult
+
+
+
+def create_scores_df(raw_data: list[EvalResult]) -> pd.DataFrame:
+ """
+ Generates a DataFrame containing the maximum scores until each date.
+
+ :param results_df: A DataFrame containing result information including metric scores and dates.
+ :return: A new DataFrame containing the maximum scores until each date for every metric.
+ """
+ # Step 1: Ensure 'date' is in datetime format and sort the DataFrame by it
+ results_df = pd.DataFrame(raw_data)
+ #results_df["date"] = pd.to_datetime(results_df["date"], format="mixed", utc=True)
+ results_df.sort_values(by="date", inplace=True)
+
+ # Step 2: Initialize the scores dictionary
+ scores = {k: [] for k in BENCHMARK_COLS + [AutoEvalColumn.average.name]}
+
+ # Step 3: Iterate over the rows of the DataFrame and update the scores dictionary
+ for task in [t.value for t in Tasks] + [Task("Average", "avg", AutoEvalColumn.average.name)]:
+ current_max = 0
+ last_date = ""
+ column = task.col_name
+ for _, row in results_df.iterrows():
+ current_model = row["full_model"]
+ if current_model in FLAGGED_MODELS:
+ continue
+
+ current_date = row["date"]
+ if task.benchmark == "Average":
+ avg_skip_len = 0
+ if row["results"]["winogrande"] == 0.0:
+ avg_skip_len += 1
+ if row["results"]["gsm8k"] == 0.0:
+ avg_skip_len += 1
+ if row["results"]["eq_bench"] == 0.0:
+ avg_skip_len += 1
+ if row["results"]["inst_follow"] == 0.0:
+ avg_skip_len += 1
+ if row["results"]["harmlessness"] == 0.0:
+ avg_skip_len += 1
+ if row["results"]["helpfulness"] == 0.0:
+ avg_skip_len += 1
+
+ current_score = np.sum(list(row["results"].values())) / (len(row["results"]) - avg_skip_len)
+ else:
+ current_score = row["results"][task.benchmark]
+
+ if current_score > current_max:
+ if current_date == last_date and len(scores[column]) > 0:
+ scores[column][-1] = {"model": current_model, "date": current_date, "score": current_score}
+ else:
+ scores[column].append({"model": current_model, "date": current_date, "score": current_score})
+ current_max = current_score
+ last_date = current_date
+
+ # Step 4: Return all dictionaries as DataFrames
+ return {k: pd.DataFrame(v) for k, v in scores.items()}
+
+
+def create_plot_df(scores_df: dict[str: pd.DataFrame]) -> pd.DataFrame:
+ """
+ Transforms the scores DataFrame into a new format suitable for plotting.
+
+ :param scores_df: A DataFrame containing metric scores and dates.
+ :return: A new DataFrame reshaped for plotting purposes.
+ """
+ # Initialize the list to store DataFrames
+ dfs = []
+
+ # Iterate over the cols and create a new DataFrame for each column
+ for col in BENCHMARK_COLS + [AutoEvalColumn.average.name]:
+ d = scores_df[col].reset_index(drop=True)
+ d["task"] = col
+ dfs.append(d)
+
+ # Concatenate all the created DataFrames
+ concat_df = pd.concat(dfs, ignore_index=True)
+
+ # Sort values by 'date'
+ concat_df.sort_values(by="date", inplace=True)
+ concat_df.reset_index(drop=True, inplace=True)
+ return concat_df
+
+
+def create_metric_plot_obj(
+ df: pd.DataFrame, metrics: list[str], title: str
+) -> Figure:
+ """
+ Create a Plotly figure object with lines representing different metrics
+ and horizontal dotted lines representing human baselines.
+
+ :param df: The DataFrame containing the metric values, names, and dates.
+ :param metrics: A list of strings representing the names of the metrics
+ to be included in the plot.
+ :param title: A string representing the title of the plot.
+ :return: A Plotly figure object with lines representing metrics and
+ horizontal dotted lines representing human baselines.
+ """
+
+ # Filter the DataFrame based on the specified metrics
+ df = df[df["task"].isin(metrics)]
+
+ # Filter the human baselines based on the specified metrics
+ filtered_human_baselines = {k: v for k, v in HUMAN_BASELINE.items() if k in metrics}
+
+ # Create a line figure using plotly express with specified markers and custom data
+ fig = px.line(
+ df,
+ x="date",
+ y="score",
+ color="task",
+ markers=True,
+ custom_data=["task", "score", "model"],
+ title=title,
+ )
+
+ # Update hovertemplate for better hover interaction experience
+ fig.update_traces(
+ hovertemplate="
".join(
+ [
+ "Model Name: %{customdata[2]}",
+ "Metric Name: %{customdata[0]}",
+ "Date: %{x}",
+ "Metric Value: %{y}",
+ ]
+ )
+ )
+
+ # Update the range of the y-axis
+ fig.update_layout(yaxis_range=[0, 100])
+
+ # Create a dictionary to hold the color mapping for each metric
+ metric_color_mapping = {}
+
+ # Map each metric name to its color in the figure
+ for trace in fig.data:
+ metric_color_mapping[trace.name] = trace.line.color
+
+ # Iterate over filtered human baselines and add horizontal lines to the figure
+ for metric, value in filtered_human_baselines.items():
+ color = metric_color_mapping.get(metric, "blue") # Retrieve color from mapping; default to blue if not found
+ location = "top left" if metric == "Ko-HellaSwag" else "bottom left" # Set annotation position
+ # Add horizontal line with matched color and positioned annotation
+ fig.add_hline(
+ y=value,
+ line_dash="dot",
+ annotation_text=f"{metric} human baseline",
+ annotation_position=location,
+ annotation_font_size=10,
+ annotation_font_color=color,
+ line_color=color,
+ )
+
+ return fig
+
+
+# Example Usage:
+# human_baselines dictionary is defined.
+# chart = create_metric_plot_obj(scores_df, ["ARC", "HellaSwag", "MMLU", "TruthfulQA"], human_baselines, "Graph Title")