{ "cells": [ { "cell_type": "code", "execution_count": 1, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:43:21.422682Z", "iopub.status.busy": "2025-06-03T18:43:21.422323Z", "iopub.status.idle": "2025-06-03T18:43:21.433788Z", "shell.execute_reply": "2025-06-03T18:43:21.433436Z", "shell.execute_reply.started": "2025-06-03T18:43:21.422664Z" } }, "outputs": [], "source": [ "# setup autoload\n", "%load_ext autoreload\n", "%autoreload 2" ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:58:21.040680Z", "start_time": "2024-05-16T13:58:19.777010Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:43:21.434457Z", "iopub.status.busy": "2025-06-03T18:43:21.434202Z", "iopub.status.idle": "2025-06-03T18:43:23.188013Z", "shell.execute_reply": "2025-06-03T18:43:23.187655Z", "shell.execute_reply.started": "2025-06-03T18:43:21.434445Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "The autoreload extension is already loaded. To reload it, use:\n", " %reload_ext autoreload\n" ] } ], "source": [ "import ast\n", "import os\n", "import shutil\n", "import sys\n", "from collections import defaultdict\n", "\n", "import json\n", "import numpy as np\n", "import pandas as pd\n", "from preference_data_preparation_4min_30b_task import *\n", "from preference_helper import *\n", "from sklearn.model_selection import train_test_split\n", "from suno_utils.utils.s3 import download_s3_files\n", "from suno_utils.utils.text import read_json, read_jsonl, write_json, write_jsonl\n", "from tqdm import tqdm\n", "\n", "pd.set_option(\"display.max_rows\", 500)\n", "pd.set_option(\"display.max_columns\", 500)\n", "pd.set_option(\"display.width\", 1000)\n", "\n", "# setup autoload\n", "%load_ext autoreload\n", "%autoreload 2\n", "\n", "\n", "def custom_parse(x):\n", " try:\n", " return json.loads(x)\n", " except:\n", " return {}" ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:58:21.082172Z", "start_time": "2024-05-16T13:58:21.041926Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:43:23.188700Z", "iopub.status.busy": "2025-06-03T18:43:23.188396Z", "iopub.status.idle": "2025-06-03T18:43:23.245326Z", "shell.execute_reply": "2025-06-03T18:43:23.244958Z", "shell.execute_reply.started": "2025-06-03T18:43:23.188686Z" } }, "outputs": [], "source": [ "OUT_DATA_DIR = \"/app/suno/data/dpo/auk_mix_t1_v14\"\n", "os.makedirs(OUT_DATA_DIR, exist_ok=True)\n", "shutil.copyfile(\n", " \"/app/suno/data/dpo/7v_v20_full/tokenizer_60k.json\",\n", " os.path.join(OUT_DATA_DIR, \"tokenizer_60k.json\"),\n", ")\n", "NPZ_DIR = \"/app/suno/data/dpo/30b_npz\"\n", "# NOTE FOR 30b we increase this from 6016 up\n", "N_TOKENS_AUDIO = 6016" ] }, { "cell_type": "code", "execution_count": 4, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:58:53.962528Z", "start_time": "2024-05-16T13:58:21.105919Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:43:23.246328Z", "iopub.status.busy": "2025-06-03T18:43:23.246152Z", "iopub.status.idle": "2025-06-03T18:43:37.586617Z", "shell.execute_reply": "2025-06-03T18:43:37.586255Z", "shell.execute_reply.started": "2025-06-03T18:43:23.246316Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Preference data shape (931852, 90)\n" ] } ], "source": [ "df = pd.read_pickle(\n", " \"/home/tony/Data/Preference/30b_v6/interesting_clips_v4_h_t_6_20250501_full_long.pkl\"\n", ")\n", "print(\"Preference data shape\", df.shape)" ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:43:37.587228Z", "iopub.status.busy": "2025-06-03T18:43:37.587012Z", "iopub.status.idle": "2025-06-03T18:43:39.597769Z", "shell.execute_reply": "2025-06-03T18:43:39.597399Z", "shell.execute_reply.started": "2025-06-03T18:43:37.587215Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "(931852, 90) (465684, 90)\n", "after date cut (465684, 90)\n" ] } ], "source": [ "df[\"created_at\"] = pd.to_datetime(df[\"created_at\"], utc=True)\n", "cutoff_date = pd.to_datetime(\"2025-04-12\", utc=True)\n", "# cutoff_date = pd.to_datetime(\"2025-04-17\", utc=True)\n", "print(df.shape, df[df[\"created_at\"] >= cutoff_date].shape)\n", "df = df[(df[\"created_at\"] >= cutoff_date)].copy()\n", "print(\"after date cut\", df.shape)" ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:43:39.602984Z", "iopub.status.busy": "2025-06-03T18:43:39.602745Z", "iopub.status.idle": "2025-06-03T18:43:40.429557Z", "shell.execute_reply": "2025-06-03T18:43:40.429193Z", "shell.execute_reply.started": "2025-06-03T18:43:39.602970Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "after dropna (465684, 85)\n" ] } ], "source": [ "df = df.dropna(axis=1, how=\"all\")\n", "print(\"after dropna\", df.shape)" ] }, { "cell_type": "code", "execution_count": 7, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:58:56.199480Z", "start_time": "2024-05-16T13:58:53.963687Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:43:40.430543Z", "iopub.status.busy": "2025-06-03T18:43:40.430301Z", "iopub.status.idle": "2025-06-03T18:49:02.664583Z", "shell.execute_reply": "2025-06-03T18:49:02.664215Z", "shell.execute_reply.started": "2025-06-03T18:43:40.430530Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "6442462\n", "6442462\n", "pre-downloaded df (465684, 85)\n", "downloaded df (465683, 85)\n" ] } ], "source": [ "converted_paths = os.listdir(NPZ_DIR)\n", "print(len(converted_paths))\n", "\n", "converted_paths = set([f.replace(\".npz\", \"\") for f in converted_paths])\n", "print(len(converted_paths))\n", "\n", "print(\"pre-downloaded df\", df.shape)\n", "df[df[\"s3_id\"].isin(converted_paths)].shape\n", "df = df[df[\"s3_id\"].isin(converted_paths)].copy()\n", "print(\"downloaded df\", df.shape)" ] }, { "cell_type": "code", "execution_count": 8, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:58:56.467253Z", "start_time": "2024-05-16T13:58:56.207647Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:49:02.665136Z", "iopub.status.busy": "2025-06-03T18:49:02.664973Z", "iopub.status.idle": "2025-06-03T18:49:03.822609Z", "shell.execute_reply": "2025-06-03T18:49:03.822260Z", "shell.execute_reply.started": "2025-06-03T18:49:02.665123Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "is_30b\n", "True 465683\n", "Name: count, dtype: int64\n" ] }, { "data": { "text/plain": [ "task\n", "cover 256083\n", "artist_consistency 182810\n", "infill 26786\n", "artist_extend 4\n", "Name: count, dtype: int64" ] }, "execution_count": 8, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df[\"is_30b\"] = df[\"model_name\"].str.contains(\"-t\")\n", "print(df[\"is_30b\"].value_counts())\n", "df[\"task\"].value_counts()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# LET's do the data prep" ] }, { "cell_type": "code", "execution_count": 9, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:49:03.823240Z", "iopub.status.busy": "2025-06-03T18:49:03.823001Z", "iopub.status.idle": "2025-06-03T18:49:03.836953Z", "shell.execute_reply": "2025-06-03T18:49:03.836653Z", "shell.execute_reply.started": "2025-06-03T18:49:03.823228Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "(465683, 86)\n" ] } ], "source": [ "# drop extend for now\n", "# reason is -- cause they are likely caused by extend from 13b\n", "# we don't want contamination\n", "# df = df[df[\"task\"] != \"extend\"].copy()\n", "print(df.shape)" ] }, { "cell_type": "code", "execution_count": 10, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:58:56.592883Z", "start_time": "2024-05-16T13:58:56.470781Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:49:03.837446Z", "iopub.status.busy": "2025-06-03T18:49:03.837271Z", "iopub.status.idle": "2025-06-03T18:49:03.996198Z", "shell.execute_reply": "2025-06-03T18:49:03.995838Z", "shell.execute_reply.started": "2025-06-03T18:49:03.837434Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "preference model_name \n", "False chirp-v4-h-t-6 232842\n", "True chirp-v4-h-t-6 232841\n", "Name: count, dtype: int64\n", "before filter on model name (465683, 86)\n", "after filter on model name (465683, 86)\n" ] } ], "source": [ "## for 13b this is easy for now\n", "print(df.groupby([\"preference\"])[\"model_name\"].value_counts())\n", "print(\"before filter on model name\", df.shape)\n", "df = df[df[\"model_name\"].isin([\"chirp-v4-h-t-6\", \"chirp-v3p5-engine-t-6\"])]\n", "# df = df[df[\"model_name\"].isin([\"chirp-v3p5-engine-t-6\"])]\n", "print(\"after filter on model name\", df.shape)" ] }, { "cell_type": "code", "execution_count": 11, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:49:03.996770Z", "iopub.status.busy": "2025-06-03T18:49:03.996573Z", "iopub.status.idle": "2025-06-03T18:49:04.012270Z", "shell.execute_reply": "2025-06-03T18:49:04.011995Z", "shell.execute_reply.started": "2025-06-03T18:49:03.996758Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "is_public\n", "False 454090\n", "True 11593\n", "Name: count, dtype: int64\n" ] } ], "source": [ "print(df[\"is_public\"].value_counts())\n", "# remove public for now cause fucking users\n", "# df = df[~df[\"is_public\"]]" ] }, { "cell_type": "code", "execution_count": 12, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:58:56.909539Z", "start_time": "2024-05-16T13:58:56.595736Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:49:04.012734Z", "iopub.status.busy": "2025-06-03T18:49:04.012579Z", "iopub.status.idle": "2025-06-03T18:49:04.624021Z", "shell.execute_reply": "2025-06-03T18:49:04.623628Z", "shell.execute_reply.started": "2025-06-03T18:49:04.012723Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "before filter on request id pairs (465683, 86)\n", "after filter on request id pairs (465682, 86)\n", "preference model_name \n", "False chirp-v4-h-t-6 232841\n", "True chirp-v4-h-t-6 232841\n", "Name: count, dtype: int64\n" ] } ], "source": [ "print(\"before filter on request id pairs\", df.shape)\n", "df = df[\n", " df[\"request_id\"].isin(\n", " df[\"request_id\"].value_counts().index[df[\"request_id\"].value_counts() == 2]\n", " )\n", "]\n", "print(\"after filter on request id pairs\", df.shape)\n", "print(df.groupby([\"preference\"])[\"model_name\"].value_counts())\n", "assert df.shape[0] == df[\"request_id\"].nunique() * 2" ] }, { "cell_type": "code", "execution_count": 13, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:49:04.625796Z", "iopub.status.busy": "2025-06-03T18:49:04.625607Z", "iopub.status.idle": "2025-06-03T18:50:15.511358Z", "shell.execute_reply": "2025-06-03T18:50:15.510996Z", "shell.execute_reply.started": "2025-06-03T18:49:04.625783Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "unique_requests 232841\n", "before removing duplicates (465682, 160)\n", "after removing duplicates (465682, 153)\n" ] } ], "source": [ "# Let's use the old selection for now -- for quality assurance\n", "# expand the metadata columns -- this takes forever...~ 6 mins\n", "# test_slice = df[\"metadata\"].apply(lambda x: ast.literal_eval(str(x)))\n", "# test_slice = df[\"metadata\"].apply(lambda x: custom_parse(x))\n", "test_slice = df[\"metadata\"] # .apply(lambda x: custom_parse(x))\n", "test_slice_series = test_slice.apply(pd.Series)\n", "df = pd.concat([df, test_slice_series], axis=1, join=\"inner\")\n", "print(\"unique_requests\", df[\"request_id\"].nunique())\n", "# remove the duplicates\n", "print(\"before removing duplicates\", df.shape)\n", "df = df.loc[:, ~df.columns.duplicated()].copy()\n", "print(\"after removing duplicates\", df.shape)" ] }, { "cell_type": "code", "execution_count": 14, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:50:15.512060Z", "iopub.status.busy": "2025-06-03T18:50:15.511749Z", "iopub.status.idle": "2025-06-03T18:50:16.319307Z", "shell.execute_reply": "2025-06-03T18:50:16.318970Z", "shell.execute_reply.started": "2025-06-03T18:50:15.512047Z" } }, "outputs": [ { "data": { "text/plain": [ "task\n", "cover 256082\n", "artist_consistency 182810\n", "infill 26786\n", "artist_extend 4\n", "Name: count, dtype: int64" ] }, "execution_count": 14, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df[\"task\"].value_counts()" ] }, { "cell_type": "code", "execution_count": 15, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:50:16.319885Z", "iopub.status.busy": "2025-06-03T18:50:16.319685Z", "iopub.status.idle": "2025-06-03T18:50:17.187823Z", "shell.execute_reply": "2025-06-03T18:50:17.187497Z", "shell.execute_reply.started": "2025-06-03T18:50:16.319873Z" } }, "outputs": [ { "data": { "text/plain": [ "pos_diff_preference\n", "1.0 161927\n", "2.0 70914\n", "Name: count, dtype: int64" ] }, "execution_count": 15, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df = df.sort_values(by=[\"request_id\", \"preference\", \"diff_preference\"])\n", "df[\"pos_diff_preference\"] = df[\"diff_preference\"].diff()\n", "# df[\"cer_diff_preference\"] = df[\"cer\"].diff()\n", "df[df[\"preference\"]][\"pos_diff_preference\"].value_counts()" ] }, { "cell_type": "code", "execution_count": 16, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:50:17.188415Z", "iopub.status.busy": "2025-06-03T18:50:17.188181Z", "iopub.status.idle": "2025-06-03T18:50:17.489079Z", "shell.execute_reply": "2025-06-03T18:50:17.488721Z", "shell.execute_reply.started": "2025-06-03T18:50:17.188402Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "positive param_experiment\n", "text_1 5758\n", "temp_s_80 5605\n", "temp_s_70 5489\n", "Name: count, dtype: int64\n" ] } ], "source": [ "try:\n", " print(\"positive\", df[df[\"preference\"]][\"param_experiment\"].value_counts())\n", "except:\n", " pass" ] }, { "cell_type": "code", "execution_count": 17, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:50:17.489680Z", "iopub.status.busy": "2025-06-03T18:50:17.489479Z", "iopub.status.idle": "2025-06-03T18:50:30.989567Z", "shell.execute_reply": "2025-06-03T18:50:30.989209Z", "shell.execute_reply.started": "2025-06-03T18:50:17.489667Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Found 39551 duplicated prompts 19776 unique requests\n", "Found 9451 request_ids with duplicate prompts but not highest play counts in their group\n", "['fa5aba0d-4424-438b-9f87-ba5d0ef503cb', '115a0270-2735-43a2-80d4-05b371926471', '14e7fadb-f34c-4f26-9bda-c378283f529a', '90a5dfe0-d1ba-4a62-9813-936e80ccabb8', '5c18e378-278c-4a9c-a60c-7c9508bbfaf3', 'f3edfa94-39fe-4c6b-b00d-7aea44f16447', '6897cd2a-59ee-4a54-831f-4e0f56ea74e1', '19726b9e-0f6e-4638-91f2-18d0a9d02050', '266d9dbb-ace9-42fe-b92e-e04d0d9bd3a8', 'a982e31f-1ff4-4503-83de-14cbe0a11f55']\n", "Before dedup user gen requests 465682\n", "After dedup user gen requests 465682\n" ] } ], "source": [ "# Find duplicated prompts with count > 2\n", "duplicate_entries = df.groupby(\n", " [\"user_id\", \"prompt_text\", \"tags\", \"task\", \"edited_clip_id\"]\n", ").filter(lambda x: len(x) > 2)\n", "print(\n", " \"Found\",\n", " len(duplicate_entries),\n", " \"duplicated prompts\",\n", " len(duplicate_entries[\"request_id\"].unique()),\n", " \"unique requests\",\n", ")\n", "\n", "# Group by user_id, prompt_text, and tags to find duplicate prompt groups\n", "prompt_groups = duplicate_entries.groupby(\n", " [\"user_id\", \"prompt_text\", \"tags\", \"task\", \"edited_clip_id\"]\n", ")\n", "\n", "# For each prompt group, find the request_id with the highest total reaction_play_count\n", "low_play_count_request_ids = []\n", "for prompt_key, prompt_group in prompt_groups:\n", " # Get the sum of reaction_play_count for each request_id in this group\n", " request_play_counts = prompt_group.groupby(\"request_id\")[\n", " \"reaction_play_count\"\n", " ].sum()\n", "\n", " # Find the max play count in this group\n", " max_play_count = request_play_counts.max()\n", "\n", " # Add request_ids that don't have the max play count to our filter list\n", " lower_play_count_request_ids = request_play_counts[\n", " request_play_counts < max_play_count\n", " ].index.tolist()\n", " low_play_count_request_ids.extend(lower_play_count_request_ids)\n", "\n", "# Display the filtered request IDs\n", "print(\n", " f\"Found {len(low_play_count_request_ids)} request_ids with duplicate prompts but not highest play counts in their group\"\n", ")\n", "print(\n", " low_play_count_request_ids[:10]\n", " if len(low_play_count_request_ids) > 10\n", " else low_play_count_request_ids\n", ")\n", "print(\"Before dedup user gen requests\", df.shape[0])\n", "# df = df[~df[\"request_id\"].isin(low_play_count_request_ids)]\n", "print(\"After dedup user gen requests\", df.shape[0])" ] }, { "cell_type": "code", "execution_count": 18, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:40.799375Z", "start_time": "2024-05-16T13:59:36.394236Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:50:30.990093Z", "iopub.status.busy": "2025-06-03T18:50:30.989949Z", "iopub.status.idle": "2025-06-03T18:50:33.107667Z", "shell.execute_reply": "2025-06-03T18:50:33.107316Z", "shell.execute_reply.started": "2025-06-03T18:50:30.990080Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "11875\n", "good_continue_at\n", "True 465682\n", "Name: count, dtype: int64\n", "\n", " Check some basics... \n", " preference\n", "False 232841\n", "True 232841\n", "Name: count, dtype: int64 is_30b\n", "True 465682\n", "Name: count, dtype: int64 model_name\n", "chirp-v4-h-t-6 465682\n", "Name: count, dtype: int64 preference model_name \n", "False chirp-v4-h-t-6 232841\n", "True chirp-v4-h-t-6 232841\n", "Name: count, dtype: int64\n", "task\n", "cover 256082\n", "artist_consistency 182810\n", "infill 26786\n", "artist_extend 4\n", "Name: count, dtype: int64\n" ] } ], "source": [ "df[\"id\"] = df[\"str_id\"]\n", "# get the original duration of the clips, if they are concacted\n", "df[\"original_duration_s\"] = df[\"total_start_s\"] + df[\"duration\"]\n", "# classify the continue at behavoirs by the duration choice\n", "audio_prompt_id_to_continue_at = {}\n", "\n", "for _, row in df[~df[\"continued_parent\"].isna()].iterrows():\n", " audio_prompt_id = row[\"continued_parent\"]\n", " if audio_prompt_id not in audio_prompt_id_to_continue_at:\n", " audio_prompt_id_to_continue_at[audio_prompt_id] = row[\"continue_at\"]\n", " else:\n", " # pick the max\n", " audio_prompt_id = max(\n", " audio_prompt_id_to_continue_at[audio_prompt_id], row[\"continue_at\"]\n", " )\n", "print(len(audio_prompt_id_to_continue_at))\n", "df[\"has_continue_and_start_continue_at\"] = df[\"id\"].apply(\n", " lambda x: audio_prompt_id_to_continue_at.get(x)\n", ")\n", "# we want continue at to be at most of the clip...\n", "df[\"good_continue_at\"] = (\n", " (df[\"has_continue_and_start_continue_at\"] / df[\"duration\"]) > 0.9\n", ") | df[\"has_continue_and_start_continue_at\"].isna()\n", "print(df[\"good_continue_at\"].value_counts())\n", "\n", "\n", "print(\n", " \"\\n Check some basics... \\n\",\n", " df[\"preference\"].value_counts(),\n", " df[\"is_30b\"].value_counts(),\n", " df[\"model_name\"].value_counts(),\n", " df.groupby([\"preference\"])[\"model_name\"].value_counts(),\n", ")\n", "\n", "df = df.sort_values(by=[\"request_id\", \"preference\"])\n", "df[\"duration_rel_diff\"] = df[\"duration\"].diff()\n", "df[\"play_rel_diff\"] = df[\"reaction_play_count\"].diff()\n", "print(df[\"task\"].value_counts())\n", "\n", "df[\"post_infill_duration\"] = (\n", " df[\"duration\"]\n", " + df[\"infill_context_end_s\"]\n", " - df[\"infill_context_start_s\"]\n", " - df[\"include_future_s\"]\n", " - df[\"include_history_s\"]\n", " - df[\"infill_dur_s\"]\n", ")" ] }, { "cell_type": "code", "execution_count": 19, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:50:33.108248Z", "iopub.status.busy": "2025-06-03T18:50:33.108048Z", "iopub.status.idle": "2025-06-03T18:50:33.276257Z", "shell.execute_reply": "2025-06-03T18:50:33.275933Z", "shell.execute_reply.started": "2025-06-03T18:50:33.108235Z" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "df[\"post_infill_duration\"].hist(bins=np.linspace(-5, 360, 100))\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": 20, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.035167Z", "start_time": "2024-05-16T13:59:40.801098Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:50:33.276843Z", "iopub.status.busy": "2025-06-03T18:50:33.276689Z", "iopub.status.idle": "2025-06-03T18:50:34.560313Z", "shell.execute_reply": "2025-06-03T18:50:34.559953Z", "shell.execute_reply.started": "2025-06-03T18:50:33.276830Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "after duration 0.9955162535807697\n", "after infill duration 0.9978977070189529\n", "neg_filter_reaction_play_count 1.0\n", "neg_filter_upvote_count 0.9901\n", "neg_filter_norm_play_frac 1.0\n", "neg_filter_continues 1.0\n", "----------------\n", "pos_filter_continues 1.0\n", "pos_filter_reaction_play_count 1.0\n", "pos_filter_relative_play_count 0.9744\n", "pos_filter_cer_diff_preference 1.0\n", "pos_filter_bad_flags 0.9999\n", "after filter on play counts 0.9731\n", "after filter on higher quality 0.3144\n", "----------------\n", "negative 229029 positive 63875\n", "----------------\n", "total pair requests 232841 --> selected pair requests 63272 frac 0.272\n" ] } ], "source": [ "normal_pos_play_count = 3\n", "# this is lower, cause a concat is probably already ensuring that it is good\n", "concat_pos_play_count = 1\n", "# this is a filter on the concated clip\n", "concat_total_play_count = 3\n", "\n", "all_fitlers = (df[\"duration\"] >= 10) & (df[\"duration\"] <= 240)\n", "print(\"after duration\", all_fitlers.sum() / df.shape[0])\n", "infill_duration_filter = (\n", " ~df[\"task\"].isin(\n", " [\n", " \"infill\",\n", " \"infill_intro\",\n", " \"infill_outro\",\n", " ]\n", " )\n", ") | (df[\"post_infill_duration\"] <= 239)\n", "print(\"after infill duration\", infill_duration_filter.sum() / df.shape[0])\n", "# negative fitlers\n", "total_negative = df[~df[\"preference\"]].shape[0]\n", "neg_filter_reaction_play_count = (~df[\"preference\"]) & (df[\"reaction_play_count\"] >= 1)\n", "print(\n", " \"neg_filter_reaction_play_count\",\n", " round(neg_filter_reaction_play_count.sum() / total_negative, 4),\n", ")\n", "neg_filter_upvote_count = (~df[\"preference\"]) & (df[\"upvote_count\"] == 0)\n", "print(\n", " \"neg_filter_upvote_count\",\n", " round(neg_filter_upvote_count.sum() / total_negative, 4),\n", ")\n", "neg_filter_norm_play_frac = (~df[\"preference\"]) & (df[\"norm_play_frac\"] <= 3.1)\n", "print(\n", " \"neg_filter_norm_play_frac\",\n", " round(neg_filter_norm_play_frac.sum() / total_negative, 4),\n", ")\n", "neg_filter_continues = (~df[\"preference\"]) & (\n", " df[\"has_continue_and_start_continue_at\"].isna()\n", ")\n", "print(\n", " \"neg_filter_continues\",\n", " round(neg_filter_continues.sum() / total_negative, 4),\n", ")\n", "\n", "neg_filter_selection_mask = (\n", " all_fitlers\n", " & infill_duration_filter\n", " & neg_filter_reaction_play_count\n", " & neg_filter_upvote_count\n", " & neg_filter_norm_play_frac\n", " & neg_filter_continues\n", ")\n", "\n", "print(\"----------------\")\n", "total_positive = df[df[\"preference\"]].shape[0]\n", "assert total_positive == total_negative\n", "pos_filter_continues = (df[\"preference\"]) & (df[\"good_continue_at\"])\n", "print(\"pos_filter_continues\", round(pos_filter_continues.sum() / total_positive, 4))\n", "pos_filter_reaction_play_count = (df[\"preference\"]) & (df[\"reaction_play_count\"] >= 1)\n", "print(\n", " \"pos_filter_reaction_play_count\",\n", " round(pos_filter_reaction_play_count.sum() / total_positive, 4),\n", ")\n", "pos_filter_relative_play_count = (df[\"preference\"]) & (df[\"play_rel_diff\"] >= 0)\n", "print(\n", " \"pos_filter_relative_play_count\",\n", " round(pos_filter_relative_play_count.sum() / total_positive, 4),\n", ")\n", "pos_filter_cer_diff_preference = (\n", " df[\n", " \"preference\"\n", " ] # & (df[\"pos_diff_preference\"] == 2) # & (df[\"cer_diff_preference\"] < 0.5) & (df[\"cer\"] < 0.99)\n", ")\n", "print(\n", " \"pos_filter_cer_diff_preference\",\n", " round(pos_filter_cer_diff_preference.sum() / total_positive, 4),\n", ")\n", "pos_filter_bad_flags = (\n", " (df[\"preference\"]) & (df[\"flag_count\"] == 0) & (df[\"dislike_count\"] == 0)\n", ")\n", "print(\n", " \"pos_filter_bad_flags\",\n", " round(pos_filter_bad_flags.sum() / total_positive, 4),\n", ")\n", "pos_filter_play_counts = (df[\"preference\"]) & (\n", " (\n", " (df[\"part_of_concat\"])\n", " & (df[\"reaction_play_count\"] >= concat_pos_play_count)\n", " & (df[\"concat_play_counts\"] >= concat_total_play_count)\n", " )\n", " | (\n", " (~df[\"part_of_concat\"]) & (df[\"reaction_play_count\"] >= normal_pos_play_count)\n", " # & (df[\"norm_play_frac\"] >= 2.1) # this is a bit of a luxury cut...\n", " )\n", " | (df[\"task\"].isin([\"infill\", \"infill_intro\", \"infill_outro\"]))\n", ")\n", "print(\n", " \"after filter on play counts\",\n", " round(pos_filter_play_counts.sum() / total_positive, 4),\n", ")\n", "high_quality_tasks_filter = (\n", " (df[\"task\"].isin([\"cover\", \"extend\", \"artist_consistency\", \"\"]))\n", " & (\n", " (df[\"upvote_count\"] >= 1) # (df[\"upvote_count\"] >= 1)\n", " | (df[\"reaction_play_count\"] >= 5)\n", " | (df[\"concat_play_counts\"] >= 5)\n", " )\n", " & (\n", " (df[\"part_of_concat\"])\n", " | (\n", " (~df[\"part_of_concat\"])\n", " & (df[\"norm_play_frac\"] >= 5.1) # this is a bit of a luxury cut...\n", " & (\n", " df[\"norm_play_frac\"] >= df[\"reaction_play_count\"] / 3\n", " ) # play duration is not low on average\n", " )\n", " )\n", ")\n", "medium_quality_tasks_filter = (\n", " df[\"task\"].isin(\n", " [\n", " \"infill\",\n", " \"infill_intro\",\n", " \"infill_outro\",\n", " ]\n", " )\n", ") & ( # let more infill through only in this case...\n", " (\n", " df[\"upvote_count\"] >= 0\n", " ) # (df[\"upvote_count\"] >= 1) (df[\"pos_diff_preference\"] == 2)\n", " | (df[\"reaction_play_count\"] >= 1)\n", " | (df[\"concat_play_counts\"] >= 1)\n", ")\n", "pos_filter_higher_quality = (df[\"preference\"]) & (\n", " high_quality_tasks_filter | medium_quality_tasks_filter\n", ")\n", "print(\n", " \"after filter on higher quality\",\n", " round(pos_filter_higher_quality.sum() / total_positive, 4),\n", ")\n", "\n", "user_gen_filter = (\n", " df[\"user_n_clips\"] >= 100\n", ") # user needs to have genereated at least 100 over the time period\n", "\n", "print(\"----------------\")\n", "pos_filter_selectin_mask = (\n", " (df[\"preference\"]) # get basics aligned\n", " & all_fitlers\n", " & infill_duration_filter\n", " & pos_filter_continues\n", " & pos_filter_reaction_play_count\n", " & pos_filter_relative_play_count\n", " & pos_filter_cer_diff_preference\n", " & pos_filter_bad_flags\n", " & pos_filter_play_counts\n", " & pos_filter_higher_quality\n", " & user_gen_filter\n", ")\n", "print(\n", " \"negative\",\n", " sum(neg_filter_selection_mask),\n", " \"positive\",\n", " sum(pos_filter_selectin_mask),\n", ")\n", "\n", "neg_filter_requests = df[neg_filter_selection_mask][\"request_id\"].unique()\n", "pos_filter_requests = df[pos_filter_selectin_mask][\"request_id\"].unique()\n", "# looking for very strong signal here:\n", "# listen to the positive/negative more than once\n", "# disliked one of the clips\n", "unique_requests = set(pos_filter_requests).intersection(neg_filter_requests)\n", "print(\"----------------\")\n", "print(\n", " \"total pair requests\",\n", " df[\"request_id\"].nunique(),\n", " \" --> selected pair requests\",\n", " len(unique_requests),\n", " f\"frac {len(unique_requests) / df['request_id'].nunique():.3f}\",\n", ")" ] }, { "cell_type": "code", "execution_count": 21, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.250737Z", "start_time": "2024-05-16T13:59:41.036434Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:50:34.560925Z", "iopub.status.busy": "2025-06-03T18:50:34.560763Z", "iopub.status.idle": "2025-06-03T18:50:35.016290Z", "shell.execute_reply": "2025-06-03T18:50:35.015931Z", "shell.execute_reply.started": "2025-06-03T18:50:34.560912Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "auk_mix_t1_v14 requests 63272 clips 126544 total khrs 5.382; N gpus for 1000 iters 7.909; 4 gpus for x iters 1977.250; n unique users 26926 n pro users 26111\n" ] } ], "source": [ "df_slice = df[df[\"request_id\"].isin(set(unique_requests))].copy()\n", "print(\n", " f\"{os.path.basename(OUT_DATA_DIR)} requests\",\n", " df_slice[\"request_id\"].nunique(),\n", " \"clips\",\n", " df_slice.shape[0],\n", " f\"total khrs {sum(df_slice['duration'] / 3600 / 1000):.3f};\",\n", " f\"N gpus for 1000 iters {df_slice.shape[0] / 8 / 2 / 1000:.3f};\",\n", " f\"4 gpus for x iters {df_slice.shape[0] / 8 / 2 / 4:.3f};\",\n", " f\"n unique users {df_slice['user_id'].nunique()}\",\n", " f\"n pro users {df_slice[df_slice['is_pro_user']]['user_id'].nunique()}\",\n", ")\n", "# 76171 152342 total khrs 2.880 n gpus for 1250 iters 3.809\n", "# v10 has 78866\n", "# v14 has 110402\n", "# 30b_t4_v12 requests 47755 clips 95510 total khrs 4.555; N gpus for 1000 iters 5.969; 4 gpus for x iters 1492.344; n unique users 13790 n pro users 13526\n", "# 30b_t4_v20 requests 18306 clips 36612 total khrs 1.754; N gpus for 1000 iters 2.288; 4 gpus for x iters 572.062; n unique users 6943 n pro users 6714\n", "# 30b_t5_v5 requests 21318 clips 42636 total khrs 1.938; N gpus for 1000 iters 2.665; 4 gpus for x iters 666.188; n unique users 11661 n pro users 9347\n", "# 30b_t5_v6 requests 22833 clips 45666 total khrs 2.072; N gpus for 1000 iters 2.854; 4 gpus for x iters 713.531; n unique users 12576 n pro users 9910\n", "# 30b_t5_v7 requests 33197 clips 66394 total khrs 2.846; N gpus for 1000 iters 4.150; 4 gpus for x iters 1037.406; n unique users 16665 n pro users 13050\n", "# 30b_t5_v8 requests 20907 clips 41814 total khrs 1.629; N gpus for 1000 iters 2.613; 4 gpus for x iters 653.344; n unique users 11102 n pro users 9052\n", "# 30b_t5_v9 requests 24645 clips 49290 total khrs 1.918; N gpus for 1000 iters 3.081; 4 gpus for x iters 770.156; n unique users 12661 n pro users 10229\n", "# 30b_t5_v11 requests 21975 clips 43950 total khrs 1.508; N gpus for 1000 iters 2.747; 4 gpus for x iters 686.719; n unique users 11173 n pro users 9270\n", "# 30b_t5_v12 requests 21436 clips 42872 total khrs 1.582; N gpus for 1000 iters 2.679; 4 gpus for x iters 669.875; n unique users 11553 n pro users 9374\n", "# 30b_t5_v13 requests 25912 clips 51824 total khrs 1.922; N gpus for 1000 iters 3.239; 4 gpus for x iters 809.750; n unique users 13429 n pro users 10737\n", "# 30b_t5_v14 requests 27687 clips 55374 total khrs 2.036; N gpus for 1000 iters 3.461; 4 gpus for x iters 865.219; n unique users 14196 n pro users 11227\n", "# 30b_t5_v15 requests 29563 clips 59126 total khrs 2.225; N gpus for 1000 iters 3.695; 4 gpus for x iters 923.844; n unique users 14684 n pro users 11642\n", "# 30b_t6_v1 requests 24678 clips 49356 total khrs 1.934; N gpus for 1000 iters 3.085; 4 gpus for x iters 771.188; n unique users 11350 n pro users 10065\n", "# 30b_t6_v2 requests 49341 clips 98682 total khrs 4.332; N gpus for 1000 iters 6.168; 4 gpus for x iters 1541.906; n unique users 19992 n pro users 16972\n", "# 30b_t6_v3 requests 54967 clips 109934 total khrs 4.826; N gpus for 1000 iters 6.871; 4 gpus for x iters 1717.719; n unique users 21653 n pro users 18266\n", "# 30b_t6_v5 requests 29459 clips 58918 total khrs 2.657; N gpus for 1000 iters 3.682; 4 gpus for x iters 920.594; n unique users 14091 n pro users 11884\n", "# 30b_t6_v7 requests 40056 clips 80112 total khrs 3.354; N gpus for 1000 iters 5.007; 4 gpus for x iters 1251.750; n unique users 17440 n pro users 14476\n", "# 30b_t6_v8 requests 31681 clips 63362 total khrs 2.681; N gpus for 1000 iters 3.960; 4 gpus for x iters 990.031; n unique users 15180 n pro users 15078\n", "# 30b_t6_v9 requests 55785 clips 111570 total khrs 4.727; N gpus for 1000 iters 6.973; 4 gpus for x iters 1743.281; n unique users 23114 n pro users 22871\n", "# 30b_t6_v10 requests 52209 clips 104418 total khrs 4.515; N gpus for 1000 iters 6.526; 4 gpus for x iters 1631.531; n unique users 22216 n pro users 21893\n", "# 30b_t6_v11 requests 31338 clips 62676 total khrs 2.493; N gpus for 1000 iters 3.917; 4 gpus for x iters 979.312; n unique users 15190 n pro users 14973\n", "# 30b_t6_v13 requests 50856 clips 101712 total khrs 3.987; N gpus for 1000 iters 6.357; 4 gpus for x iters 1589.250; n unique users 23547 n pro users 23097\n", "# 30b_t6_v15 requests 65213 clips 130426 total khrs 5.217; N gpus for 1000 iters 8.152; 4 gpus for x iters 2037.906; n unique users 28763 n pro users 27964\n", "# 30b_t6_v16 requests 37907 clips 75814 total khrs 2.956; N gpus for 1000 iters 4.738; 4 gpus for x iters 1184.594; n unique users 18448 n pro users 17212\n", "# 30b_t6_v17 requests 51543 clips 103086 total khrs 4.154; N gpus for 1000 iters 6.443; 4 gpus for x iters 1610.719; n unique users 24223 n pro users 21829\n", "# 30b_t6_v18 requests 94965 clips 189930 total khrs 7.483; N gpus for 1000 iters 11.871; 4 gpus for x iters 2967.656; n unique users 23852 n pro users 21767\n", "# 30b_t6_v19 requests 34289 clips 68578 total khrs 2.846; N gpus for 1000 iters 4.286; 4 gpus for x iters 1071.531; n unique users 11802 n pro users 10830\n", "# 30b_t6_v20 requests 53258 clips 106516 total khrs 4.270; N gpus for 1000 iters 6.657; 4 gpus for x iters 1664.312; n unique users 22096 n pro users 21171\n", "# 30b_t6_v22 requests 29971 clips 59942 total khrs 1.992; N gpus for 1000 iters 3.746; 4 gpus for x iters 936.594; n unique users 13611 n pro users 13036\n", "# 30b_t6_v24 requests 45333 clips 90666 total khrs 3.343; N gpus for 1000 iters 5.667; 4 gpus for x iters 1416.656; n unique users 18853 n pro users 18049\n", "# 30b_t6_v25 requests 28154 clips 56308 total khrs 2.401; N gpus for 1000 iters 3.519; 4 gpus for x iters 879.812; n unique users 12946 n pro users 12361\n", "# 30b_t6_v26 requests 30428 clips 60856 total khrs 2.599; N gpus for 1000 iters 3.804; 4 gpus for x iters 950.875; n unique users 13822 n pro users 13163\n", "# 30b_t6_v27 requests 43731 clips 87462 total khrs 3.862; N gpus for 1000 iters 5.466; 4 gpus for x iters 1366.594; n unique users 17898 n pro users 16865\n", "# 30b_t6_v28 requests 56869 clips 113738 total khrs 5.064; N gpus for 1000 iters 7.109; 4 gpus for x iters 1777.156; n unique users 22004 n pro users 20244\n", "# 30b_t6_v29 requests 32729 clips 65458 total khrs 2.604; N gpus for 1000 iters 4.091; 4 gpus for x iters 1022.781; n unique users 13739 n pro users 13503\n", "# 30b_t6_v30 requests 32553 clips 65106 total khrs 2.592; N gpus for 1000 iters 4.069; 4 gpus for x iters 1017.281; n unique users 13712 n pro users 13478\n", "# 30b_t6_v31 requests 41523 clips 83046 total khrs 3.164; N gpus for 1000 iters 5.190; 4 gpus for x iters 1297.594; n unique users 16266 n pro users 15967\n", "# 30b_t6_v32 requests 55942 clips 111884 total khrs 4.361; N gpus for 1000 iters 6.993; 4 gpus for x iters 1748.188; n unique users 20808 n pro users 20271\n", "# 30b_t6_v33 requests 182096 clips 364192 total khrs 13.745; N gpus for 1000 iters 22.762; 4 gpus for x iters 5690.500; n unique users 40437 n pro users 34524\n", "# auk_mix_t1_v2 requests 102002 clips 204004 total khrs 9.191; N gpus for 1000 iters 12.750; 4 gpus for x iters 3187.562; n unique users 36408 n pro users 34038\n", "# auk_mix_t1_v13 requests 53011 clips 106022 total khrs 4.496; N gpus for 1000 iters 6.626; 4 gpus for x iters 1656.594; n unique users 23713 n pro users 23117\n", "# auk_mix_t1_v14 requests 63272 clips 126544 total khrs 5.382; N gpus for 1000 iters 7.909; 4 gpus for x iters 1977.250; n unique users 26926 n pro users 26111" ] }, { "cell_type": "code", "execution_count": 22, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.277006Z", "start_time": "2024-05-16T13:59:41.252105Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:50:35.016936Z", "iopub.status.busy": "2025-06-03T18:50:35.016692Z", "iopub.status.idle": "2025-06-03T18:50:35.067301Z", "shell.execute_reply": "2025-06-03T18:50:35.066966Z", "shell.execute_reply.started": "2025-06-03T18:50:35.016922Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "positive in playlist (13493, 159)\n", "task\n", "cover 53704\n", "artist_consistency 51122\n", "infill 21718\n", "Name: count, dtype: int64\n" ] } ], "source": [ "test_mask = (df_slice[\"preference\"]) & (\n", " (df_slice[\"is_in_playlist\"]) | (df_slice[\"concat_in_playlist\"])\n", ")\n", "print(\"positive in playlist\", df_slice[test_mask].shape)\n", "print(df_slice[\"task\"].value_counts())" ] }, { "cell_type": "code", "execution_count": 23, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:50:35.067867Z", "iopub.status.busy": "2025-06-03T18:50:35.067676Z", "iopub.status.idle": "2025-06-03T18:50:35.201249Z", "shell.execute_reply": "2025-06-03T18:50:35.200924Z", "shell.execute_reply.started": "2025-06-03T18:50:35.067854Z" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "df_slice[\"post_infill_duration\"].hist(bins=np.linspace(-5, 300, 100))\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": 24, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.323409Z", "start_time": "2024-05-16T13:59:41.278278Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:50:35.201815Z", "iopub.status.busy": "2025-06-03T18:50:35.201622Z", "iopub.status.idle": "2025-06-03T18:50:35.215331Z", "shell.execute_reply": "2025-06-03T18:50:35.215061Z", "shell.execute_reply.started": "2025-06-03T18:50:35.201801Z" } }, "outputs": [], "source": [ "# interesting_clips_must_be_positive_mask = (\n", "# (df_slice[\"upvoted\"] == True)\n", "# | (df_slice[\"has_action\"] == True)\n", "# | (df_slice[\"part_of_concat\"] == True)\n", "# )\n", "# interesting_clips_must_be_not_negative_mask = (df_slice[\"downvoted\"] == False) # & (df_slice[\"dislike_count\"] < 1)\n", "# interesting_clips_mask = interesting_clips_must_be_positive_mask & interesting_clips_must_be_not_negative_mask\n", "# assert interesting_clips_mask.eq(df_slice[\"preference\"]).all()" ] }, { "cell_type": "code", "execution_count": 25, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.392244Z", "start_time": "2024-05-16T13:59:41.324472Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:50:35.215816Z", "iopub.status.busy": "2025-06-03T18:50:35.215644Z", "iopub.status.idle": "2025-06-03T18:50:35.225703Z", "shell.execute_reply": "2025-06-03T18:50:35.225448Z", "shell.execute_reply.started": "2025-06-03T18:50:35.215805Z" } }, "outputs": [], "source": [ "# save positive ids\n", "# positive_preference_ids = df_slice[df_slice[\"preference\"] == False][\"s3_id\"].to_json(orient='values')\n", "# with open('/home/tony/Data/Preference/7b_v2/7v_v20_full_recut_id_negative.json', 'w') as file:\n", "# file.write(positive_preference_ids)" ] }, { "cell_type": "code", "execution_count": 26, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:50:35.226170Z", "iopub.status.busy": "2025-06-03T18:50:35.226004Z", "iopub.status.idle": "2025-06-03T18:50:35.235969Z", "shell.execute_reply": "2025-06-03T18:50:35.235713Z", "shell.execute_reply.started": "2025-06-03T18:50:35.226159Z" } }, "outputs": [], "source": [ "# df_slice_2 = pd.read_pickle(\"/home/tony/Data/Preference/30b_v6/interesting_clips_v4_t_6_20241118_full_slice.pkl\")\n", "# df_total = pd.concat([df_slice, df_slice_2])\n", "# print(df_total.shape)\n", "# df_slice = df_total" ] }, { "cell_type": "code", "execution_count": 27, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:50:35.236465Z", "iopub.status.busy": "2025-06-03T18:50:35.236269Z", "iopub.status.idle": "2025-06-03T18:50:35.246348Z", "shell.execute_reply": "2025-06-03T18:50:35.246094Z", "shell.execute_reply.started": "2025-06-03T18:50:35.236454Z" } }, "outputs": [], "source": [ "# df_slice_prev = pd.read_pickle(\n", "# \"/home/tony/Data/Preference/30b_v2/interesting_clips_v4_t_3_v13_20240902_slice.pkl\"\n", "# )\n", "# df_slice_prev = df_slice_prev[\n", "# (\n", "# (df_slice_prev[\"task\"] != \"infill\")\n", "# & (df_slice_prev[\"task\"] != \"cover\")\n", "# & (df_slice_prev[\"task\"] != \"artist_consistency\")\n", "# )\n", "# ].copy()\n", "# df_slice = pd.concat([df_slice, df_slice_prev])\n", "# print(df_slice.shape)\n", "# print(df_slice[\"task\"].value_counts())" ] }, { "cell_type": "code", "execution_count": 28, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:50:35.246835Z", "iopub.status.busy": "2025-06-03T18:50:35.246651Z", "iopub.status.idle": "2025-06-03T18:50:35.256718Z", "shell.execute_reply": "2025-06-03T18:50:35.256463Z", "shell.execute_reply.started": "2025-06-03T18:50:35.246823Z" } }, "outputs": [], "source": [ "# df_slice[(df_slice[\"preference\"]) & ((df_slice[\"task\"].str.strip() == \"\") | (df_slice[\"task\"].str.strip() == \"cover\"))].to_pickle(\n", "# \"/home/tony/Data/Preference/30b_v5/interesting_clips_v4_t_5_20241018_full_with_cer_pos_gen.pkl\"\n", "# )\n", "# df_slice[(df_slice[\"preference\"]) & ((df_slice[\"task\"].str.strip() == \"\") | (df_slice[\"task\"].str.strip() == \"cover\"))].shape" ] }, { "cell_type": "code", "execution_count": 29, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:50:35.257162Z", "iopub.status.busy": "2025-06-03T18:50:35.257016Z", "iopub.status.idle": "2025-06-03T18:50:35.267032Z", "shell.execute_reply": "2025-06-03T18:50:35.266774Z", "shell.execute_reply.started": "2025-06-03T18:50:35.257151Z" } }, "outputs": [], "source": [ "# df_slice.to_pickle(\"/home/tony/Data/Preference/30b_v6/interesting_clips_v4_h_t_6_20250108_full_slice.pkl\")" ] }, { "cell_type": "code", "execution_count": 30, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:50:35.267490Z", "iopub.status.busy": "2025-06-03T18:50:35.267325Z", "iopub.status.idle": "2025-06-03T18:50:35.277243Z", "shell.execute_reply": "2025-06-03T18:50:35.276985Z", "shell.execute_reply.started": "2025-06-03T18:50:35.267479Z" } }, "outputs": [], "source": [ "# with open(\"/home/tony/Data/Preference/30b_v6/similarities_30b_t6.json\", \"r\") as fp:\n", "# all_similarities = json.load(fp)\n", "\n", "# df_slice[\"similarity\"] = df_slice[\"id\"].map(all_similarities)\n", "# df_cover = df_slice[df_slice[\"task\"] == \"cover\"].copy()\n", "# df_artist = df_slice[df_slice[\"task\"] == \"artist_consistency\"].copy()" ] }, { "cell_type": "code", "execution_count": 31, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:50:35.277739Z", "iopub.status.busy": "2025-06-03T18:50:35.277546Z", "iopub.status.idle": "2025-06-03T18:50:35.287627Z", "shell.execute_reply": "2025-06-03T18:50:35.287374Z", "shell.execute_reply.started": "2025-06-03T18:50:35.277728Z" } }, "outputs": [], "source": [ "# df_cover[df_cover[\"preference\"]][\"similarity\"].hist(bins=200, label=\"positive\", alpha=0.5)\n", "# df_cover[~df_cover[\"preference\"]][\"similarity\"].hist(bins=200, label=\"negative\", alpha=0.5)\n", "# plt.legend()\n", "# plt.show()\n", "# df_cover[\"similarity_diff\"] = df_cover[\"similarity\"].diff()\n", "# df_cover[df_cover[\"preference\"]][\"similarity_diff\"].hist(bins=200)\n", "# plt.show()\n", "# df_cover[\"continued_parent\"] = None\n", "# df_cover[\"continue_at\"] = -1\n", "# too_similar = (df_cover[\"similarity\"] >= 0.99)\n", "# too_different = ( df_cover[\"similarity\"] < 0.25)\n", "# too_big_difference = (df_cover[\"similarity_diff\"] < -0.2)\n", "# df_cover_drops_id = df_cover[\n", "# (df_cover[\"preference\"])\n", "# & (too_similar | too_different | too_big_difference)\n", "# ][\"s3_id\"].unique()\n", "# print(\"total rows\", df_cover.shape[0], \"total drops\", len(df_cover_drops_id))" ] }, { "cell_type": "code", "execution_count": 32, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:50:35.288082Z", "iopub.status.busy": "2025-06-03T18:50:35.287920Z", "iopub.status.idle": "2025-06-03T18:50:35.298005Z", "shell.execute_reply": "2025-06-03T18:50:35.297748Z", "shell.execute_reply.started": "2025-06-03T18:50:35.288071Z" } }, "outputs": [], "source": [ "# df_artist[df_artist[\"preference\"]][\"similarity\"].hist(bins=200, label=\"positive\", alpha=0.5)\n", "# df_artist[~df_artist[\"preference\"]][\"similarity\"].hist(bins=200, label=\"negative\", alpha=0.5)\n", "# plt.legend()\n", "# plt.show()\n", "# df_artist[\"similarity_diff\"] = df_artist[\"similarity\"].diff()\n", "# df_artist[df_artist[\"preference\"]][\"similarity_diff\"].hist(bins=200)\n", "# plt.show()\n", "# too_similar = (df_artist[\"similarity\"] >= 0.95)\n", "# too_different = ( df_artist[\"similarity\"] < 0.4)\n", "# too_big_difference = (df_artist[\"similarity_diff\"] < -0.2)\n", "# df_artist_drops_id = df_artist[\n", "# (df_artist[\"preference\"])\n", "# & (too_similar | too_different | too_big_difference)\n", "# ][\"s3_id\"].unique()\n", "# print(\"total rows\", df_artist.shape[0], \"total drops\", len(df_artist_drops_id))" ] }, { "cell_type": "code", "execution_count": 33, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:50:35.298447Z", "iopub.status.busy": "2025-06-03T18:50:35.298301Z", "iopub.status.idle": "2025-06-03T18:50:35.308335Z", "shell.execute_reply": "2025-06-03T18:50:35.308084Z", "shell.execute_reply.started": "2025-06-03T18:50:35.298436Z" } }, "outputs": [], "source": [ "# print(\"before filter on cover similarity\", df_slice.shape)\n", "# df_slice = df_slice[~df_slice[\"s3_id\"].isin(df_cover_drops_id)].copy()\n", "# print(\"after filter on cover similarity\", df_slice.shape)\n", "# df_slice = df_slice[~df_slice[\"s3_id\"].isin(df_artist_drops_id)].copy()\n", "# print(\"after filter on artist similarity\", df_slice.shape)\n", "# df_slice = df_slice[\n", "# df_slice[\"request_id\"].isin(\n", "# df_slice[\"request_id\"].value_counts().index[df_slice[\"request_id\"].value_counts() == 2]\n", "# )\n", "# ]\n", "# print(\"after filter on request id pairs\", df_slice.shape)\n", "# print(df_slice.groupby([\"preference\"])[\"model_name\"].value_counts())\n", "# assert df_slice.shape[0] == df_slice[\"request_id\"].nunique() * 2" ] }, { "cell_type": "code", "execution_count": 34, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:50:35.308788Z", "iopub.status.busy": "2025-06-03T18:50:35.308627Z", "iopub.status.idle": "2025-06-03T18:50:35.319960Z", "shell.execute_reply": "2025-06-03T18:50:35.319700Z", "shell.execute_reply.started": "2025-06-03T18:50:35.308777Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "is_public\n", "False 121494\n", "True 5050\n", "Name: count, dtype: int64\n" ] } ], "source": [ "print(df_slice[\"is_public\"].value_counts())" ] }, { "cell_type": "code", "execution_count": 35, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:50:35.320436Z", "iopub.status.busy": "2025-06-03T18:50:35.320253Z", "iopub.status.idle": "2025-06-03T18:50:35.352175Z", "shell.execute_reply": "2025-06-03T18:50:35.351860Z", "shell.execute_reply.started": "2025-06-03T18:50:35.320425Z" } }, "outputs": [], "source": [ "df_slice[\"npz_path\"] = df_slice[\"s3_id\"].map(lambda x: f\"{NPZ_DIR}/{x}.npz\")" ] }, { "cell_type": "code", "execution_count": 36, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T14:00:20.866354Z", "start_time": "2024-05-16T14:00:12.443344Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:50:35.352703Z", "iopub.status.busy": "2025-06-03T18:50:35.352523Z", "iopub.status.idle": "2025-06-03T18:50:35.997156Z", "shell.execute_reply": "2025-06-03T18:50:35.995850Z", "shell.execute_reply.started": "2025-06-03T18:50:35.352691Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "(126544, 160)\n", "task\n", "cover 53704\n", "artist_consistency 51122\n", "infill 21718\n", "Name: count, dtype: int64\n", "(104826, 160)\n", "task\n", "cover 53704\n", "artist_consistency 51122\n", "Name: count, dtype: int64\n" ] }, { "ename": "NameError", "evalue": "name 'BREAK' is not defined", "output_type": "error", "traceback": [ "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", "\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)", "Cell \u001b[0;32mIn[36], line 9\u001b[0m\n\u001b[1;32m 7\u001b[0m \u001b[38;5;28mprint\u001b[39m(df_slice\u001b[38;5;241m.\u001b[39mshape)\n\u001b[1;32m 8\u001b[0m \u001b[38;5;28mprint\u001b[39m(df_slice[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtask\u001b[39m\u001b[38;5;124m\"\u001b[39m]\u001b[38;5;241m.\u001b[39mvalue_counts())\n\u001b[0;32m----> 9\u001b[0m \u001b[43mBREAK\u001b[49m\n", "\u001b[0;31mNameError\u001b[0m: name 'BREAK' is not defined" ] } ], "source": [ "# df_slice.to_pickle(\n", "# \"/home/tony/Data/Preference/30b_v6/interesting_clips_v4_h_t_6_20250426_full_long_slice.pkl\"\n", "# )\n", "print(df_slice.shape)\n", "print(df_slice[\"task\"].value_counts())\n", "df_slice = df_slice[df_slice[\"task\"].isin([\"cover\", \"artist_consistency\"])].copy()\n", "print(df_slice.shape)\n", "print(df_slice[\"task\"].value_counts())\n", "BREAK" ] }, { "cell_type": "code", "execution_count": 37, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:51:22.836920Z", "iopub.status.busy": "2025-06-03T18:51:22.836662Z", "iopub.status.idle": "2025-06-03T18:51:29.217557Z", "shell.execute_reply": "2025-06-03T18:51:29.217200Z", "shell.execute_reply.started": "2025-06-03T18:51:22.836904Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "(201960, 168)\n", "task\n", " 193040\n", "upload_extend 4992\n", "extend 3928\n", "Name: count, dtype: int64\n" ] } ], "source": [ "df_13b_slice = pd.read_pickle(\n", " \"/home/tony/Data/Preference/13b_v32/interesting_clips_v4_h_s_32_20250412_20250501_full_long_slice.pkl\"\n", ")\n", "print(df_13b_slice.shape)\n", "print(df_13b_slice[\"task\"].value_counts())" ] }, { "cell_type": "code", "execution_count": 38, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:51:29.218356Z", "iopub.status.busy": "2025-06-03T18:51:29.218166Z", "iopub.status.idle": "2025-06-03T18:51:29.297846Z", "shell.execute_reply": "2025-06-03T18:51:29.297497Z", "shell.execute_reply.started": "2025-06-03T18:51:29.218343Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "(104826, 160)\n", "task\n", "cover 53704\n", "artist_consistency 51122\n", "Name: count, dtype: int64\n" ] } ], "source": [ "df_30b_slice = df_slice.copy()\n", "print(df_30b_slice.shape)\n", "print(df_30b_slice[\"task\"].value_counts())" ] }, { "cell_type": "code", "execution_count": 39, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:51:29.298429Z", "iopub.status.busy": "2025-06-03T18:51:29.298226Z", "iopub.status.idle": "2025-06-03T18:51:31.598782Z", "shell.execute_reply": "2025-06-03T18:51:31.598434Z", "shell.execute_reply.started": "2025-06-03T18:51:29.298416Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "(306786, 174)\n", "task\n", " 193040\n", "cover 53704\n", "artist_consistency 51122\n", "upload_extend 4992\n", "extend 3928\n", "Name: count, dtype: int64\n" ] } ], "source": [ "df_total = pd.concat([df_13b_slice, df_30b_slice])\n", "print(df_total.shape)\n", "print(df_total[\"task\"].value_counts())" ] }, { "cell_type": "code", "execution_count": 40, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:51:31.599367Z", "iopub.status.busy": "2025-06-03T18:51:31.599166Z", "iopub.status.idle": "2025-06-03T18:51:32.358937Z", "shell.execute_reply": "2025-06-03T18:51:32.358525Z", "shell.execute_reply.started": "2025-06-03T18:51:31.599353Z" } }, "outputs": [], "source": [ "df_slice = df_total.copy()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Need to kick out the ones has gpt prompt -- these are pairs with different text inputs" ] }, { "cell_type": "code", "execution_count": 41, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.932296Z", "start_time": "2024-05-16T13:59:41.932287Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:51:32.360028Z", "iopub.status.busy": "2025-06-03T18:51:32.359826Z", "iopub.status.idle": "2025-06-03T18:51:32.384119Z", "shell.execute_reply": "2025-06-03T18:51:32.383751Z", "shell.execute_reply.started": "2025-06-03T18:51:32.360015Z" } }, "outputs": [], "source": [ "# don't have continue at\n", "df_slice[\"request_id\"] = df_slice[\"request_id\"].astype(str)\n", "# df_slice[\"npz_path\"] = df_slice[\"npz_path\"].apply(lambda x: str(x).replace(\"_npz\", \"_npz/\"))\n", "# df_slice[df_slice[\"continue_at\"].isna()][\"request_id\"].nunique(), df_slice[\"request_id\"].nunique()" ] }, { "cell_type": "code", "execution_count": 42, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:51:32.384670Z", "iopub.status.busy": "2025-06-03T18:51:32.384492Z", "iopub.status.idle": "2025-06-03T18:51:33.187908Z", "shell.execute_reply": "2025-06-03T18:51:33.187547Z", "shell.execute_reply.started": "2025-06-03T18:51:32.384658Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "(306786, 174)\n", "(306786, 174)\n", "(306786, 174)\n" ] } ], "source": [ "print(df_slice.shape)\n", "df_slice = df_slice[df_slice[\"request_id\"].apply(lambda x: len(x) > 3)]\n", "print(df_slice.shape)\n", "# df_slice = df_slice[df_slice[\"is_pro_user\"]].copy()\n", "print(df_slice.shape)" ] }, { "cell_type": "code", "execution_count": 43, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.932966Z", "start_time": "2024-05-16T13:59:41.932957Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:51:33.188429Z", "iopub.status.busy": "2025-06-03T18:51:33.188303Z", "iopub.status.idle": "2025-06-03T18:51:33.273381Z", "shell.execute_reply": "2025-06-03T18:51:33.273015Z", "shell.execute_reply.started": "2025-06-03T18:51:33.188416Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "153393\n" ] } ], "source": [ "final_filtered_requests = df_slice[\"request_id\"].astype(str).unique()\n", "# final_filtered_requests = df_slice[df_slice[\"is_pro_user\"]][\"request_id\"].astype(str).unique()\n", "print(len(final_filtered_requests))" ] }, { "cell_type": "code", "execution_count": 44, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.933558Z", "start_time": "2024-05-16T13:59:41.933550Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:51:33.273958Z", "iopub.status.busy": "2025-06-03T18:51:33.273768Z", "iopub.status.idle": "2025-06-03T18:51:33.289902Z", "shell.execute_reply": "2025-06-03T18:51:33.289634Z", "shell.execute_reply.started": "2025-06-03T18:51:33.273945Z" } }, "outputs": [], "source": [ "# df_slice.to_csv(\"/home/tony/Data/Preference/30b_v2/interesting_clips_v4_t_3_20240902_slice.csv\", index=False)" ] }, { "cell_type": "code", "execution_count": 45, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:51:33.290361Z", "iopub.status.busy": "2025-06-03T18:51:33.290209Z", "iopub.status.idle": "2025-06-03T18:51:33.300501Z", "shell.execute_reply": "2025-06-03T18:51:33.300246Z", "shell.execute_reply.started": "2025-06-03T18:51:33.290349Z" } }, "outputs": [], "source": [ "# df_slice[\"continue_at\"] = -1" ] }, { "cell_type": "code", "execution_count": 46, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.934277Z", "start_time": "2024-05-16T13:59:41.934268Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:51:33.300975Z", "iopub.status.busy": "2025-06-03T18:51:33.300802Z", "iopub.status.idle": "2025-06-03T18:51:34.807160Z", "shell.execute_reply": "2025-06-03T18:51:34.806809Z", "shell.execute_reply.started": "2025-06-03T18:51:33.300964Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "151859 1534\n", "(303718, 174) (3068, 174)\n" ] } ], "source": [ "train_requests, val_requests = train_test_split(\n", " sorted(list(final_filtered_requests)), test_size=0.01, random_state=42\n", ")\n", "print(len(train_requests), len(val_requests))\n", "\n", "train_df = df_slice[df_slice[\"request_id\"].astype(str).isin(set(train_requests))].copy()\n", "val_df = df_slice[df_slice[\"request_id\"].astype(str).isin(set(val_requests))].copy()\n", "train_df = train_df.sort_values(by=[\"request_id\", \"preference\"])\n", "train_df = train_df # .reset_index()\n", "val_df = val_df.sort_values(by=[\"request_id\", \"preference\"])\n", "val_df = val_df # .reset_index()\n", "train_df = train_df.reset_index(drop=True)\n", "val_df = val_df.reset_index(drop=True)\n", "print(train_df.shape, val_df.shape)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Actually make" ] }, { "cell_type": "code", "execution_count": 47, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.935620Z", "start_time": "2024-05-16T13:59:41.935613Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:51:34.807744Z", "iopub.status.busy": "2025-06-03T18:51:34.807545Z", "iopub.status.idle": "2025-06-03T18:51:41.670198Z", "shell.execute_reply": "2025-06-03T18:51:41.669838Z", "shell.execute_reply.started": "2025-06-03T18:51:34.807731Z" } }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "100%|██████████████████████████████████████████████████████████████████████████████████████████████████████| 303718/303718 [00:06<00:00, 44362.36it/s]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "15,638 hours of 303718 clips, 18.982375 nodes, 3163.7291666666665 iters\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\n" ] } ], "source": [ "total_duration = 0\n", "for i, row in tqdm(train_df.iterrows(), total=len(train_df)):\n", " # we need to alternate between preference: neg, pos\n", " # print(i, row)\n", " try:\n", " assert row[\"preference\"] == (i % 2 == 1)\n", " total_duration += row[\"duration\"]\n", " except Exception as E:\n", " print(i, row)\n", " print(E)\n", " raise ValueError()\n", "\n", "print(\n", " f\"{round(total_duration / 60 / 60):,} hours of {train_df.shape[0]} clips, {train_df.shape[0] / 8 / 2 / 1000} nodes, {train_df.shape[0] / 8 / 2 / 6} iters\"\n", ")" ] }, { "cell_type": "code", "execution_count": 48, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.936268Z", "start_time": "2024-05-16T13:59:41.936260Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:51:41.670782Z", "iopub.status.busy": "2025-06-03T18:51:41.670586Z", "iopub.status.idle": "2025-06-03T18:52:25.994438Z", "shell.execute_reply": "2025-06-03T18:52:25.994082Z", "shell.execute_reply.started": "2025-06-03T18:51:41.670768Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "t_data_memmap is set to: 6016\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3068/3068 [00:44<00:00, 69.29it/s]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Total 3068 clips, 0 different prompts\n", "89 hours of False\n", "87 hours of True\n", "gen: 105.0 hours\n", "cover: 30.6 hours\n", "artist_consistency: 34.1 hours\n", "extend: 6.6 hours\n", "Done\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\n" ] } ], "source": [ "make_dataset(\n", " val_df, OUT_DATA_DIR, is_val=True, npz_dir=NPZ_DIR, t_data_memmap=N_TOKENS_AUDIO\n", ")" ] }, { "cell_type": "code", "execution_count": 49, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T18:52:25.994970Z", "iopub.status.busy": "2025-06-03T18:52:25.994821Z", "iopub.status.idle": "2025-06-03T18:52:26.781829Z", "shell.execute_reply": "2025-06-03T18:52:26.781430Z", "shell.execute_reply.started": "2025-06-03T18:52:25.994958Z" } }, "outputs": [], "source": [ "# test_npz = np.load(\"/app/suno/data/dpo/30b_npz/26d19085-18da-4701-af43-122684543891.npz\")\n", "# for k in test_npz.keys():\n", "# print(k)" ] }, { "cell_type": "code", "execution_count": 50, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.936964Z", "start_time": "2024-05-16T13:59:41.936957Z" }, "execution": { "iopub.execute_input": "2025-06-03T18:52:26.783432Z", "iopub.status.busy": "2025-06-03T18:52:26.783241Z", "iopub.status.idle": "2025-06-03T20:07:51.453039Z", "shell.execute_reply": "2025-06-03T20:07:51.452579Z", "shell.execute_reply.started": "2025-06-03T18:52:26.783420Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "t_data_memmap is set to: 6016\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ " 35%|████████████████████████████████████▍ | 105513/303718 [25:56<45:15, 72.99it/s]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "weird, /app/suno/data/dpo/30b_npz/c1150a74-6d63-43b1-b3e5-41d74a4b3c63.npz, with only v3.0\n", "105499, 'artist_arr is not a file in the archive', artist_consistency, /app/suno/data/dpo/30b_npz/c1150a74-6d63-43b1-b3e5-41d74a4b3c63.npz.\n", "WTF --> 105499, 'artist_arr is not a file in the archive', artist_consistency, /app/suno/data/dpo/30b_npz/c1150a74-6d63-43b1-b3e5-41d74a4b3c63.npz.\n", "WTF --> 105500, skip, preference: False, 937b20a3-a0fd-4d34-a0fc-7b86472dc271, task: .\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ " 61%|████████████████████████████████████████████████████████████████▎ | 186087/303718 [45:35<26:48, 73.12it/s]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "186074, 'history_arr is not a file in the archive', upload_extend, /app/suno/data/dpo/13b_s32_npz/cf0c29c4-267e-4aa3-a3f8-009876019300.npz.\n", "WTF --> 186075, skip, preference: True, 3cb56754-6b0a-4492-9bce-299dff540fe1, task: upload_extend.\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ " 70%|█████████████████████████████████████████████████████████████████████████▎ | 212013/303718 [51:53<21:37, 70.66it/s]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "211998, 'history_arr is not a file in the archive', upload_extend, /app/suno/data/dpo/13b_s32_npz/a538ccbe-a313-4acf-99c7-317806e9d062.npz.\n", "WTF --> 211999, skip, preference: True, 83054a12-5b17-42f2-b702-7d025c5e3fe9, task: upload_extend.\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "100%|███████████████████████████████████████████████████████████████████████████████████████████████████████| 303718/303718 [1:15:24<00:00, 67.13it/s]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Total 303712 clips, 30 different prompts\n", "8,807 hours of False\n", "8,683 hours of True\n", "gen: 10217.9 hours\n", "artist_consistency: 3355.6 hours\n", "extend: 497.6 hours\n", "cover: 3419.3 hours\n", "🚨 Error artist_consistency: 1\n", "🚨 Error upload_extend: 2\n", "Done\n" ] } ], "source": [ "make_dataset(\n", " train_df, OUT_DATA_DIR, is_val=False, npz_dir=NPZ_DIR, t_data_memmap=N_TOKENS_AUDIO\n", ")" ] }, { "cell_type": "markdown", "metadata": { "ExecuteTime": { "end_time": "2024-01-29T19:46:47.549860Z", "start_time": "2024-01-29T19:46:47.548015Z" } }, "source": [ "# Validation" ] }, { "cell_type": "code", "execution_count": 51, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.937879Z", "start_time": "2024-05-16T13:59:41.937870Z" }, "execution": { "iopub.execute_input": "2025-06-03T20:07:51.453690Z", "iopub.status.busy": "2025-06-03T20:07:51.453543Z", "iopub.status.idle": "2025-06-03T20:07:52.489344Z", "shell.execute_reply": "2025-06-03T20:07:52.488848Z", "shell.execute_reply.started": "2025-06-03T20:07:51.453675Z" } }, "outputs": [], "source": [ "# verify\n", "mm = np.memmap(os.path.join(OUT_DATA_DIR, f\"data_val.bin\"), dtype=np.uint16, mode=\"r\")\n", "test_metas = read_jsonl(os.path.join(OUT_DATA_DIR, f\"meta_val.jsonl\"))\n", "test_info = read_json(os.path.join(OUT_DATA_DIR, f\"info_val.json\"))\n", "mm = mm.reshape(-1, N_TOKENS_AUDIO, 13)\n", "assert len(mm) == len(test_metas)\n", "assert mm[:100, :, 0].min() >= 0\n", "assert mm[:100, :, 0].max() <= 4000\n", "assert mm[:100, :, 1:].min() >= 0\n", "assert mm[:100, :, 1:].max() <= 2048" ] }, { "cell_type": "code", "execution_count": 52, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T20:07:52.489976Z", "iopub.status.busy": "2025-06-03T20:07:52.489841Z", "iopub.status.idle": "2025-06-03T20:07:52.508923Z", "shell.execute_reply": "2025-06-03T20:07:52.508626Z", "shell.execute_reply.started": "2025-06-03T20:07:52.489962Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Counter({None: 1968, 'artist_consistency': 512, 'cover': 472, 'extend': 116})\n" ] } ], "source": [ "task_counts = Counter()\n", "for test_meta in test_metas:\n", " task_counts[test_meta.get(\"task\")] += 1\n", "print(task_counts)" ] }, { "cell_type": "code", "execution_count": 53, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.938629Z", "start_time": "2024-05-16T13:59:41.938621Z" }, "execution": { "iopub.execute_input": "2025-06-03T20:07:52.509393Z", "iopub.status.busy": "2025-06-03T20:07:52.509280Z", "iopub.status.idle": "2025-06-03T20:07:52.520565Z", "shell.execute_reply": "2025-06-03T20:07:52.520277Z", "shell.execute_reply.started": "2025-06-03T20:07:52.509381Z" } }, "outputs": [], "source": [ "# # randomly listen to some stuff\n", "# from suno_utils.tasks.dac_2c_12cb import preload_models as preload_codec_models\n", "# from suno_utils.tasks.dac_2c_12cb import (\n", "# encode as codec_encode,\n", "# decode_stream_to_full_audio as codec_decode,\n", "# EMBEDDING_RATE as CODEC_EMBEDDING_RATE,\n", "# decode as decode\n", "# )\n", "# os.environ[\"CUDA_VISIBLE_DEVICES\"] = \"0\"\n", "# _ = preload_codec_models(\"/app/suno/data/dpo/models/dac_2c_25x12.pt\", device=\"cuda\")\n", "# assert len(test_metas) == len(mm)\n", "# idx_list = list(range(len(test_metas)))\n", "# # random.shuffle(idx_list)\n", "# # idx_list = [idx for idx in idx_list if \"text\" in test_metas[idx]]\n", "# print(len(mm))" ] }, { "cell_type": "code", "execution_count": 54, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.939205Z", "start_time": "2024-05-16T13:59:41.939198Z" }, "execution": { "iopub.execute_input": "2025-06-03T20:07:52.521024Z", "iopub.status.busy": "2025-06-03T20:07:52.520913Z", "iopub.status.idle": "2025-06-03T20:07:52.531794Z", "shell.execute_reply": "2025-06-03T20:07:52.531502Z", "shell.execute_reply.started": "2025-06-03T20:07:52.521013Z" } }, "outputs": [], "source": [ "# import random\n", "# idx = random.choice(test_info[\"perference_0\"][\"idx_list\"])\n", "# assert \"original_duration_s\" in test_metas[idx]\n", "# # positive index should be shifted by 1\n", "# pos_idx = idx + 1\n", "# print(\n", "# \"tags:\",\n", "# test_metas[idx].get(\"tags\") == test_metas[pos_idx].get(\"tags\"),\n", "# test_metas[idx].get(\"tags\"),\n", "# )\n", "# arr = mm[idx, 1:].copy().astype(np.int16)[:, 1:]\n", "# pos_arr = mm[pos_idx, 1:].copy().astype(np.int16)[:, 1:]\n", "# pad_idx_arr = np.where(arr == COARSE_PAD_TOKEN)[0]\n", "# if len(pad_idx_arr) > 0:\n", "# arr = arr[: pad_idx_arr[0], :]\n", "# pos_pad_idx_arr = np.where(pos_arr == COARSE_PAD_TOKEN)[0]\n", "# if len(pos_pad_idx_arr) > 0:\n", "# pos_arr = pos_arr[: pos_pad_idx_arr[0], :]\n", "# a = decode(arr)\n", "# print(\"\\n negative example \\n\", test_metas[idx])\n", "# a.play(compress=False)\n", "# pos_a = decode(pos_arr)\n", "# print(\"\\n positive example \\n\", test_metas[pos_idx])\n", "# pos_a.play(compress=False)\n", "# print(\n", "# \"text:\",\n", "# test_metas[idx].get(\"text\") == test_metas[pos_idx].get(\"text\"),\n", "# test_metas[idx].get(\"text\"),\n", "# )" ] }, { "cell_type": "code", "execution_count": 55, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.939977Z", "start_time": "2024-05-16T13:59:41.939969Z" }, "execution": { "iopub.execute_input": "2025-06-03T20:07:52.532238Z", "iopub.status.busy": "2025-06-03T20:07:52.532130Z", "iopub.status.idle": "2025-06-03T20:07:52.542621Z", "shell.execute_reply": "2025-06-03T20:07:52.542340Z", "shell.execute_reply.started": "2025-06-03T20:07:52.532227Z" } }, "outputs": [], "source": [ "# val_df[val_df[\"tags\"] == 'a vibrant blend of experimental jazz fusion, drum-and-bass and swagger fuzzed-out guitars']" ] }, { "cell_type": "code", "execution_count": 56, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.940610Z", "start_time": "2024-05-16T13:59:41.940603Z" }, "execution": { "iopub.execute_input": "2025-06-03T20:07:52.543056Z", "iopub.status.busy": "2025-06-03T20:07:52.542950Z", "iopub.status.idle": "2025-06-03T20:07:52.553203Z", "shell.execute_reply": "2025-06-03T20:07:52.552918Z", "shell.execute_reply.started": "2025-06-03T20:07:52.543045Z" } }, "outputs": [], "source": [ "# from collections import Counter\n", "# c = Counter()\n", "# for _, row in df_slice.iterrows():\n", "# # print(row[\"metadata\"])\n", "# for k in ast.literal_eval(row[\"metadata\"]).keys():\n", "# c[k] += 1" ] }, { "cell_type": "code", "execution_count": 57, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.941167Z", "start_time": "2024-05-16T13:59:41.941159Z" }, "execution": { "iopub.execute_input": "2025-06-03T20:07:52.553652Z", "iopub.status.busy": "2025-06-03T20:07:52.553542Z", "iopub.status.idle": "2025-06-03T20:07:52.564230Z", "shell.execute_reply": "2025-06-03T20:07:52.563947Z", "shell.execute_reply.started": "2025-06-03T20:07:52.553641Z" } }, "outputs": [], "source": [ "# original_npz_path = f\"/app/suno/data/dpo/7b_npz/{test_metas[idx]['id']}.npz\"\n", "# original_npz_path = \"/app/suno/data/dpo/7b_npz/729c3011-f672-4ccd-8d82-1cbf2b52ff69.npz\"\n", "# original_arr = np.load(original_npz_path)[\"v2_raw\"]" ] }, { "cell_type": "code", "execution_count": 58, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.941801Z", "start_time": "2024-05-16T13:59:41.941793Z" }, "execution": { "iopub.execute_input": "2025-06-03T20:07:52.564680Z", "iopub.status.busy": "2025-06-03T20:07:52.564572Z", "iopub.status.idle": "2025-06-03T20:07:52.579004Z", "shell.execute_reply": "2025-06-03T20:07:52.578725Z", "shell.execute_reply.started": "2025-06-03T20:07:52.564669Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "1534 0\n" ] } ], "source": [ "def validation_on_metas(input_metas):\n", " total_bad = 0\n", " total_good = 0\n", " for idx in range(len(input_metas)):\n", " if idx % 2 == 0:\n", " pos_idx = idx + 1\n", " if input_metas[idx].get(\"tags\") != input_metas[pos_idx].get(\"tags\"):\n", " # print(test_metas[idx].get(\"text\") == test_metas[pos_idx].get(\"text\"), test_metas[idx].get(\"tags\"), test_metas[pos_idx].get(\"tags\"))\n", " total_bad += 1\n", " else:\n", " total_good += 1\n", " print(total_good, total_bad)\n", " return\n", "\n", "\n", "validation_on_metas(test_metas)" ] }, { "cell_type": "code", "execution_count": 59, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.942520Z", "start_time": "2024-05-16T13:59:41.942511Z" }, "execution": { "iopub.execute_input": "2025-06-03T20:07:52.579456Z", "iopub.status.busy": "2025-06-03T20:07:52.579345Z", "iopub.status.idle": "2025-06-03T20:07:52.616801Z", "shell.execute_reply": "2025-06-03T20:07:52.616424Z", "shell.execute_reply.started": "2025-06-03T20:07:52.579444Z" } }, "outputs": [], "source": [ "train_info = read_json(os.path.join(OUT_DATA_DIR, f\"info_tr.json\"))" ] }, { "cell_type": "code", "execution_count": 60, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.943072Z", "start_time": "2024-05-16T13:59:41.943065Z" }, "execution": { "iopub.execute_input": "2025-06-03T20:07:52.617329Z", "iopub.status.busy": "2025-06-03T20:07:52.617210Z", "iopub.status.idle": "2025-06-03T20:07:52.644453Z", "shell.execute_reply": "2025-06-03T20:07:52.644091Z", "shell.execute_reply.started": "2025-06-03T20:07:52.617317Z" } }, "outputs": [], "source": [ "n_neg_tr = train_info[\"perference_0\"][\"idx_list\"]\n", "n_pos_tr = train_info[\"perference_1\"][\"idx_list\"]\n", "assert len(n_pos_tr) == len(n_neg_tr)\n", "# make sure they are offset by 1 and exactly 1\n", "for i, j in zip(n_neg_tr, n_pos_tr):\n", " assert i == j - 1" ] }, { "cell_type": "code", "execution_count": 61, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.944246Z", "start_time": "2024-05-16T13:59:41.944237Z" }, "execution": { "iopub.execute_input": "2025-06-03T20:07:52.644970Z", "iopub.status.busy": "2025-06-03T20:07:52.644851Z", "iopub.status.idle": "2025-06-03T20:07:52.658030Z", "shell.execute_reply": "2025-06-03T20:07:52.657745Z", "shell.execute_reply.started": "2025-06-03T20:07:52.644957Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "total samples 303712 (303718, 174)\n" ] } ], "source": [ "total_iters = len(n_neg_tr) + len(n_pos_tr)\n", "print(\"total samples\", total_iters, train_df.shape)" ] }, { "cell_type": "code", "execution_count": 62, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.945249Z", "start_time": "2024-05-16T13:59:41.945241Z" }, "execution": { "iopub.execute_input": "2025-06-03T20:07:52.658485Z", "iopub.status.busy": "2025-06-03T20:07:52.658373Z", "iopub.status.idle": "2025-06-03T20:07:52.669600Z", "shell.execute_reply": "2025-06-03T20:07:52.669327Z", "shell.execute_reply.started": "2025-06-03T20:07:52.658474Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "1 epoch per batch 4, total 4745.5\n" ] } ], "source": [ "print(\"1 epoch per batch 4, total\", total_iters / 8 / 2 / 4)" ] }, { "cell_type": "code", "execution_count": 63, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.945972Z", "start_time": "2024-05-16T13:59:41.945964Z" }, "execution": { "iopub.execute_input": "2025-06-03T20:07:52.670039Z", "iopub.status.busy": "2025-06-03T20:07:52.669932Z", "iopub.status.idle": "2025-06-03T20:07:52.680339Z", "shell.execute_reply": "2025-06-03T20:07:52.680059Z", "shell.execute_reply.started": "2025-06-03T20:07:52.670027Z" } }, "outputs": [], "source": [ "# !cd /home/tony/Work/tony/slurm/sem_4p5_dpo && sbatch sbatch_ipo_4p5" ] }, { "cell_type": "code", "execution_count": 64, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T20:07:52.680780Z", "iopub.status.busy": "2025-06-03T20:07:52.680672Z", "iopub.status.idle": "2025-06-03T20:07:52.702362Z", "shell.execute_reply": "2025-06-03T20:07:52.702062Z", "shell.execute_reply.started": "2025-06-03T20:07:52.680769Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Cache kept!\n" ] } ], "source": [ "import shutil\n", "\n", "# Basic file copy\n", "shutil.copy(\n", " \"/home/tony/Work/tony/Preference/make_dataset_auk_mixed_2.ipynb\",\n", " os.path.join(OUT_DATA_DIR, \"make_dataset.ipynb\"),\n", ")\n", "print(\"Cache kept!\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# some gymathtics loading prev data" ] }, { "cell_type": "code", "execution_count": 65, "metadata": { "ExecuteTime": { "end_time": "2024-05-16T13:59:41.946562Z", "start_time": "2024-05-16T13:59:41.946555Z" }, "execution": { "iopub.execute_input": "2025-06-03T20:07:52.702825Z", "iopub.status.busy": "2025-06-03T20:07:52.702715Z", "iopub.status.idle": "2025-06-03T20:07:52.714040Z", "shell.execute_reply": "2025-06-03T20:07:52.713754Z", "shell.execute_reply.started": "2025-06-03T20:07:52.702814Z" } }, "outputs": [], "source": [ "# prev_v3_data = \"/app/suno/data/dpo/7v_v20_full/\"\n", "\n", "# test_val_metas = read_jsonl(os.path.join(prev_v3_data, f\"meta_val.jsonl\"))\n", "# test_tr_metas = read_jsonl(os.path.join(prev_v3_data, f\"meta_tr.jsonl\"))\n", "\n", "# all_ids = set()\n", "# for meta in test_val_metas:\n", "# all_ids.add(meta[\"id\"])\n", "# for meta in test_tr_metas:\n", "# all_ids.add(meta[\"id\"])\n", "# print(len(all_ids), len(test_val_metas) + len(test_tr_metas))\n", "\n", "# all_ids = list(all_ids)\n", "# with open(\"/home/tony/Data/Preference/7b_v2/7v_v20_full_recut_id.json\", \"w\") as fp:\n", "# json.dump(all_ids, fp)" ] }, { "cell_type": "code", "execution_count": 66, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T20:07:52.714481Z", "iopub.status.busy": "2025-06-03T20:07:52.714374Z", "iopub.status.idle": "2025-06-03T20:07:52.725047Z", "shell.execute_reply": "2025-06-03T20:07:52.724758Z", "shell.execute_reply.started": "2025-06-03T20:07:52.714470Z" } }, "outputs": [], "source": [ "# x_data = train_df[train_df[\"preference\"]][\"similarity\"]\n", "# y_data = train_df[~train_df[\"preference\"]][\"similarity\"]\n", "# from matplotlib.colors import LogNorm\n", "\n", "# fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(24, 10))\n", "\n", "# # 2D Histogram\n", "# h = ax1.hist2d(\n", "# x_data,\n", "# y_data,\n", "# bins=(50, 50),\n", "# cmap=\"coolwarm\",\n", "# range=[[0, 1], [0, 1]],\n", "# norm=LogNorm(),\n", "# )\n", "\n", "# ax1.set_xlabel(\"Semantic Distance (Preferred)\")\n", "# ax1.set_ylabel(\"Semantic Distance (Non-Preferred)\")\n", "# ax1.set_title(\n", "# \"2D Histogram of Semantic Distances: Preferred vs Non-Preferred (Log Scale)\"\n", "# )\n", "\n", "# cbar1 = plt.colorbar(h[3], ax=ax1)\n", "# cbar1.set_label(\"Number of Request IDs (Log Scale)\")\n", "\n", "# # Scatter plot\n", "# ax2.scatter(x_data, y_data, alpha=0.1, s=1)\n", "# ax2.set_xlabel(\"Semantic Distance (Preferred)\")\n", "# ax2.set_ylabel(\"Semantic Distance (Non-Preferred)\")\n", "# ax2.set_title(\"Scatter Plot of Semantic Distances: Preferred vs Non-Preferred\")\n", "# ax2.set_xlim(0, 1)\n", "# ax2.set_ylim(0, 1)\n", "\n", "# plt.tight_layout()\n", "# plt.show()" ] }, { "cell_type": "code", "execution_count": 67, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T20:07:52.725490Z", "iopub.status.busy": "2025-06-03T20:07:52.725382Z", "iopub.status.idle": "2025-06-03T20:07:52.735771Z", "shell.execute_reply": "2025-06-03T20:07:52.735493Z", "shell.execute_reply.started": "2025-06-03T20:07:52.725478Z" } }, "outputs": [], "source": [ "# train_metas = read_jsonl(os.path.join(OUT_DATA_DIR, f\"meta_tr.jsonl\"))" ] }, { "cell_type": "code", "execution_count": 68, "metadata": { "execution": { "iopub.execute_input": "2025-06-03T20:07:52.736200Z", "iopub.status.busy": "2025-06-03T20:07:52.736093Z", "iopub.status.idle": "2025-06-03T20:07:52.747156Z", "shell.execute_reply": "2025-06-03T20:07:52.746905Z", "shell.execute_reply.started": "2025-06-03T20:07:52.736189Z" } }, "outputs": [ { "data": { "text/plain": [ "dict_keys(['perference_0', 'perference_1'])" ] }, "execution_count": 68, "metadata": {}, "output_type": "execute_result" } ], "source": [ "train_info.keys()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "suno_env_dev", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.10.15" }, "toc": { "base_numbering": 1, "nav_menu": {}, "number_sections": true, "sideBar": true, "skip_h1_title": false, "title_cell": "Table of Contents", "title_sidebar": "Contents", "toc_cell": false, "toc_position": {}, "toc_section_display": true, "toc_window_display": false } }, "nbformat": 4, "nbformat_minor": 4 }