csvance / armchair-expert Goto Github PK
View Code? Open in Web Editor NEWMachine Learning Chatbot
License: MIT License
Machine Learning Chatbot
License: MIT License
Seems like sometimes the twitter backend does a double reply for a single input. Investigate and fix.
Traceback (most recent call last): File "import_text_file.py", line 43, in <module> main() File "import_text_file.py", line 28, in main data_manager = ImportTrainingDataManager() NameError: name 'ImportTrainingDataManager' is not defined
When I start the bot I get it to the RUNNING state, however, I don't believe it is storing anything from my discord channel.
I noticed that when I check the weights directory it is empty. Whenever I place the test trump data that you provided in the weights folder, I get responses from the bot in discord. Should the python program be creating the markov.json.zlib and structure-model.h5 in the weights directory on start-up?
Images for reference:
Thanks!
Hi !
I dropping your code into a docker container and used pigar to build a requirements.txt file and for some reason it pulled down common and installed it via pip3.
Took me a while to figure out why common.nlp couldn't be found.
Could you at a later date change your namespace to be a bit more exotic so other less capable python folk don't trip over this.
I'll update this with the docker container when it is up on dockerhub
Backends for Twitter, Slack, and IRC should be easy to integrate and run in parallel with the rest of the codebase.
Instead of just having one feature that says a word is at the start of a sentence, an RNN potentially allows for more accuracy in the reproduction of capitalization style for each PoS.
The process of setting up the bot and training it initially doesn't provide much useful information about the learned knowledge of the bot. Features should be added to provide statistics on what the bot knows to the user.
When setting up this bot with my discord server, I get this exception randomly whenever someone sends the bot a message:
divide
p_values = distance_magnitudes / sums
/Users/hwashere/Desktop/PROJECTS/python-copebot/common/ml.py:10: RuntimeWarning: divide by zero encountered in log
preds = np.log(preds) / temperature
Traceback (most recent call last):
File "copebot_python_edition.py", line 289, in <module>
cpe.start(retrain_structure=args.retrain_structure, retrain_markov=args.retrain_markov)
File "copebot_python_edition.py", line 95, in start
self._main()
File "copebot_python_edition.py", line 247, in _main
reply = connector.generate(message, doc=doc)
File "/Users/hwashere/Desktop/PROJECTS/python-copebot/connectors/connector_common.py", line 164, in generate
return self._reply_generator.generate(message, doc)
File "/Users/hwashere/Desktop/PROJECTS/python-copebot/connectors/bot_instance.py", line 17, in generate
reply = ConnectorReplyGenerator.generate(self, message, doc, ignore_topics=[BOT_USERNAME.split('#')[0]])
File "/Users/hwashere/Desktop/PROJECTS/python-copebot/connectors/connector_common.py", line 59, in generate
sentences = generator.generate(db=self._markov_model)
File "/Users/hwashere/Desktop/PROJECTS/python-copebot/markov_engine.py", line 364, in generate
if not self._generate_words(db):
File "/Users/hwashere/Desktop/PROJECTS/python-copebot/markov_engine.py", line 504, in _generate_words
handle_projections()
File "/Users/hwashere/Desktop/PROJECTS/python-copebot/markov_engine.py", line 473, in handle_projections
word_choice_idx = temp(p_values, temperature=MARKOV_MODEL_TEMPERATURE)
File "/Users/hwashere/Desktop/PROJECTS/python-copebot/common/ml.py", line 13, in temp
probas = np.random.multinomial(1, preds, 1)
File "mtrand.pyx", line 3863, in numpy.random.mtrand.RandomState.multinomial
File "common.pyx", line 323, in numpy.random.common.check_array_constraint
ValueError: pvals < 0, pvals > 1 or pvals contains NaNs```
Any idea what could be causing this crash?
INFO:discord.http:A rate limit bucket has been exhausted
Is there a lever for lowering the responses or is this perhaps bad training data that is causing my bot to run rampant ?
The current model's performance is very poor for generation. Create a new markov engine using a trie database. Instead of a neighbor system, we should embed the frequency of the distance the word occurs from others so we can calculate the probability it will be used in a certain position.
Hi, firstly apologies because I hate to create issues, but I cant seem to fix this error by myself. I'm a bit of a noob.
It seems to finish loading and display the graphical info on training stats in the console, but after a few seconds another message appears that seems to be an error that stops the bot from working correctly.
Using Python 3.7, all modules installed and updated, on discord.
Console:
D:\Mirai\armchair-expert-master>c:\Users\Jake\AppData\Local\Programs\Python\Python37\python.exe D:\Mirai\armchair-expert-master\armchair_expert.py
INFO:ArmchairExpert:Status: STARTING_UP
INFO:ArmchairExpert:Loaded Discord Connector.
INFO:ArmchairExpert:Loading spaCy model
INFO:ArmchairExpert:Training begin
INFO:ArmchairExpert:Training_Preprocessing_Markov(Import)
INFO:ArmchairExpert:Training_Preprocessing_Markov(Discord)
INFO:ArmchairExpert:Training(Markov)
INFO:ArmchairExpert:Training_Preprocessing_Structure(Import)
INFO:ArmchairExpert:Training_Preprocessing_Structure(Discord)
INFO:ArmchairExpert:Training(Structure)
Using TensorFlow backend.
Using TensorFlow backend.
WARNING:tensorflow:From c:\Users\Jake\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow\python\framework\op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.
Instructions for updating:
Colocations handled automatically by placer.
WARNING:tensorflow:From c:\Users\Jake\AppData\Local\Programs\Python\Python37\lib\site-packages\keras\backend\tensorflow_backend.py:3445: calling dropout (from tensorflow.python.ops.nn_ops) with keep_prob is deprecated and will be removed in a future version.
Instructions for updating:
Please use `rate` instead of `keep_prob`. Rate should be set to `rate = 1 - keep_prob`.
INFO:ArmchairExpert:Training end
INFO:ArmchairExpert:Status: RUNNING
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
embedding_1 (Embedding) (None, 16, 120) 14400
_________________________________________________________________
lstm_1 (LSTM) (None, 128) 127488
_________________________________________________________________
dense_1 (Dense) (None, 120) 15480
=================================================================
Total params: 157,368
Trainable params: 157,368
Non-trainable params: 0
_________________________________________________________________
Task exception was never retrieved
future: <Task finished coro=<ConnectionState._delay_ready() done, defined at c:\Users\Jake\AppData\Local\Programs\Python\Python37\lib\site-packages\discord\state.py:286> exception=AttributeError("'bool' object has no attribute 'set'")>
Traceback (most recent call last):
File "c:\Users\Jake\AppData\Local\Programs\Python\Python37\lib\site-packages\discord\state.py", line 322, in _delay_ready
self.call_handlers('ready')
File "c:\Users\Jake\AppData\Local\Programs\Python\Python37\lib\site-packages\discord\state.py", line 139, in call_handlers
func(*args, **kwargs)
File "c:\Users\Jake\AppData\Local\Programs\Python\Python37\lib\site-packages\discord\client.py", line 207, in _handle_ready
self._ready.set()
AttributeError: 'bool' object has no attribute 'set'
After this the bot still appears to be online, and direct messaging it yields further console text every time a message is sent, either direct or in a dm:
Ignoring exception in on_message
Traceback (most recent call last):
File "c:\Users\Jake\AppData\Local\Programs\Python\Python37\lib\site-packages\discord\client.py", line 255, in _run_event
await coro(*args, **kwargs)
File "D:\Mirai\armchair-expert-master\connectors\discord.py", line 54, in on_message
if message.server is None and DISCORD_LEARN_FROM_DIRECT_MESSAGE:
AttributeError: 'Message' object has no attribute 'server'
The bot never sends any messages out to discord. :(
Thanks,
-I.V.
NOUN -> ADJ
NOUN -> VERB
NOUN -> NUM
NOUN -> SYM
NOUN -> HASHTAG
NOUN -> EMOJI
NOUN -> URL
VERB -> ADV
VERB -> NOUN
VERB -> NUM
VERB -> SYM
VERB -> HASHTAG
VERB -> EMOJI
VERB -> URL
HASHTAG -> NOUN
HASHTAG -> VERB
HASHTAG -> ADV
HASHTAG -> ADJ
HASHTAG -> SYM
HASHTAG -> EMOJI
HASHTAG -> URL
Hello, when I tried running this and using it, it ever worked. It was just stuck on this page.
C:\Users\Mohit\PycharmProjects\SuperAIPR\venv\Scripts\python.exe C:/Users/Mohit/PycharmProjects/SuperAIPR/armchair-expert/armchair_expert.py
INFO:ArmchairExpert:Status: STARTING_UP
INFO:ArmchairExpert:Loaded Discord Connector.
INFO:ArmchairExpert:Loading spaCy model
2020-11-10 13:41:58.346187: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cudart64_101.dll
INFO:ArmchairExpert:Training begin
INFO:ArmchairExpert:Training_Preprocessing_Markov(Import)
INFO:ArmchairExpert:Training_Preprocessing_Markov(Discord)
INFO:ArmchairExpert:Training(Markov)
INFO:ArmchairExpert:Training_Preprocessing_Structure(Import)
INFO:ArmchairExpert:Training_Preprocessing_Structure(Discord)
INFO:ArmchairExpert:Training(Structure)
2020-11-10 13:41:58.930290: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cudart64_101.dll
INFO:ArmchairExpert:Training end
INFO:ArmchairExpert:Status: RUNNING
2020-11-10 13:42:08.089287: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library nvcuda.dll
2020-11-10 13:42:08.173283: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1716] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: GeForce GTX 680 computeCapability: 3.0
coreClock: 1.0585GHz coreCount: 8 deviceMemorySize: 2.00GiB deviceMemoryBandwidth: 179.05GiB/s
2020-11-10 13:42:08.174598: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1716] Found device 1 with properties:
pciBusID: 0000:20:00.0 name: Quadro P400 computeCapability: 6.1
coreClock: 1.2525GHz coreCount: 2 deviceMemorySize: 2.00GiB deviceMemoryBandwidth: 29.88GiB/s
2020-11-10 13:42:08.174914: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cudart64_101.dll
2020-11-10 13:42:08.234582: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cublas64_10.dll
2020-11-10 13:42:08.263904: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cufft64_10.dll
2020-11-10 13:42:08.272198: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library curand64_10.dll
2020-11-10 13:42:08.350489: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cusolver64_10.dll
2020-11-10 13:42:08.376081: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cusparse64_10.dll
2020-11-10 13:42:08.379728: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cudnn64_7.dll
2020-11-10 13:42:08.380804: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1812] Ignoring visible gpu device (device: 0, name: GeForce GTX 680, pci bus id: 0000:01:00.0, compute capability: 3.0) with Cuda compute capability 3.0. The minimum required Cuda capability is 3.5.
2020-11-10 13:42:08.381653: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1843] Ignoring visible gpu device (device: 1, name: Quadro P400, pci bus id: 0000:20:00.0, compute capability: 6.1) with core count: 2. The minimum required count is 8. You can adjust this requirement with the env var TF_MIN_GPU_MULTIPROCESSOR_COUNT.
2020-11-10 13:42:08.432344: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x1715f876120 initialized for platform Host (this does not guarantee that XLA will be used). Devices:
2020-11-10 13:42:08.432594: I tensorflow/compiler/xla/service/service.cc:176] StreamExecutor device (0): Host, Default Version
2020-11-10 13:42:08.434576: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1257] Device interconnect StreamExecutor with strength 1 edge matrix:
2020-11-10 13:42:08.434815: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1263]
Model: "sequential"
embedding (Embedding) (None, 16, 120) 14400
lstm (LSTM) (None, 128) 127488
Total params: 157,368
Trainable params: 157,368
Non-trainable params: 0
2020-11-10 13:42:09.270123: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1716] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: GeForce GTX 680 computeCapability: 3.0
coreClock: 1.0585GHz coreCount: 8 deviceMemorySize: 2.00GiB deviceMemoryBandwidth: 179.05GiB/s
2020-11-10 13:42:09.270433: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1716] Found device 1 with properties:
pciBusID: 0000:20:00.0 name: Quadro P400 computeCapability: 6.1
coreClock: 1.2525GHz coreCount: 2 deviceMemorySize: 2.00GiB deviceMemoryBandwidth: 29.88GiB/s
2020-11-10 13:42:09.272560: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cudart64_101.dll
2020-11-10 13:42:09.273504: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cublas64_10.dll
2020-11-10 13:42:09.275615: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cufft64_10.dll
2020-11-10 13:42:09.275742: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library curand64_10.dll
2020-11-10 13:42:09.275866: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cusolver64_10.dll
2020-11-10 13:42:09.275991: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cusparse64_10.dll
2020-11-10 13:42:09.276120: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library cudnn64_7.dll
2020-11-10 13:42:09.277873: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1812] Ignoring visible gpu device (device: 0, name: GeForce GTX 680, pci bus id: 0000:01:00.0, compute capability: 3.0) with Cuda compute capability 3.0. The minimum required Cuda capability is 3.5.
2020-11-10 13:42:09.278200: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1843] Ignoring visible gpu device (device: 1, name: Quadro P400, pci bus id: 0000:20:00.0, compute capability: 6.1) with core count: 2. The minimum required count is 8. You can adjust this requirement with the env var TF_MIN_GPU_MULTIPROCESSOR_COUNT.
2020-11-10 13:42:09.279155: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1257] Device interconnect StreamExecutor with strength 1 edge matrix:
2020-11-10 13:42:09.279284: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1263] 0 1
2020-11-10 13:42:09.279366: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1276] 0: N N
2020-11-10 13:42:09.279511: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1276] 1: N N
2020-11-10 13:42:09.281033: I tensorflow/compiler/xla/service/platform_util.cc:139] StreamExecutor cuda device (0) is of insufficient compute capability: 3.5 required, device is 3.0
2020-11-10 13:42:09.282005: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x17167ef3790 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:
2020-11-10 13:42:09.282233: I tensorflow/compiler/xla/service/service.cc:176] StreamExecutor device (0): Quadro P400, Compute Capability 6.1
Its been like this for 20 minutes and I don't know what to do
Log data through discord to a database to use for scheduled training sessions
Train markov module on new data while running
Train neural nets on new data on startup
Detect when the userstream needs to be restarted and restart it.
I abandoned running armchair-expert in a docker container for now but thought I'd play with using youtube captions as a source for data. I've managed to load the data ok, but no matter what I say to the bot, it either replies back with the a single noun of what I wrote or a 'huh ?' .
I'd love to hear ideas on how to debug this. Thanks !
The punctuation marks . (sometimes), !, and ? indicate the end of a sentence. Store these in the PoS tree for use in generation later so the generated sentence matches the original classification of sentence in both punctuation and PoS. Check if spaCy can differentiate between period used in abbreviations and as a sentence terminator
It looks like the dual LSTM architecture and hidden size may be complete overkill, and hence causing very slow / ineffective training. Going to test moving changing the dim to 64 and removing one of the LSTM's entirely.
Also the number of epochs should be scaled by the amount of training data available.
Usernames with underscores don't seem to survive intact. Swap twitter handles out before doing filtering, then swap them back in.
Have the bot fill in a mad lib template based on a subject.
We have run into the performance ceiling with RDBMS for loading data and generating replies.
Going to look into creating an optimized Markov chain database. Some ideas:
Hey. I've been trying to set this bot up for a Discord channel. I have very little experience with Python, so this was probably a bit too complex to dive into for me, but I thought I was making good progress. I got all the requirements and dependencies set up and managed to feed the training module a chunk of data that it processed just fine. But when it comes to actually connecting it throws up this:
Traceback (most recent call last):
File "C:\DiscordBot\Armchair\armchair_expert.py", line 334, in
ae.start(retrain_structure=args.retrain_structure, retrain_markov=args.retrain_markov)
File "C:\DiscordBot\Armchair\armchair_expert.py", line 106, in start
connector.start()
File "C:\DiscordBot\Armchair\connectors\connector_common.py", line 122, in start
self._scheduler.start()
File "C:\DiscordBot\Armchair\connectors\connector_common.py", line 101, in start
self._worker.start()
File "C:\Users\Barrin\Anaconda3\lib\multiprocessing\process.py", line 105, in start
self._popen = self._Popen(self)
File "C:\Users\Barrin\Anaconda3\lib\multiprocessing\context.py", line 223, in _Popen
return _default_context.get_context().Process._Popen(process_obj)
File "C:\Users\Barrin\Anaconda3\lib\multiprocessing\context.py", line 322, in _Popen
return Popen(process_obj)
File "C:\Users\Barrin\Anaconda3\lib\multiprocessing\popen_spawn_win32.py", line 65, in init
reduction.dump(process_obj, to_child)
File "C:\Users\Barrin\Anaconda3\lib\multiprocessing\reduction.py", line 60, in dump
ForkingPickler(file, protocol).dump(obj)
TypeError: can't pickle _thread.RLock objects
Traceback (most recent call last):
File "", line 1, in
File "C:\Users\Barrin\Anaconda3\lib\multiprocessing\spawn.py", line 105, in spawn_main
exitcode = _main(fd)
File "C:\Users\Barrin\Anaconda3\lib\multiprocessing\spawn.py", line 115, in _main
self = reduction.pickle.load(from_parent)
EOFError: Ran out of input
Add a thread to the twitter frontend which pulls and stores new data for learning.
Traceback (most recent call last):
File "C:\Users\Admin\CBL.tech\MARKPVNBOT 2\FTBot-master\ftbot_discord.py", line 1, in
from ftbot import *
File "C:\Users\Admin\CBL.tech\MARKPVNBOT 2\FTBot-master\ftbot.py", line 2, in
from search import *
File "C:\Users\Admin\CBL.tech\MARKPVNBOT 2\FTBot-master\search.py", line 1, in
from googleapiclient.discovery import build
ImportError: No module named 'googleapiclient'
Implement some sort of logging framework, especially for exceptions and debug information.
This appears while the bot is using Discord and responds to anyone..
After that it answers just sometimes..
Implement a dump of discord channel history if possible with API
Allow replyrate, shutup, wakeup to be done on a per channel basis for chat programs like Discord, Slack, IRC, etc. Source level configuration for private message / dm. Create DB tables for source (Discord, Slack, Twitter, etc), server (where applicable), and channel. Source and server can both store private message settings. First server will be checked, and if it does not exist or has NULL entires for the options, it is loaded from the source instead.
Mix random and greediness for better output with a system like this:
def sample(preds, temperature=1.0):
preds = np.asarray(preds).astype('float64')
preds = np.log(preds) / temperature
exp_preds = np.exp(preds)
preds = exp_preds / np.sum(exp_preds)
probas = np.random.multinomial(1, preds, 1)
return np.argmax(probas)
I downloaded and updated all the modules that armchair requires. I imported my chat log using your text import tool. It works fine. But I get this after I start the bot.
INFO:ArmchairExpert:Training_Preprocessing_Markov(Import): 25.458330%
Traceback (most recent call last):
File "armchair_expert.py", line 344, in
ae.start(retrain_structure=args.retrain_structure, retrain_markov=args.retrain_markov)
File "armchair_expert.py", line 98, in start
self.train(retrain_structure=True, retrain_markov=retrain_markov)
File "armchair_expert.py", line 273, in train
self._train_markov(retrain_markov)
File "armchair_expert.py", line 223, in _train_markov
spacy_preprocessor = self._preprocess_markov_data(all_training_data=retrain)
File "armchair_expert.py", line 180, in _preprocess_markov_data
doc = self._nlp(MarkovFilters.filter_input(message[0].decode()))
File "/opt/rh/rh-python36/root/usr/lib64/python3.6/site-packages/spacy/language.py", line 402, in call
doc = proc(doc, **component_cfg.get(name, {}))
File "/opt/rh/rh-python36/root/usr/lib/python3.6/site-packages/spacymoji/init.py", line 87, in call
retokenizer.merge(span)
File "_retokenize.pyx", line 56, in spacy.tokens._retokenize.Retokenizer.merge
ValueError: [E102] Can't merge non-disjoint spans. '๐ฆ' is already part of tokens to merge.
Need to add Discord support back into the bot though its own frontend
I built tensorflow (r2.0) from source to add AVX2 FMA cpu support (and because basic pip installation wasn't working either).
Here what i got when i start python armchair_expert.py --retrain-structure
(running on docker Ubuntu 18.04) :
INFO:ArmchairExpert:Status: STARTING_UP
/usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/dtypes.py:516: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint8 = np.dtype([("qint8", np.int8, 1)])
/usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/dtypes.py:517: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint8 = np.dtype([("quint8", np.uint8, 1)])
/usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/dtypes.py:518: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint16 = np.dtype([("qint16", np.int16, 1)])
/usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/dtypes.py:519: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint16 = np.dtype([("quint16", np.uint16, 1)])
/usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/dtypes.py:520: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint32 = np.dtype([("qint32", np.int32, 1)])
/usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_resource = np.dtype([("resource", np.ubyte, 1)])
/usr/local/lib/python3.6/dist-packages/tensorboard/compat/tensorflow_stub/dtypes.py:541: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint8 = np.dtype([("qint8", np.int8, 1)])
/usr/local/lib/python3.6/dist-packages/tensorboard/compat/tensorflow_stub/dtypes.py:542: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint8 = np.dtype([("quint8", np.uint8, 1)])
/usr/local/lib/python3.6/dist-packages/tensorboard/compat/tensorflow_stub/dtypes.py:543: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint16 = np.dtype([("qint16", np.int16, 1)])
/usr/local/lib/python3.6/dist-packages/tensorboard/compat/tensorflow_stub/dtypes.py:544: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint16 = np.dtype([("quint16", np.uint16, 1)])
/usr/local/lib/python3.6/dist-packages/tensorboard/compat/tensorflow_stub/dtypes.py:545: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint32 = np.dtype([("qint32", np.int32, 1)])
/usr/local/lib/python3.6/dist-packages/tensorboard/compat/tensorflow_stub/dtypes.py:550: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_resource = np.dtype([("resource", np.ubyte, 1)])
Using TensorFlow backend.
WARNING: Logging before flag parsing goes to stderr.
W0805 14:12:44.053263 139799969265472 deprecation_wrapper.py:118] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:74: The name tf.get_default_graph is deprecated. Please use tf.compat.v1.get_default_graph instead.W0805 14:12:44.061141 139799969265472 deprecation_wrapper.py:118] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:517: The name tf.placeholder is deprecated. Please use tf.compat.v1.placeholder instead.
W0805 14:12:44.062849 139799969265472 deprecation_wrapper.py:118] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:4138: The name tf.random_uniform is deprecated. Please use tf.random.uniform instead.
W0805 14:12:44.131823 139799969265472 deprecation_wrapper.py:118] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:133: The name tf.placeholder_with_default is deprecated. Please use tf.compat.v1.placeholder_with_default instead.
W0805 14:12:44.137647 139799969265472 deprecation.py:506] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:3445: calling dropout (from tensorflow.python.ops.nn_ops) with keep_prob is deprecated and will be removed in a future version.
Instructions for updating:
Please userate
instead ofkeep_prob
. Rate should be set torate = 1 - keep_prob
.
W0805 14:12:44.326126 139799969265472 deprecation_wrapper.py:118] From /usr/local/lib/python3.6/dist-packages/keras/optimizers.py:790: The name tf.train.Optimizer is deprecated. Please use tf.compat.v1.train.Optimizer instead.W0805 14:12:44.340184 139799969265472 deprecation_wrapper.py:118] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:3341: The name tf.log is deprecated. Please use tf.math.log instead.
2019-08-05 14:12:44.352372: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 3795880000 Hz
2019-08-05 14:12:44.352505: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x1cbc580 executing computations on platform Host. Devices:
2019-08-05 14:12:44.352545: I tensorflow/compiler/xla/service/service.cc:175] StreamExecutor device (0): ,
The training databases should only be indexed on whether the data is learned yet or not. Having a ton of (mostly unused) indexes makes inserts slower for no good reason.
CHAOS should match "aBcD" instead of "aB" like it currently does.
We should have a group of scripts for importing data into the DB lines table, and a seperate group of scripts for training the neural networks. Some scripts which import data into the DB lines table could use arguments instead of hardcoded paths.
Each background thread should be torn down when the main threat / process receives a signal to shut down, ie. CTRL+C, SIGTERM, etc.
Most of the markov, reaction model, pos tree, and capitalization model config options should be in a separate file so they don't clutter up the main config.
Since we are implementing word embedding neural nets in place of relational database for most learning and generation, have the database take care of things we don't need to have running in a neural net. The engine can use SQLite again, and separate tables can be created for each PoS we need to store to keep performance good.
The new tables should just contain id, text, and count fields. The sum(count) for calculating p-values can be calculated on startup and managed in memory instead of querying the database for it every time.
When someone @'s the bot it seems to intake its ID string as "<@!7456453436456356>" instead of "@chatbot", and when it sends a message in discord it will say "<@!7456453436456356>" or someone else's username ID instead of "@chatbot" or "@taggeduser", any idea on how to fix this?
Hey i know this project is old but would it be possible to add a webserver you can post messages to and get a response and learn from all the messages
i done everything but do i ping her so she learns or what do i do to make her speak please help
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.