mirror of
https://github.com/ltcptgeneral/cse151b-final-project.git
synced 2025-10-22 18:49:21 +00:00
Compare commits
11 Commits
ethan-test
...
f40301cac9
| Author | SHA1 | Date | |
|---|---|---|---|
|
f40301cac9 | ||
|
fc197acb6e | ||
|
|
e799c14ece | ||
|
|
bbe9a1891c | ||
|
|
9172326013 | ||
|
|
4836be8121 | ||
|
|
5672169073 | ||
|
|
5ec123e0f1 | ||
|
|
e9622b6f68 | ||
|
83e81722d2 | ||
|
|
320f2f81b7 |
5
.gitignore
vendored
5
.gitignore
vendored
@@ -1 +1,6 @@
|
|||||||
**/data/*
|
**/data/*
|
||||||
|
**/*.zip
|
||||||
|
**/__pycache__
|
||||||
|
/env
|
||||||
|
**/runs/*
|
||||||
|
**/wandb/*
|
||||||
2996
dqn_letter_gssr.ipynb
Normal file
2996
dqn_letter_gssr.ipynb
Normal file
File diff suppressed because it is too large
Load Diff
338
dqn_wordle.ipynb
Normal file
338
dqn_wordle.ipynb
Normal file
@@ -0,0 +1,338 @@
|
|||||||
|
{
|
||||||
|
"cells": [
|
||||||
|
{
|
||||||
|
"cell_type": "code",
|
||||||
|
"execution_count": 1,
|
||||||
|
"metadata": {},
|
||||||
|
"outputs": [],
|
||||||
|
"source": [
|
||||||
|
"import gym\n",
|
||||||
|
"import gym_wordle\n",
|
||||||
|
"from stable_baselines3 import DQN, PPO, common\n",
|
||||||
|
"import numpy as np\n",
|
||||||
|
"import tqdm"
|
||||||
|
]
|
||||||
|
},
|
||||||
|
{
|
||||||
|
"cell_type": "code",
|
||||||
|
"execution_count": 2,
|
||||||
|
"metadata": {},
|
||||||
|
"outputs": [
|
||||||
|
{
|
||||||
|
"name": "stdout",
|
||||||
|
"output_type": "stream",
|
||||||
|
"text": [
|
||||||
|
"<Monitor<WordleEnv instance>>\n"
|
||||||
|
]
|
||||||
|
}
|
||||||
|
],
|
||||||
|
"source": [
|
||||||
|
"env = gym_wordle.wordle.WordleEnv()\n",
|
||||||
|
"env = common.monitor.Monitor(env)\n",
|
||||||
|
"\n",
|
||||||
|
"print(env)"
|
||||||
|
]
|
||||||
|
},
|
||||||
|
{
|
||||||
|
"cell_type": "code",
|
||||||
|
"execution_count": 3,
|
||||||
|
"metadata": {},
|
||||||
|
"outputs": [
|
||||||
|
{
|
||||||
|
"name": "stdout",
|
||||||
|
"output_type": "stream",
|
||||||
|
"text": [
|
||||||
|
"Using cuda device\n",
|
||||||
|
"Wrapping the env in a DummyVecEnv.\n"
|
||||||
|
]
|
||||||
|
},
|
||||||
|
{
|
||||||
|
"data": {
|
||||||
|
"application/vnd.jupyter.widget-view+json": {
|
||||||
|
"model_id": "6921a0721569456abf5bceac7e7b6b34",
|
||||||
|
"version_major": 2,
|
||||||
|
"version_minor": 0
|
||||||
|
},
|
||||||
|
"text/plain": [
|
||||||
|
"Output()"
|
||||||
|
]
|
||||||
|
},
|
||||||
|
"metadata": {},
|
||||||
|
"output_type": "display_data"
|
||||||
|
},
|
||||||
|
{
|
||||||
|
"name": "stdout",
|
||||||
|
"output_type": "stream",
|
||||||
|
"text": [
|
||||||
|
"----------------------------------\n",
|
||||||
|
"| rollout/ | |\n",
|
||||||
|
"| ep_len_mean | 4.97 |\n",
|
||||||
|
"| ep_rew_mean | -63.8 |\n",
|
||||||
|
"| exploration_rate | 0.05 |\n",
|
||||||
|
"| time/ | |\n",
|
||||||
|
"| episodes | 10000 |\n",
|
||||||
|
"| fps | 1628 |\n",
|
||||||
|
"| time_elapsed | 30 |\n",
|
||||||
|
"| total_timesteps | 49995 |\n",
|
||||||
|
"----------------------------------\n"
|
||||||
|
]
|
||||||
|
},
|
||||||
|
{
|
||||||
|
"name": "stdout",
|
||||||
|
"output_type": "stream",
|
||||||
|
"text": [
|
||||||
|
"----------------------------------\n",
|
||||||
|
"| rollout/ | |\n",
|
||||||
|
"| ep_len_mean | 5 |\n",
|
||||||
|
"| ep_rew_mean | -70.5 |\n",
|
||||||
|
"| exploration_rate | 0.05 |\n",
|
||||||
|
"| time/ | |\n",
|
||||||
|
"| episodes | 20000 |\n",
|
||||||
|
"| fps | 662 |\n",
|
||||||
|
"| time_elapsed | 150 |\n",
|
||||||
|
"| total_timesteps | 99992 |\n",
|
||||||
|
"| train/ | |\n",
|
||||||
|
"| learning_rate | 0.0001 |\n",
|
||||||
|
"| loss | 11.7 |\n",
|
||||||
|
"| n_updates | 12497 |\n",
|
||||||
|
"----------------------------------\n"
|
||||||
|
]
|
||||||
|
},
|
||||||
|
{
|
||||||
|
"data": {
|
||||||
|
"text/html": [
|
||||||
|
"<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"></pre>\n"
|
||||||
|
],
|
||||||
|
"text/plain": []
|
||||||
|
},
|
||||||
|
"metadata": {},
|
||||||
|
"output_type": "display_data"
|
||||||
|
},
|
||||||
|
{
|
||||||
|
"data": {
|
||||||
|
"text/html": [
|
||||||
|
"<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">\n",
|
||||||
|
"</pre>\n"
|
||||||
|
],
|
||||||
|
"text/plain": [
|
||||||
|
"\n"
|
||||||
|
]
|
||||||
|
},
|
||||||
|
"metadata": {},
|
||||||
|
"output_type": "display_data"
|
||||||
|
},
|
||||||
|
{
|
||||||
|
"data": {
|
||||||
|
"text/plain": [
|
||||||
|
"<stable_baselines3.dqn.dqn.DQN at 0x1bfd6cc0210>"
|
||||||
|
]
|
||||||
|
},
|
||||||
|
"execution_count": 3,
|
||||||
|
"metadata": {},
|
||||||
|
"output_type": "execute_result"
|
||||||
|
}
|
||||||
|
],
|
||||||
|
"source": [
|
||||||
|
"total_timesteps = 100_000\n",
|
||||||
|
"model = DQN(\"MlpPolicy\", env, verbose=1, device='cuda')\n",
|
||||||
|
"model.learn(total_timesteps=total_timesteps, log_interval=10_000, progress_bar=True)"
|
||||||
|
]
|
||||||
|
},
|
||||||
|
{
|
||||||
|
"cell_type": "code",
|
||||||
|
"execution_count": 4,
|
||||||
|
"metadata": {},
|
||||||
|
"outputs": [],
|
||||||
|
"source": [
|
||||||
|
"model.save(\"dqn_new_state\")"
|
||||||
|
]
|
||||||
|
},
|
||||||
|
{
|
||||||
|
"cell_type": "code",
|
||||||
|
"execution_count": 5,
|
||||||
|
"metadata": {},
|
||||||
|
"outputs": [
|
||||||
|
{
|
||||||
|
"name": "stderr",
|
||||||
|
"output_type": "stream",
|
||||||
|
"text": [
|
||||||
|
"c:\\Repository\\cse151b-final-project\\env\\Lib\\site-packages\\stable_baselines3\\common\\save_util.py:166: UserWarning: Could not deserialize object lr_schedule. Consider using `custom_objects` argument to replace this object.\n",
|
||||||
|
"Exception: code() argument 13 must be str, not int\n",
|
||||||
|
" warnings.warn(\n",
|
||||||
|
"c:\\Repository\\cse151b-final-project\\env\\Lib\\site-packages\\stable_baselines3\\common\\save_util.py:166: UserWarning: Could not deserialize object exploration_schedule. Consider using `custom_objects` argument to replace this object.\n",
|
||||||
|
"Exception: code() argument 13 must be str, not int\n",
|
||||||
|
" warnings.warn(\n"
|
||||||
|
]
|
||||||
|
}
|
||||||
|
],
|
||||||
|
"source": [
|
||||||
|
"# model = DQN.load(\"dqn_wordle\")"
|
||||||
|
]
|
||||||
|
},
|
||||||
|
{
|
||||||
|
"cell_type": "code",
|
||||||
|
"execution_count": 7,
|
||||||
|
"metadata": {},
|
||||||
|
"outputs": [
|
||||||
|
{
|
||||||
|
"name": "stdout",
|
||||||
|
"output_type": "stream",
|
||||||
|
"text": [
|
||||||
|
"[1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1. 0. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1. 0. 1.\n",
|
||||||
|
" 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1.\n",
|
||||||
|
" 0. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 0. 1. 0. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 0. 1. 0. 1. 1. 1. 1. 1. 0. 0. 0. 0. 0. 1. 1. 0. 0. 0. 0. 0. 0. 0.\n",
|
||||||
|
" 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
|
||||||
|
"[1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1.\n",
|
||||||
|
" 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 0. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 0. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 1. 1. 0. 1. 1. 1. 1. 1. 0. 0. 0. 0. 0. 1. 1. 0. 0. 0. 0. 0. 0. 0.\n",
|
||||||
|
" 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
|
||||||
|
"[1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 0. 1. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 0. 1. 1.\n",
|
||||||
|
" 1. 1. 1. 1. 1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 0.\n",
|
||||||
|
" 1. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 0. 0. 1. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 0. 0. 1. 1. 1. 1. 1. 1. 0. 0. 0. 0. 0. 0. 1. 0. 0. 0. 0. 0. 0. 0.\n",
|
||||||
|
" 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
|
||||||
|
"[1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 0. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 0. 1.\n",
|
||||||
|
" 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0.\n",
|
||||||
|
" 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 1. 0. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 1. 0. 0. 1. 1. 1. 1. 1. 0. 0. 0. 0. 0. 0. 1. 0. 0. 0. 0. 0. 0. 0.\n",
|
||||||
|
" 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
|
||||||
|
"[1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1. 1.\n",
|
||||||
|
" 1. 1. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.\n",
|
||||||
|
" 1. 0. 0. 0. 0. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 0. 0. 0. 0. 0. 1. 1. 0. 0. 0. 0. 0. 0. 0.\n",
|
||||||
|
" 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 1. 0. 0.]\n",
|
||||||
|
"[1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 0. 0. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 0. 0. 1.\n",
|
||||||
|
" 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 0.\n",
|
||||||
|
" 0. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 0. 0. 0. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 0. 0. 0. 1. 1. 1. 1. 1. 0. 0. 0. 0. 0. 1. 1. 0. 0. 0. 0. 0. 0. 0.\n",
|
||||||
|
" 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
|
||||||
|
"[1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1. 1.\n",
|
||||||
|
" 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0.\n",
|
||||||
|
" 1. 1. 1. 1. 1. 1. 0. 0. 0. 0. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.\n",
|
||||||
|
" 0. 0. 0. 0. 0. 0. 0. 0. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.\n",
|
||||||
|
" 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 1. 0.]\n",
|
||||||
|
"[1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 0. 0. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 0. 0. 1.\n",
|
||||||
|
" 1. 1. 1. 1. 1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 0.\n",
|
||||||
|
" 0. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 0. 0. 0. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 0. 0. 0. 1. 1. 1. 1. 1. 0. 0. 0. 0. 0. 0. 1. 0. 0. 0. 0. 0. 0. 0.\n",
|
||||||
|
" 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
|
||||||
|
"[1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 0. 0. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 0. 0. 1.\n",
|
||||||
|
" 1. 1. 1. 1. 1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 0.\n",
|
||||||
|
" 0. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 0. 0. 0. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 0. 0. 0. 1. 1. 1. 1. 1. 0. 0. 0. 0. 1. 0. 1. 0. 0. 0. 0. 0. 0. 0.\n",
|
||||||
|
" 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
|
||||||
|
"[1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 1. 1. 1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1.\n",
|
||||||
|
" 1. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 0. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.\n",
|
||||||
|
" 1. 1. 0. 1. 1. 1. 1. 1. 1. 1. 0. 0. 0. 0. 0. 0. 1. 0. 0. 0. 0. 0. 0. 0.\n",
|
||||||
|
" 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
|
||||||
|
"0\n"
|
||||||
|
]
|
||||||
|
}
|
||||||
|
],
|
||||||
|
"source": [
|
||||||
|
"env = gym_wordle.wordle.WordleEnv()\n",
|
||||||
|
"\n",
|
||||||
|
"for i in range(1000):\n",
|
||||||
|
" \n",
|
||||||
|
" state, info = env.reset()\n",
|
||||||
|
"\n",
|
||||||
|
" done = False\n",
|
||||||
|
"\n",
|
||||||
|
" wins = 0\n",
|
||||||
|
"\n",
|
||||||
|
" while not done:\n",
|
||||||
|
"\n",
|
||||||
|
" action, _states = model.predict(state, deterministic=True)\n",
|
||||||
|
"\n",
|
||||||
|
" state, reward, done, truncated, info = env.step(action)\n",
|
||||||
|
"\n",
|
||||||
|
" print(state)\n",
|
||||||
|
" if info[\"correct\"]:\n",
|
||||||
|
" wins += 1\n",
|
||||||
|
"\n",
|
||||||
|
"print(wins)"
|
||||||
|
]
|
||||||
|
},
|
||||||
|
{
|
||||||
|
"cell_type": "code",
|
||||||
|
"execution_count": 6,
|
||||||
|
"metadata": {},
|
||||||
|
"outputs": [
|
||||||
|
{
|
||||||
|
"data": {
|
||||||
|
"text/plain": [
|
||||||
|
"(array([1., 0., 1., 1., 1., 0., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n",
|
||||||
|
" 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 0., 1., 1., 1., 0., 1., 1.,\n",
|
||||||
|
" 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n",
|
||||||
|
" 1., 1., 0., 1., 1., 1., 0., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n",
|
||||||
|
" 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 0., 1., 1., 1., 0., 1.,\n",
|
||||||
|
" 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n",
|
||||||
|
" 1., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
|
||||||
|
" 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
|
||||||
|
" 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
|
||||||
|
" 0., 0., 0., 0., 0., 0., 0., 1.]),\n",
|
||||||
|
" -50)"
|
||||||
|
]
|
||||||
|
},
|
||||||
|
"execution_count": 6,
|
||||||
|
"metadata": {},
|
||||||
|
"output_type": "execute_result"
|
||||||
|
}
|
||||||
|
],
|
||||||
|
"source": [
|
||||||
|
"state, reward"
|
||||||
|
]
|
||||||
|
},
|
||||||
|
{
|
||||||
|
"cell_type": "code",
|
||||||
|
"execution_count": null,
|
||||||
|
"metadata": {},
|
||||||
|
"outputs": [],
|
||||||
|
"source": []
|
||||||
|
}
|
||||||
|
],
|
||||||
|
"metadata": {
|
||||||
|
"kernelspec": {
|
||||||
|
"display_name": "Python 3 (ipykernel)",
|
||||||
|
"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": "3.11.5"
|
||||||
|
}
|
||||||
|
},
|
||||||
|
"nbformat": 4,
|
||||||
|
"nbformat_minor": 2
|
||||||
|
}
|
||||||
129
eric_wordle/.gitignore
vendored
Normal file
129
eric_wordle/.gitignore
vendored
Normal file
@@ -0,0 +1,129 @@
|
|||||||
|
# Byte-compiled / optimized / DLL files
|
||||||
|
__pycache__/
|
||||||
|
*.py[cod]
|
||||||
|
*$py.class
|
||||||
|
|
||||||
|
# C extensions
|
||||||
|
*.so
|
||||||
|
|
||||||
|
# Distribution / packaging
|
||||||
|
.Python
|
||||||
|
build/
|
||||||
|
develop-eggs/
|
||||||
|
dist/
|
||||||
|
downloads/
|
||||||
|
eggs/
|
||||||
|
.eggs/
|
||||||
|
lib/
|
||||||
|
lib64/
|
||||||
|
parts/
|
||||||
|
sdist/
|
||||||
|
var/
|
||||||
|
wheels/
|
||||||
|
pip-wheel-metadata/
|
||||||
|
share/python-wheels/
|
||||||
|
*.egg-info/
|
||||||
|
.installed.cfg
|
||||||
|
*.egg
|
||||||
|
MANIFEST
|
||||||
|
|
||||||
|
# PyInstaller
|
||||||
|
# Usually these files are written by a python script from a template
|
||||||
|
# before PyInstaller builds the exe, so as to inject date/other infos into it.
|
||||||
|
*.manifest
|
||||||
|
*.spec
|
||||||
|
|
||||||
|
# Installer logs
|
||||||
|
pip-log.txt
|
||||||
|
pip-delete-this-directory.txt
|
||||||
|
|
||||||
|
# Unit test / coverage reports
|
||||||
|
htmlcov/
|
||||||
|
.tox/
|
||||||
|
.nox/
|
||||||
|
.coverage
|
||||||
|
.coverage.*
|
||||||
|
.cache
|
||||||
|
nosetests.xml
|
||||||
|
coverage.xml
|
||||||
|
*.cover
|
||||||
|
*.py,cover
|
||||||
|
.hypothesis/
|
||||||
|
.pytest_cache/
|
||||||
|
|
||||||
|
# Translations
|
||||||
|
*.mo
|
||||||
|
*.pot
|
||||||
|
|
||||||
|
# Django stuff:
|
||||||
|
*.log
|
||||||
|
local_settings.py
|
||||||
|
db.sqlite3
|
||||||
|
db.sqlite3-journal
|
||||||
|
|
||||||
|
# Flask stuff:
|
||||||
|
instance/
|
||||||
|
.webassets-cache
|
||||||
|
|
||||||
|
# Scrapy stuff:
|
||||||
|
.scrapy
|
||||||
|
|
||||||
|
# Sphinx documentation
|
||||||
|
docs/_build/
|
||||||
|
|
||||||
|
# PyBuilder
|
||||||
|
target/
|
||||||
|
|
||||||
|
# Jupyter Notebook
|
||||||
|
.ipynb_checkpoints
|
||||||
|
|
||||||
|
# IPython
|
||||||
|
profile_default/
|
||||||
|
ipython_config.py
|
||||||
|
|
||||||
|
# pyenv
|
||||||
|
.python-version
|
||||||
|
|
||||||
|
# pipenv
|
||||||
|
# According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
|
||||||
|
# However, in case of collaboration, if having platform-specific dependencies or dependencies
|
||||||
|
# having no cross-platform support, pipenv may install dependencies that don't work, or not
|
||||||
|
# install all needed dependencies.
|
||||||
|
#Pipfile.lock
|
||||||
|
|
||||||
|
# PEP 582; used by e.g. github.com/David-OConnor/pyflow
|
||||||
|
__pypackages__/
|
||||||
|
|
||||||
|
# Celery stuff
|
||||||
|
celerybeat-schedule
|
||||||
|
celerybeat.pid
|
||||||
|
|
||||||
|
# SageMath parsed files
|
||||||
|
*.sage.py
|
||||||
|
|
||||||
|
# Environments
|
||||||
|
.env
|
||||||
|
.venv
|
||||||
|
env/
|
||||||
|
venv/
|
||||||
|
ENV/
|
||||||
|
env.bak/
|
||||||
|
venv.bak/
|
||||||
|
|
||||||
|
# Spyder project settings
|
||||||
|
.spyderproject
|
||||||
|
.spyproject
|
||||||
|
|
||||||
|
# Rope project settings
|
||||||
|
.ropeproject
|
||||||
|
|
||||||
|
# mkdocs documentation
|
||||||
|
/site
|
||||||
|
|
||||||
|
# mypy
|
||||||
|
.mypy_cache/
|
||||||
|
.dmypy.json
|
||||||
|
dmypy.json
|
||||||
|
|
||||||
|
# Pyre type checker
|
||||||
|
.pyre/
|
||||||
11
eric_wordle/README.md
Normal file
11
eric_wordle/README.md
Normal file
@@ -0,0 +1,11 @@
|
|||||||
|
# N-dle Solver
|
||||||
|
|
||||||
|
A solver designed to beat New York Time's Wordle (link [here](https://www.nytimes.com/games/wordle/index.html)). If you are bored enough, can extend to solve the more general N-dle problem (for quordle, octordle, etc.)
|
||||||
|
|
||||||
|
I originally made this out of frustration for the game (and my own lack of lingual talent). One day, my friend thought she could beat my bot. To her dismay, she learned that she is no better than a machine. Let's see if you can do any better (the average number of attempts is 3.6).
|
||||||
|
|
||||||
|
## Usage:
|
||||||
|
1. Run `python main.py --n 1`
|
||||||
|
2. Follow the prompts
|
||||||
|
|
||||||
|
Currently only supports solving for 1 word at a time (i.e. wordle).
|
||||||
126
eric_wordle/ai.py
Normal file
126
eric_wordle/ai.py
Normal file
@@ -0,0 +1,126 @@
|
|||||||
|
import re
|
||||||
|
import string
|
||||||
|
|
||||||
|
import numpy as np
|
||||||
|
|
||||||
|
|
||||||
|
class AI:
|
||||||
|
def __init__(self, vocab_file, num_letters=5, num_guesses=6):
|
||||||
|
self.vocab_file = vocab_file
|
||||||
|
self.num_letters = num_letters
|
||||||
|
self.num_guesses = 6
|
||||||
|
|
||||||
|
self.vocab, self.vocab_scores, self.letter_scores = self.get_vocab(self.vocab_file)
|
||||||
|
self.best_words = sorted(list(self.vocab_scores.items()), key=lambda tup: tup[1])[::-1]
|
||||||
|
|
||||||
|
self.domains = None
|
||||||
|
self.possible_letters = None
|
||||||
|
|
||||||
|
self.reset()
|
||||||
|
|
||||||
|
def solve(self):
|
||||||
|
num_guesses = 0
|
||||||
|
while [len(e) for e in self.domains] != [1 for _ in range(self.num_letters)]:
|
||||||
|
num_guesses += 1
|
||||||
|
word = self.sample()
|
||||||
|
|
||||||
|
# # Always start with these two words
|
||||||
|
# if num_guesses == 1:
|
||||||
|
# word = 'soare'
|
||||||
|
# elif num_guesses == 2:
|
||||||
|
# word = 'culti'
|
||||||
|
|
||||||
|
print('-----------------------------------------------')
|
||||||
|
print(f'Guess #{num_guesses}/{self.num_guesses}: {word}')
|
||||||
|
print('-----------------------------------------------')
|
||||||
|
self.arc_consistency(word)
|
||||||
|
|
||||||
|
print(f'You did it! The word is {"".join([e[0] for e in self.domains])}')
|
||||||
|
|
||||||
|
|
||||||
|
def arc_consistency(self, word):
|
||||||
|
print(f'Performing arc consistency check on {word}...')
|
||||||
|
print(f'Specify 0 for completely nonexistent letter at the specified index, 1 for existent letter but incorrect index, and 2 for correct letter at correct index.')
|
||||||
|
results = []
|
||||||
|
|
||||||
|
# Collect results
|
||||||
|
for l in word:
|
||||||
|
while True:
|
||||||
|
result = input(f'{l}: ')
|
||||||
|
if result not in ['0', '1', '2']:
|
||||||
|
print('Incorrect option. Try again.')
|
||||||
|
continue
|
||||||
|
results.append(result)
|
||||||
|
break
|
||||||
|
|
||||||
|
self.possible_letters += [word[i] for i in range(len(word)) if results[i] == '1']
|
||||||
|
|
||||||
|
for i in range(len(word)):
|
||||||
|
if results[i] == '0':
|
||||||
|
if word[i] in self.possible_letters:
|
||||||
|
if word[i] in self.domains[i]:
|
||||||
|
self.domains[i].remove(word[i])
|
||||||
|
else:
|
||||||
|
for j in range(len(self.domains)):
|
||||||
|
if word[i] in self.domains[j] and len(self.domains[j]) > 1:
|
||||||
|
self.domains[j].remove(word[i])
|
||||||
|
if results[i] == '1':
|
||||||
|
if word[i] in self.domains[i]:
|
||||||
|
self.domains[i].remove(word[i])
|
||||||
|
if results[i] == '2':
|
||||||
|
self.domains[i] = [word[i]]
|
||||||
|
|
||||||
|
|
||||||
|
def reset(self):
|
||||||
|
self.domains = [list(string.ascii_lowercase) for _ in range(self.num_letters)]
|
||||||
|
self.possible_letters = []
|
||||||
|
|
||||||
|
def sample(self):
|
||||||
|
"""
|
||||||
|
Samples a best word given the current domains
|
||||||
|
:return:
|
||||||
|
"""
|
||||||
|
# Compile a regex of possible words with the current domain
|
||||||
|
regex_string = ''
|
||||||
|
for domain in self.domains:
|
||||||
|
regex_string += ''.join(['[', ''.join(domain), ']', '{1}'])
|
||||||
|
pattern = re.compile(regex_string)
|
||||||
|
|
||||||
|
# From the words with the highest scores, only return the best word that match the regex pattern
|
||||||
|
for word, _ in self.best_words:
|
||||||
|
if pattern.match(word) and False not in [e in word for e in self.possible_letters]:
|
||||||
|
return word
|
||||||
|
|
||||||
|
def get_vocab(self, vocab_file):
|
||||||
|
vocab = []
|
||||||
|
with open(vocab_file, 'r') as f:
|
||||||
|
for l in f:
|
||||||
|
vocab.append(l.strip())
|
||||||
|
|
||||||
|
# Count letter frequencies at each index
|
||||||
|
letter_freqs = [{letter: 0 for letter in string.ascii_lowercase} for _ in range(self.num_letters)]
|
||||||
|
for word in vocab:
|
||||||
|
for i, l in enumerate(word):
|
||||||
|
letter_freqs[i][l] += 1
|
||||||
|
|
||||||
|
# Assign a score to each letter at each index by the probability of it appearing
|
||||||
|
letter_scores = [{letter: 0 for letter in string.ascii_lowercase} for _ in range(self.num_letters)]
|
||||||
|
for i in range(len(letter_scores)):
|
||||||
|
max_freq = np.max(list(letter_freqs[i].values()))
|
||||||
|
for l in letter_scores[i].keys():
|
||||||
|
letter_scores[i][l] = letter_freqs[i][l] / max_freq
|
||||||
|
|
||||||
|
# Find a sorted list of words ranked by sum of letter scores
|
||||||
|
vocab_scores = {} # (score, word)
|
||||||
|
for word in vocab:
|
||||||
|
score = 0
|
||||||
|
for i, l in enumerate(word):
|
||||||
|
score += letter_scores[i][l]
|
||||||
|
|
||||||
|
# # Optimization: If repeating letters, deduct a couple points
|
||||||
|
# if len(set(word)) < len(word):
|
||||||
|
# score -= 0.25 * (len(word) - len(set(word)))
|
||||||
|
|
||||||
|
vocab_scores[word] = score
|
||||||
|
|
||||||
|
return vocab, vocab_scores, letter_scores
|
||||||
37
eric_wordle/dist.py
Normal file
37
eric_wordle/dist.py
Normal file
@@ -0,0 +1,37 @@
|
|||||||
|
import string
|
||||||
|
|
||||||
|
import numpy as np
|
||||||
|
|
||||||
|
words = []
|
||||||
|
with open('words.txt', 'r') as f:
|
||||||
|
for l in f:
|
||||||
|
words.append(l.strip())
|
||||||
|
|
||||||
|
# Count letter frequencies at each index
|
||||||
|
letter_freqs = [{letter: 0 for letter in string.ascii_lowercase} for _ in range(5)]
|
||||||
|
for word in words:
|
||||||
|
for i, l in enumerate(word):
|
||||||
|
letter_freqs[i][l] += 1
|
||||||
|
|
||||||
|
# Assign a score to each letter at each index by the probability of it appearing
|
||||||
|
letter_scores = [{letter: 0 for letter in string.ascii_lowercase} for _ in range(5)]
|
||||||
|
for i in range(len(letter_scores)):
|
||||||
|
max_freq = np.max(list(letter_freqs[i].values()))
|
||||||
|
for l in letter_scores[i].keys():
|
||||||
|
letter_scores[i][l] = letter_freqs[i][l] / max_freq
|
||||||
|
|
||||||
|
# Find a sorted list of words ranked by sum of letter scores
|
||||||
|
word_scores = [] # (score, word)
|
||||||
|
for word in words:
|
||||||
|
score = 0
|
||||||
|
for i, l in enumerate(word):
|
||||||
|
score += letter_scores[i][l]
|
||||||
|
word_scores.append((score, word))
|
||||||
|
|
||||||
|
sorted_by_second = sorted(word_scores, key=lambda tup: tup[0])[::-1]
|
||||||
|
print(sorted_by_second[:10])
|
||||||
|
|
||||||
|
for i, (score, word) in enumerate(sorted_by_second):
|
||||||
|
if word == 'soare':
|
||||||
|
print(f'{word} with a score of {score} is found at index {i}')
|
||||||
|
|
||||||
18
eric_wordle/main.py
Normal file
18
eric_wordle/main.py
Normal file
@@ -0,0 +1,18 @@
|
|||||||
|
import argparse
|
||||||
|
from ai import AI
|
||||||
|
|
||||||
|
|
||||||
|
def main(args):
|
||||||
|
if args.n is None:
|
||||||
|
raise Exception('Need to specify n (i.e. n = 1 for wordle, n = 4 for quordle, n = 16 for sedecordle).')
|
||||||
|
|
||||||
|
ai = AI(args.vocab_file)
|
||||||
|
ai.solve()
|
||||||
|
|
||||||
|
|
||||||
|
if __name__ == '__main__':
|
||||||
|
parser = argparse.ArgumentParser()
|
||||||
|
parser.add_argument('--n', dest='n', type=int, default=None)
|
||||||
|
parser.add_argument('--vocab_file', dest='vocab_file', type=str, default='wordle_words.txt')
|
||||||
|
args = parser.parse_args()
|
||||||
|
main(args)
|
||||||
15
eric_wordle/process.py
Normal file
15
eric_wordle/process.py
Normal file
@@ -0,0 +1,15 @@
|
|||||||
|
import pandas
|
||||||
|
|
||||||
|
print('Loading in words dictionary; this may take a while...')
|
||||||
|
df = pandas.read_json('words_dictionary.json')
|
||||||
|
print('Done loading words dictionary.')
|
||||||
|
words = []
|
||||||
|
for word in df.axes[0].tolist():
|
||||||
|
if len(word) != 5:
|
||||||
|
continue
|
||||||
|
words.append(word)
|
||||||
|
words.sort()
|
||||||
|
|
||||||
|
with open('words.txt', 'w') as f:
|
||||||
|
for word in words:
|
||||||
|
f.write(word + '\n')
|
||||||
15919
eric_wordle/words.txt
Normal file
15919
eric_wordle/words.txt
Normal file
File diff suppressed because it is too large
Load Diff
370104
eric_wordle/words_dictionary.json
Normal file
370104
eric_wordle/words_dictionary.json
Normal file
File diff suppressed because it is too large
Load Diff
108
letter_guess.py
Normal file
108
letter_guess.py
Normal file
@@ -0,0 +1,108 @@
|
|||||||
|
import gymnasium as gym
|
||||||
|
from gymnasium import spaces
|
||||||
|
import numpy as np
|
||||||
|
import random
|
||||||
|
import re
|
||||||
|
|
||||||
|
|
||||||
|
class LetterGuessingEnv(gym.Env):
|
||||||
|
"""
|
||||||
|
Custom Gymnasium environment for a letter guessing game with a focus on forming
|
||||||
|
valid prefixes and words from a list of valid Wordle words. The environment tracks
|
||||||
|
the current guess prefix and validates it against known valid words, ending the game
|
||||||
|
early with a negative reward for invalid prefixes.
|
||||||
|
"""
|
||||||
|
|
||||||
|
metadata = {'render_modes': ['human']}
|
||||||
|
|
||||||
|
def __init__(self, valid_words, seed=None):
|
||||||
|
self.action_space = spaces.Discrete(26)
|
||||||
|
self.observation_space = spaces.Box(low=0, high=1, shape=(26*2 + 26*4,), dtype=np.int32)
|
||||||
|
|
||||||
|
self.valid_words = valid_words # List of valid Wordle words
|
||||||
|
self.target_word = '' # Target word for the current episode
|
||||||
|
self.valid_words_str = ' '.join(self.valid_words) + ' '
|
||||||
|
self.letter_flags = None
|
||||||
|
self.letter_positions = None
|
||||||
|
self.guessed_letters = set()
|
||||||
|
self.guess_prefix = "" # Tracks the current guess prefix
|
||||||
|
|
||||||
|
self.reset()
|
||||||
|
|
||||||
|
def step(self, action):
|
||||||
|
letter_index = action % 26 # Assuming action is the letter index directly
|
||||||
|
position = len(self.guess_prefix) # The next position in the prefix is determined by its current length
|
||||||
|
letter = chr(ord('a') + letter_index)
|
||||||
|
|
||||||
|
reward = 0
|
||||||
|
done = False
|
||||||
|
|
||||||
|
# Check if the letter has already been used in the guess prefix
|
||||||
|
if letter in self.guessed_letters:
|
||||||
|
reward = -1 # Penalize for repeating letters in the prefix
|
||||||
|
else:
|
||||||
|
# Add the new letter to the prefix and update guessed letters set
|
||||||
|
self.guess_prefix += letter
|
||||||
|
self.guessed_letters.add(letter)
|
||||||
|
|
||||||
|
# Update letter flags based on whether the letter is in the target word
|
||||||
|
if self.target_word[position] == letter:
|
||||||
|
self.letter_flags[letter_index, :] = [1, 0] # Update flag for correct guess
|
||||||
|
elif letter in self.target_word:
|
||||||
|
self.letter_flags[letter_index, :] = [0, 1] # Update flag for correct guess wrong position
|
||||||
|
else:
|
||||||
|
self.letter_flags[letter_index, :] = [0, 0] # Update flag for incorrect guess
|
||||||
|
|
||||||
|
reward = 1 # Reward for adding new information by trying a new letter
|
||||||
|
|
||||||
|
# Update the letter_positions matrix to reflect the new guess
|
||||||
|
if position == 4:
|
||||||
|
self.letter_positions[:,:] = 1
|
||||||
|
else:
|
||||||
|
self.letter_positions[:, position] = 0
|
||||||
|
self.letter_positions[letter_index, position] = 1
|
||||||
|
|
||||||
|
# Use regex to check if the current prefix can lead to a valid word
|
||||||
|
if not re.search(r'\b' + self.guess_prefix, self.valid_words_str):
|
||||||
|
reward = -5 # Penalize for forming an invalid prefix
|
||||||
|
done = True # End the episode if the prefix is invalid
|
||||||
|
|
||||||
|
# guessed a full word so we reset our guess prefix to guess next round
|
||||||
|
if len(self.guess_prefix) == len(self.target_word):
|
||||||
|
self.guess_prefix = ''
|
||||||
|
self.round += 1
|
||||||
|
|
||||||
|
# end after 5 rounds of total guesses
|
||||||
|
if self.round == 2:
|
||||||
|
# reward = 5
|
||||||
|
done = True
|
||||||
|
|
||||||
|
obs = self._get_obs()
|
||||||
|
|
||||||
|
if reward < -50:
|
||||||
|
print(obs, reward, done)
|
||||||
|
|
||||||
|
return obs, reward, done, False, {}
|
||||||
|
|
||||||
|
def reset(self, seed=None):
|
||||||
|
self.target_word = random.choice(self.valid_words)
|
||||||
|
# self.target_word_encoded = self.encode_word(self.target_word)
|
||||||
|
self.letter_flags = np.ones((26, 2), dtype=np.int32)
|
||||||
|
self.letter_positions = np.ones((26, 4), dtype=np.int32)
|
||||||
|
self.guessed_letters = set()
|
||||||
|
self.guess_prefix = "" # Reset the guess prefix for the new episode
|
||||||
|
self.round = 1
|
||||||
|
return self._get_obs(), {}
|
||||||
|
|
||||||
|
def encode_word(self, word):
|
||||||
|
encoded = np.zeros((26,))
|
||||||
|
for char in word:
|
||||||
|
index = ord(char) - ord('a')
|
||||||
|
encoded[index] = 1
|
||||||
|
return encoded
|
||||||
|
|
||||||
|
def _get_obs(self):
|
||||||
|
return np.concatenate([self.letter_flags.flatten(), self.letter_positions.flatten()])
|
||||||
|
|
||||||
|
def render(self, mode='human'):
|
||||||
|
pass # Optional: Implement rendering logic if needed
|
||||||
165
test.ipynb
165
test.ipynb
File diff suppressed because one or more lines are too long
61
test.py
61
test.py
@@ -1,61 +0,0 @@
|
|||||||
|
|
||||||
from torch.utils.data import Dataset
|
|
||||||
from transformers import BertGenerationEncoder, BertGenerationDecoder, EncoderDecoderModel, BertTokenizer
|
|
||||||
from tqdm import tqdm as progress_bar
|
|
||||||
import torch
|
|
||||||
import matplotlib
|
|
||||||
|
|
||||||
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
|
||||||
print(device)
|
|
||||||
|
|
||||||
encoder = BertGenerationEncoder.from_pretrained("google-bert/bert-base-uncased", bos_token_id=101, eos_token_id=102)
|
|
||||||
# add cross attention layers and use BERT's cls token as BOS token and sep token as EOS token
|
|
||||||
decoder = BertGenerationDecoder.from_pretrained("google-bert/bert-base-uncased", add_cross_attention=True, is_decoder=True, bos_token_id=101, eos_token_id=102)
|
|
||||||
model = EncoderDecoderModel(encoder=encoder, decoder=decoder)
|
|
||||||
|
|
||||||
# create tokenizer...
|
|
||||||
tokenizer = BertTokenizer.from_pretrained("google-bert/bert-large-uncased")
|
|
||||||
|
|
||||||
import json
|
|
||||||
|
|
||||||
class CodeDataset(Dataset):
|
|
||||||
def __init__(self):
|
|
||||||
with open("data/conala-train.json") as f:
|
|
||||||
self.data = json.load(f)
|
|
||||||
|
|
||||||
def __len__(self):
|
|
||||||
return len(self.data)
|
|
||||||
|
|
||||||
def __getitem__(self, idx):
|
|
||||||
intent = self.data[idx]["rewritten_intent"] if self.data[idx]["rewritten_intent"] else self.data[idx]["intent"]
|
|
||||||
return intent, self.data[idx]["snippet"]
|
|
||||||
|
|
||||||
|
|
||||||
optimizer = torch.optim.AdamW(params=model.parameters(), lr=1e-3)
|
|
||||||
dataloader = CodeDataset()
|
|
||||||
model = model.to(device)
|
|
||||||
|
|
||||||
losses = []
|
|
||||||
epochs = 10
|
|
||||||
for i in range(epochs):
|
|
||||||
|
|
||||||
epoch_loss = 0
|
|
||||||
|
|
||||||
for idx, (question, answer) in progress_bar(enumerate(dataloader), total=len(dataloader)):
|
|
||||||
|
|
||||||
input_ids = tokenizer(question, add_special_tokens=False, return_tensors="pt").input_ids.to(device)
|
|
||||||
label_ids = tokenizer(answer, return_tensors="pt").input_ids.to(device)
|
|
||||||
|
|
||||||
loss = model(input_ids=input_ids, decoder_input_ids=label_ids, labels=label_ids).loss
|
|
||||||
|
|
||||||
optimizer.zero_grad()
|
|
||||||
loss.backward()
|
|
||||||
optimizer.step()
|
|
||||||
|
|
||||||
epoch_loss += loss.item()
|
|
||||||
|
|
||||||
losses.append(epoch_loss)
|
|
||||||
|
|
||||||
plt.plot(losses, color="green", label="Training Loss")
|
|
||||||
plt.legend(loc = 'upper left')
|
|
||||||
plt.savefig("plot.png")
|
|
||||||
2317
wordle_words.txt
Normal file
2317
wordle_words.txt
Normal file
File diff suppressed because it is too large
Load Diff
Reference in New Issue
Block a user