prefecthq / marvin Goto Github PK
View Code? Open in Web Editor NEW✨ Build AI interfaces that spark joy
Home Page: https://askmarvin.ai
License: Apache License 2.0
marvin's People
Contributors
![dependabot[bot] avatar](https://avatars.githubusercontent.com/in/29110?v=4 "dependabot[bot]")
Stargazers
Watchers
Forkers
anna-geller thomas-te michaelbernstein moerehman dexterlagan hirajanwin lesliesibanda naripok dschiffer12 csreuter codeaudit sarahmk125 chrisguidry tokenbender touristshaun dmarx jjyolz kekewind williamdeve maniklem scylla2020 joskid asehmi soltrinox discdiver lopyo-art martinrooty leiguanape eltociear michellmelloo mashirashinaa shakomatic corina26 techventurebuilder twilwa 0xdigiscore afinlayv ihshann enkaybit hhy5277 soi-20 newmedia2 eudaimoniatech appidalice rhkdgh255 lawwu kirillkazakov8 hansv905 gitrjaa stjordanis richardscottoz jaxleharper elcubonegro deniska83 set0ri vedavyas003 babyblue26 mojolab houshijie-2020 ksallee gautamveldanda johnnie88 husainshtia bigdatasciencegroup mkram amogh-opsverse brianjking foxgrover vickyliin mozayyanifar lucemia livingbio geertjan-garvis nickkatsios jjfantini prasadp4009 jeremyengram geminixiang aicodehunt hbcbh1999 mrubash1 filpano off-by-some chriscarrollsmith tzengwei danieljjh fdoperezi edskamor mbaneshi ssavala roboojack rogervaas skywolfdreamer techthiyanes 7vietiationu 1lincravite 0juvalustto denuexleomi tonywhite11 troscuzdiatomarvin's Issues
Add optional memoization for AI Functions
Add optional memoziation for AI functions to avoid repeated calls with the same inputs. This won't always be desirable, but for well structured functions with more deterministic outputs it could be.
Print the model used in all cases, not just with 3.5
First check
- I added a descriptive title to this issue.
- I used the GitHub search to look for a similar issue and didn't find it.
- I searched the Marvin documentation for this feature.
Describe the current behavior
When GPT 3.5 is used, that information is printed. If 4 is used, that info isn't printed.
Describe the proposed behavior
Print the model in all cases, as it's helpful for the user and we plan to add more models.
Example Use
No response
Additional context
No response
Prompting for "True" and "False" (capitalized) can break bool returns
@ai_fn
def classify_sentiment(tweets: list[str]) -> list[bool]:
"""
Given a list of tweets, classifies each one as
positive (True) or negative (False) and returns
a corresponding list
"""
classify_sentiment(['i love pizza', 'i hate pizza'])This errors, but changing the prompts to "true" and "false" (lowercase) works
OSError: could not get source code
I am getting this error:
root@2f43e901cf78:/# python
Python 3.10.10 (main, Mar 23 2023, 14:25:51) [GCC 10.2.1 20210110] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> from marvin import ai_fn
[03/31/23 08:59:34] INFO marvin.marvin: Using OpenAI model "gpt-3.5-turbo" logging.py:50
>>> @ai_fn
... def list_fruits(n):
... "generate a list of n fruits"
...
>>> list_fruits(5)
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/usr/local/lib/python3.10/site-packages/marvin/bots/ai_functions.py", line 141, in ai_fn_wrapper
function_def = inspect.cleandoc(inspect.getsource(fn))
File "/usr/local/lib/python3.10/inspect.py", line 1139, in getsource
lines, lnum = getsourcelines(object)
File "/usr/local/lib/python3.10/inspect.py", line 1121, in getsourcelines
lines, lnum = findsource(object)
File "/usr/local/lib/python3.10/inspect.py", line 958, in findsource
raise OSError('could not get source code')
OSError: could not get source code
Upgrading to 0.7.4 leads to database error when using `marvin chat`
context:
Name Version Build Channel
marvin 0.7.1 pypi_0 pypi
same with 0.7.4
marvin chat
results in the TUI displaying and then a very log error message that concludes with the following:
...
OperationalError: (sqlite3.OperationalError) no such column: bot_config.created_at
[SQL: SELECT bot_config.created_at, bot_config.updated_at, bot_config.plugins, bot_config.input_transformers,
bot_config.id, bot_config.name, bot_config.personality, bot_config.instructions,
bot_config.instructions_template, bot_config.description, bot_config.profile_picture
FROM bot_config
WHERE bot_config.name = ?
LIMIT ? OFFSET ?]
[parameters: ('Marvin', 1, 0)]
(Background on this error at: https://sqlalche.me/e/14/e3q8)
Running a script with an AI function completes with the correct output, but also throws the error:
sqlalchemy.exc.OperationalError: (sqlite3.OperationalError) no such column: bot_config.created_at
[SQL: SELECT bot_config.created_at, bot_config.updated_at, bot_config.plugins, bot_config.input_transformers, bot_config.id, bot_config.name, bot_config.personality, bot_config.instructions, bot_config.instructions_template, bot_config.description, bot_config.profile_picture
FROM bot_config
WHERE bot_config.name = ?
LIMIT ? OFFSET ?]
[parameters: ('DuetBot', 1, 0)]
(Background on this error at: https://sqlalche.me/e/14/e3q8)
Add PDF loaders
Include some plugin output in memory
To help avoid hallucinated links, in situations like:
user: common pitfalls running prefect on ECS?
marvin: there's a discourse post about this! (not lying here, I can see it in the DEBUG: Chroma plugin output)
user: cool, please provide a link
marvin: sure! here's a totally fake link that looks real
we may want to include plugin output in bot.history. However, to avoid eating too much of our context window, we should probably summarize the body of the plugin output while carrying forward any links present in the raw plugin output.
ai_fn response value returning title/type dict when using 3.5-turbo
@ai_fn
def watering_schedule_from_plant(plant_name: str) -> str:
"""takes a plant name and returns watering instructions"""
instruction = watering_schedule_from_plant(plant_name="ficus")
print(instruction)
returns
{"title": "Model", "type": "string"}
Support Azure OpenAI API
See: https://news.ycombinator.com/item?id=35498110
It looks like this has support in Langchain via the Chat API: https://python.langchain.com/en/latest/modules/models/chat/integrations/azure_chat_openai.html
Non-descriptive error when SSL certificate not configured correctly
After upgrading to Python 3.10, certifi was not set up correctly. After installing marvin, I could run openai-setup and set my api key, but upon running marvin chat the message would retry ~6 times and surface error APIConnectionError: Error communicating with OpenAI.
The full stacktrace revealed several SSL-related errors. I ended up running certificates.command within my root Python folder, after which SSL was set up correctly.
Consider creating more robust error messaging around APIConnectionError: Error communicating with OpenAI, as this may be a common new user failure point.
large memory footprint introduced by `chromadb` default embeddings
ChromaDB uses sentence-transformers by default for embeddings, which requires torch, which all together makes the footprint of chroma something like ~5GB.
We use OpenAI's text-embedding-ada-002 model offered via chromadb.utils.embedding_functions, but
chromadb enforces the sentence-transformers dependency at this time.
We should find a way around this if chroma doesn't make sentence-transformers an optional dep
context_aware_fillna_df doesn't fill data in example
First check
- I added a descriptive title to this issue.
- I used the GitHub search to try to find a similar issue and didn't find one.
- I searched the Marvin documentation for this issue.
Bug summary
context_aware_fillna works fine in brief manual test.
context_aware_fillna_df turns an integer column into a float and doesn't fill missing values.
Reproduction
import pandas as pd
from marvin.ai_functions.data import context_aware_fillna_df
data = pd.DataFrame(
[
["The Terminator", 1984, None],
["Minority Report", None, "Steven Spielberg"],
["Wall-E", None, "andrew Stanton"],
],
columns=["title", "release_year", "director"],
)
df_filled = context_aware_fillna_df(data)
print(df_filled)
print(type(df_filled))Error
<img width="503" alt="Screenshot 2023-04-21 at 11 29 33 PM" src="https://user-images.githubusercontent.com/7703961/233764622-6ea8dc33-78b5-4d08-979c-588c057e90c6.png">Versions
0.7.7.dev100+ge191037
Additional context
No response
Add outputtransformer
List of hooks that can take strings → return (optional) structured outputs. Could be used to call other bots to evaluate the response
("ResponseTransformer"?)
Support sets in ResponseFormatters
This currently errors:
@ai_fn
def extract_animals(text: str) -> set[str]:
"""Returns a set of all animals mentioned in the text"""
extract_animals('the fox and the dog')
Support Python 3.8
Great question from the livestream - can we support 3.9?
For posterity - the decision to support 3.10 is a legacy of this being an internal project and wanting to use the best and brightest :)
Now that it's being used so widely, I think 3.10 is too high a hurdle.
Typo in Marvin Screenshot
The code snipper shows a function names list_fruits but the function call is list_fruit
Use generators as AI functions
A generator would allow for multi-step execution and the potential for side effects, as user code would be running in sequence with AI execution.
@ai_fn
def my_fn():
x = do_stuff()
ai_result = yield x
y = do_stuff_with_ai_result(ai_result)
another_result = yield yRename bots → bot (minor internal breaking refactor)
Would have to be a 0.7.0 release
Move marvin.bots → marvin.bot # classes for making a Bot work
Move marvin.bots_lab → marvin.bots # example bots
Users would probably not be impacted as most import marvin.Bot.
render markdown links correctly in slack
regex
If I already have the default bots installed, the TUI should not have an install button
First check
- I added a descriptive title to this issue.
- I used the GitHub search to look for a similar issue and didn't find it.
- I searched the Marvin documentation for this feature.
Describe the current behavior
I already have the default bots installed. I click on "Bots". The TUI has a "Install default bots" button, which is confusing. A user might think the bots are not installed and that they need to install each bot individually.
Describe the proposed behavior
The "Install Default Bots" button should only show if the default bots are not installed.
Example Use
No response
Additional context
No response
weird recursion issue with `Document.to_excerpts` with `GitHubRepoLoader`
only appears to happen when no globs are passed in, will need to investigate
seems related to MarvinBaseModel.copy_with_updates
Example
works
prefect_source_code = GitHubRepoLoader( # gimme da source
repo="prefecthq/prefect",
include_globs=["**/*.py"],
exclude_globs=["tests/**/*", "docs/**/*", "**/migrations/**/*"],
)doesnt
prefect_source_code = GitHubRepoLoader( # gimme da source
repo="prefecthq/prefect",
# include_globs=["**/*.py"],
# exclude_globs=["tests/**/*", "docs/**/*", "**/migrations/**/*"],
)fix `plugins` for `ai_fn`
currently, bot_kwargs is set explicitly to {} in the ai_fn_wrapper scope, erasing any plugins set via the ai_fn decorator kwargs
Add temporary loaders
Add a way to temporarily and conveniently add simple knowledge to a bot, including via plugin. E.g. point a bot at a website, document, or other information source, parse it with an appropriate loader, store that in a temporary vector store, and make it available ot the bot.
"Hey bot, index example.com"
"Done! what do you want to know?"
Would provide a much lower-lift interface to exploration than having to go through formal knowledge processing
Does it require ChatGPT/GPT token to work?
Can it work offline?
Prompt Engineering: how to get better results
I love the idea of AI functions!
Can you share some tipps how I can get more reliable results though?
@ai_fn
def top3_words_in_text(text: str, criteria: str=None) -> str:
"""
give me the top 3 most often used words in the text in this format:
[
{ word: "milk", count: 10},
{ word: "fish", count: 5}
]
make sure to sort by count
"""
text = "I really really really love pancakes and cherry from cherry trees"
print(top3_words_in_text(text))
print(top3_words_in_text(text))
print(top3_words_in_text(text))
Gives me back:
[{"word": "cherry", "count": 2}, {"word": "really", "count": 3}, {"word": "love", "count": 1}]
[{"word": "really", "count": 3}, {"word": "cherry", "count": 2}, {"word": "and", "count": 1}]
[{"word": "cherry", "count": 2}, {"word": "really", "count": 3}, {"word": "love", "count": 1}]
allow JSON-style examples in plugin docstring
LLMs seem to do best with examples, and if a plugin should accept JSON, it seems easiest to just show that in the docstring
currently this errors:
In [22]: @plugin
...: def foo():
...: """ {"key": "value"} """
...: pass
...:
In [23]: foo.get_full_description()
---------------------------------------------------------------------------
KeyError Traceback (most recent call last)
Cell In[23], line 1
----> 1 foo.get_full_description()
File ~/src/open-source/marvin/src/marvin/plugins/base.py:48, in Plugin.get_full_description(self)
47 def get_full_description(self) -> str:
---> 48 description = self.description.format(**self.dict())
49 docstring = self.run.__doc__
51 result = inspect.cleandoc(
52 f"""
53 Name: {self.name}
54 Signature: {self._signature}
55 """
56 )
KeyError: '"key"'and this works fine
In [25]: @plugin
...: def foo():
...: """ "key" -> "value" """
...: pass
...:
In [26]: foo.get_full_description()
Out[26]: 'Name: foo\nSignature: ()\n "key" -> "value" 'Observation about LLM Response
I noticed that in a number of cases, these lines in your prompt:
marvin/src/marvin/bots/response_formatters.py
Lines 90 to 91 in 3cc01e4
have an outsized effect on the response from the LLM.
When the response type of an ai_fn wrapped function was str, I got this from gpt-3.5: {"title": "Model", "type": "string"}
And when the response type was dict, I got: {"title": "Model", "type": "object"}.
Failure when return type was string seems more likely. And for context, I captured these outputs directly after the response = await self._call_llm(messages=messages) line.
Perhaps this issue might only be faced with 3.5 since it doesn't always understand highly detailed instructions, but I just wanted to make note of it anyway.
Great job on the code and the tool! Pleasure to read through!
implement prd automation workflow into `marvin` repo
Discord link seems invallid?
I tried to join the discord but the link has expired.
Load existing bots from the CLI
This should load a saved bot, including personality, instructions, plugins, transformers, etc.
marvin chat -b MyBot`bot.save` fails
Enable user customization of instruction templates
Thank you for making this awesome tool available for everyone!
I noticed that the instructions templates of the bots (and its corresponding parameters in the Jinja template) are hardcoded in bots/base.py (defined in INSTRUCTIONS_TEMPLATE). I also noticed similar pattern in ai_fn defined in ai_functions.py. The instructions template for user message (defined in AI_FN_MESSAGE constant) is hardcoded, resulting in users having little control on what user message being sent to the OpenAI model.
I was wondering if there is any plan to allow users to configure their own instruction templates. I feel that the default template can be a bit restrictive sometimes as its longer length does not always go well with the shorter context length in GPT 3.5. Besides, I feel having a different bot instruction template for different use cases may result in a more efficient prompts and better performance.
Thanks :)
Issue with greenlet library on M1 Mac
Having trouble with the greenlet library on an M1 mac. (Searching around I see this issue with M1 macs for other libraries that use greenlet but haven't found a good solution yet. Specific error is:
ValueError: the greenlet library is required to use this function. dlopen(/Users/eball/miniconda3/envs/marvin/lib/python3.10/site-packages/greenlet/_greenlet.cpython-310-darwin.so, 0x0002): tried: '/Users/eball/miniconda3/envs/marvin/lib/python3.10/site-packages/greenlet/_greenlet.cpython-310-darwin.so' (mach-o file, but is an incompatible architecture (have (x86_64), need (arm64e)))
Ideas for workarounds welcome.
create context preprocessor
the list of messages provided to the LLM for a given generation should only contain the most relevant messages to producing that generation:
- most recent messages
- system messages from top of the thread dictating bot attrs
- select previous plugin outputs and other messages deemed "relevant" to the actor the generation is for
primarily this would solve the problem of overflowing the model-specific context window limit, but in general
a context preprocessor could enable arbitrary actors to "hook into" thread content that is relevant to them
this could potentially involve embedding the thread content so we could get top k messages in a thread
AttributeError: 'str' object has no attribute 'match'
First check
- I added a descriptive title to this issue.
- I used the GitHub search to try to find a similar issue and didn't find one.
- I searched the Marvin documentation for this issue.
Bug summary
When trying to use the marvin chat, I get an exception that 'str' doesn't have 'match'
Reproduction
python3 -m pip install marvin;
marvin chat;Error
(base) kevin@Kevins-MacBook-Pro ~ % marvin chat
│ │ data = { │ │
│ │ │ 'id': 'bot_01GY0MH0MTVYS7P8K6SHJSYDK6', │ │
│ │ │ 'name': 'Marvin', │ │
│ │ │ 'personality': 'Marvin is characterized by its immense intelligence, constant sense of depressio'+439, │ │
│ │ │ 'instructions': 'You are part of a library for building AI-powered software called "Marvin". One '+907, │ │
│ │ │ 'instructions_template': 'Your name is: {{ name }}\n\nYour instructions tell you how to respond to a message'+1036, │ │
│ │ │ 'plugins': [ │ │
│ │ │ │ { │ │
│ │ │ │ │ 'name': 'list_all_bots', │ │
│ │ │ │ │ 'description': '\n This plugin lets you look up the names and descriptions of all bots.\n ', │ │
│ │ │ │ │ 'discriminator': 'list_all_bots' │ │
│ │ │ │ }, │ │
│ │ │ │ { │ │
│ │ │ │ │ 'name': 'get_bot_details', │ │
│ │ │ │ │ 'description': '\n This plugin lets you look up the details of a single bot, including its\n '+55, │ │
│ │ │ │ │ 'discriminator': 'get_bot_details' │ │
│ │ │ │ }, │ │
│ │ │ │ { │ │
│ │ │ │ │ 'name': 'create_bot', │ │
│ │ │ │ │ 'description': '\n Creates a bot with the given name, description, personality, and\n instru'+226, │ │
│ │ │ │ │ 'discriminator': 'create_bot' │ │
│ │ │ │ }, │ │
│ │ │ │ { │ │
│ │ │ │ │ 'name': 'update_bot', │ │
│ │ │ │ │ 'description': "\n This plugin can be used to update a bot's description, personality, or\n "+109, │ │
│ │ │ │ │ 'discriminator': 'update_bot' │ │
│ │ │ │ }, │ │
│ │ │ │ { │ │
│ │ │ │ │ 'name': 'delete_bot', │ │
│ │ │ │ │ 'description': '\n Deletes the bot with the given name. Names must be an exact match, includin'+208, │ │
│ │ │ │ │ 'discriminator': 'delete_bot' │ │
│ │ │ │ } │ │
│ │ │ ], │ │
│ │ │ 'input_transformers': [], │ │
│ │ │ 'description': 'The Genuine People Personality you know and love.\n\nMarvin can also help you crea'+34 │ │
│ │ } │ │
│ ╰────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯ │
│ │
│ in pydantic.main.validate_model:1076 │
│ │
│ in pydantic.fields.ModelField.validate:884 │
│ │
│ in pydantic.fields.ModelField._validate_singleton:1101 │
│ │
│ in pydantic.fields.ModelField._apply_validators:1148 │
│ │
│ in pydantic.class_validators._generic_validator_basic.lambda13:318 │
│ │
│ in pydantic.types.ConstrainedStr.validate:433 │
╰─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
AttributeError: 'str' object has no attribute 'match'Versions
(base) kevin@Kevins-MacBook-Pro ~ % marvin version
WARNING: `openai_api_key` is not set. Some features may not work.
[04/14/23 12:45:45] INFO marvin.marvin: Using OpenAI model "gpt-3.5-turbo" logging.py:50
0.7.5
Additional context
.
Installation failing
I didpip3 install marvin. The installation failed. Here is the stacktrace :
pip3 install marvin
Collecting marvin
Using cached marvin-0.6.1-py3-none-any.whl (78 kB)
Collecting aiofiles~=23.1.0
Using cached aiofiles-23.1.0-py3-none-any.whl (14 kB)
Collecting aiosqlite~=0.18.0
Using cached aiosqlite-0.18.0-py3-none-any.whl (15 kB)
Collecting asyncpg~=0.27.0
Using cached asyncpg-0.27.0-cp311-cp311-macosx_11_0_arm64.whl (608 kB)
Collecting cloudpickle~=2.2.1
Using cached cloudpickle-2.2.1-py3-none-any.whl (25 kB)
Collecting datamodel-code-generator~=0.17.1
Using cached datamodel_code_generator-0.17.2-py3-none-any.whl (75 kB)
Collecting fastapi~=0.89.1
Using cached fastapi-0.89.1-py3-none-any.whl (55 kB)
Collecting httpx~=0.23.3
Using cached httpx-0.23.3-py3-none-any.whl (71 kB)
Collecting jinja2~=3.1.2
Using cached Jinja2-3.1.2-py3-none-any.whl (133 kB)
Collecting langchain>=0.0.103
Using cached langchain-0.0.128-py3-none-any.whl (465 kB)
Collecting nest-asyncio~=1.5.6
Using cached nest_asyncio-1.5.6-py3-none-any.whl (5.2 kB)
Collecting openai~=0.27.0
Using cached openai-0.27.2-py3-none-any.whl (70 kB)
Collecting pendulum~=2.1.2
Using cached pendulum-2.1.2.tar.gz (81 kB)
Installing build dependencies ... done
Getting requirements to build wheel ... done
ERROR: Exception:
Traceback (most recent call last):
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/cli/base_command.py", line 160, in exc_logging_wrapper
status = run_func(*args)
^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/cli/req_command.py", line 247, in wrapper
return func(self, options, args)
^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/commands/install.py", line 400, in run
requirement_set = resolver.resolve(
^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/resolution/resolvelib/resolver.py", line 92, in resolve
result = self._result = resolver.resolve(
^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_vendor/resolvelib/resolvers.py", line 481, in resolve
state = resolution.resolve(requirements, max_rounds=max_rounds)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_vendor/resolvelib/resolvers.py", line 373, in resolve
failure_causes = self._attempt_to_pin_criterion(name)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_vendor/resolvelib/resolvers.py", line 213, in _attempt_to_pin_criterion
criteria = self._get_updated_criteria(candidate)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_vendor/resolvelib/resolvers.py", line 204, in _get_updated_criteria
self._add_to_criteria(criteria, requirement, parent=candidate)
File "/opt/homebrew/lib/python3.11/site-packages/pip/_vendor/resolvelib/resolvers.py", line 172, in _add_to_criteria
if not criterion.candidates:
File "/opt/homebrew/lib/python3.11/site-packages/pip/_vendor/resolvelib/structs.py", line 151, in bool
return bool(self._sequence)
^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/resolution/resolvelib/found_candidates.py", line 155, in bool
return any(self)
^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/resolution/resolvelib/found_candidates.py", line 143, in
return (c for c in iterator if id(c) not in self._incompatible_ids)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/resolution/resolvelib/found_candidates.py", line 47, in _iter_built
candidate = func()
^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/resolution/resolvelib/factory.py", line 206, in _make_candidate_from_link
self._link_candidate_cache[link] = LinkCandidate(
^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/resolution/resolvelib/candidates.py", line 297, in init
super().init(
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/resolution/resolvelib/candidates.py", line 162, in init
self.dist = self._prepare()
^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/resolution/resolvelib/candidates.py", line 231, in _prepare
dist = self._prepare_distribution()
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/resolution/resolvelib/candidates.py", line 308, in _prepare_distribution
return preparer.prepare_linked_requirement(self._ireq, parallel_builds=True)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/operations/prepare.py", line 491, in prepare_linked_requirement
return self._prepare_linked_requirement(req, parallel_builds)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/operations/prepare.py", line 577, in _prepare_linked_requirement
dist = _get_prepared_distribution(
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/operations/prepare.py", line 69, in _get_prepared_distribution
abstract_dist.prepare_distribution_metadata(
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/distributions/sdist.py", line 48, in prepare_distribution_metadata
self._install_build_reqs(finder)
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/distributions/sdist.py", line 118, in _install_build_reqs
build_reqs = self._get_build_requires_wheel()
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/distributions/sdist.py", line 95, in _get_build_requires_wheel
return backend.get_requires_for_build_wheel()
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_internal/utils/misc.py", line 685, in get_requires_for_build_wheel
return super().get_requires_for_build_wheel(config_settings=cs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_vendor/pep517/wrappers.py", line 173, in get_requires_for_build_wheel
return self._call_hook('get_requires_for_build_wheel', {
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/lib/python3.11/site-packages/pip/_vendor/pep517/wrappers.py", line 319, in _call_hook
raise BackendUnavailable(data.get('traceback', ''))
pip._vendor.pep517.wrappers.BackendUnavailable: Traceback (most recent call last):
File "/opt/homebrew/lib/python3.11/site-packages/pip/_vendor/pep517/in_process/_in_process.py", line 77, in _build_backend
obj = import_module(mod_path)
^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/homebrew/Cellar/[email protected]/3.11.2_1/Frameworks/Python.framework/Versions/3.11/lib/python3.11/importlib/init.py", line 126, in import_module
return _bootstrap._gcd_import(name[level:], package, level)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "", line 1206, in _gcd_import
File "", line 1178, in _find_and_load
File "", line 1128, in _find_and_load_unlocked
File "", line 241, in _call_with_frames_removed
File "", line 1206, in _gcd_import
File "", line 1178, in _find_and_load
File "", line 1128, in _find_and_load_unlocked
File "", line 241, in _call_with_frames_removed
File "", line 1206, in _gcd_import
File "", line 1178, in _find_and_load
File "", line 1128, in _find_and_load_unlocked
File "", line 241, in _call_with_frames_removed
File "", line 1206, in _gcd_import
File "", line 1178, in _find_and_load
File "", line 1142, in _find_and_load_unlocked
ModuleNotFoundError: No module named 'poetry'
NoMatches: No nodes match
Hi,
thanks for this project. I ran marvin chat and entered my API key. However, subsequent requests seem to error out.
Here is an example:
Prompt = "What is something only a bot would say?"
I get the exception NoMatches: No nodes match. Here is a fuller view:
│ ╭───────────────────────────────────────────────────────── locals ──────────────────────────────────────────────────────────╮ │
│ │ bot = Bot( │ │
│ │ │ id='bot_01GXJJX4DND5CZWTN1G10NDB22', │ │
│ │ │ name='Marvin', │ │
│ │ │ description='The Genuine People Personality you know and love.\n\nMarvin can also help you crea'+34, │ │
│ │ │ plugins=[ │ │
│ │ │ │ list_all_bots(name='list_all_bots', discriminator='list_all_bots'), │ │
│ │ │ │ get_bot_details(name='get_bot_details', discriminator='get_bot_details'), │ │
│ │ │ │ create_bot(name='create_bot', discriminator='create_bot'), │ │
│ │ │ │ update_bot(name='update_bot', discriminator='update_bot'), │ │
│ │ │ │ delete_bot(name='delete_bot', discriminator='delete_bot') │ │
│ │ │ ] │ │
│ │ ) │ │
│ │ bot_response = BotResponse(classes={'response', 'bot-response'}, pseudo_classes={'enabled'}) │ │
│ │ bot_responses = <DOMQuery MainScreen(pseudo_classes={'focus-within', 'enabled'}) filter='BotResponse'> │ │
│ │ event = Submitted() │ │
│ │ response = BotResponse( │ │
│ │ │ id='msg_01GXJK6ZJSE5VQFXEETABGWP2F', │ │
│ │ │ role='bot', │ │
│ │ │ content="Well, that's an interesting question. As an AI language model, I can tell you th"+240, │ │
│ │ │ name='Marvin', │ │
│ │ │ timestamp=DateTime(2023, 4, 9, 8, 48, 6, 233439, tzinfo=Timezone('UTC')), │ │
│ │ │ bot_id='bot_01GXJJX4DND5CZWTN1G10NDB22', │ │
│ │ │ data={}, │ │
│ │ │ parsed_content="Well, that's an interesting question. As an AI language model, I can tell you th"+240 │ │
│ │ ) │ │
│ │ self = MainScreen(pseudo_classes={'focus-within', 'enabled'}) │ │
│ ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯ │
│ │
│ /home/lab/env/lib/python3.9/site-packages/textual/css/query.py:285 in last │
│ │
│ 282 │ │ │ The matching Widget. │
│ 283 │ │ """ │
│ 284 │ │ if not self.nodes: │
│ ❱ 285 │ │ │ raise NoMatches(f"No nodes match {self!r}") │
│ 286 │ │ last = self.nodes[-1] │
│ 287 │ │ if expect_type is not None and not isinstance(last, expect_type): │
│ 288 │ │ │ raise WrongType( │
│ │
│ ╭─────────────────────────────────────────────── locals ───────────────────────────────────────────────╮ │
│ │ expect_type = None │ │
│ │ self = <DOMQuery MainScreen(pseudo_classes={'focus-within', 'enabled'}) filter='BotResponse'> │ │
│ ╰──────────────────────────────────────────────────────────────────────────────────────────────────────╯ │
╰──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
NoMatches: No nodes match <DOMQuery MainScreen(pseudo_classes={'focus-within', 'enabled'}) filter='BotResponse'>
So I guess the API connection and almost everything works, but there might be a small bug in the rendering?!
Running this on Debian 11, Python 3.9.2, and using the latest marvin version from pip.
Thanks.
Allow AI functions to generate source code
If AI functions could generate source code and execute it, they could do significantly more on the local machine. However, this would constitute untrusted code and would require controls to avoid unintended execution.
incorrect API is being delivered to the OpenAI server.
- given the following: sk-wJOK3***************************************fjYz
- I get:
AuthenticationError: Incorrect API key provided: sk-nuLlk***************************************dhN2. You can find your API key at https://platform.openai.com/account/api-keys.
probably due I have past config from other API in my envs.
OperationalError Linux
First check
- I added a descriptive title to this issue.
- I used the GitHub search to try to find a similar issue and didn't find one.
- I searched the Marvin documentation for this issue.
Bug summary
Hello!
I installed marvin with pip install marvin (without a conda environment) on pop os 22.04. I'm getting the following error when running marvin chat:
OperationalError: (sqlite3.OperationalError) no such column: bot_config.created_at
[SQL: SELECT bot_config.created_at, bot_config.updated_at, bot_config.plugins,
bot_config.input_transformers, bot_config.id, bot_config.name, bot_config.personality,
bot_config.instructions, bot_config.instructions_template, bot_config.description,
bot_config.profile_picture
FROM bot_config
WHERE bot_config.name = ?
LIMIT ? OFFSET ?]
[parameters: ('Marvin', 1, 0)]
(Background on this error at: https://sqlalche.me/e/14/e3q8)
Any idea how to fix this?
Cheers
Reproduction
In pop os 22.04, run:
pip install marvin
marvin chat
### Error
```python3
│ ╰───────────────────────────────────────────────────────────────────────────────────────╯ │
│ │
│ /home/arnle/miniconda3/lib/python3.9/site-packages/sqlalchemy/dialects/sqlite/aiosqlite.p │
│ y:100 in execute │
│ │
│ 97 │ │ │ else: │
│ 98 │ │ │ │ self._cursor = _cursor │
│ 99 │ │ except Exception as error: │
│ ❱ 100 │ │ │ self._adapt_connection._handle_exception(error) │
│ 101 │ │
│ 102 │ def executemany(self, operation, seq_of_parameters): │
│ 103 │ │ try: │
│ │
│ ╭─────────────────────────────────────── locals ────────────────────────────────────────╮ │
│ │ _cursor = <aiosqlite.cursor.Cursor object at 0x7fc0812df730> │ │
│ │ operation = 'SELECT bot_config.created_at, bot_config.updated_at, │ │
│ │ bot_config.plugins, bot_con'+251 │ │
│ │ parameters = ('Marvin', 1, 0) │ │
│ │ self = <sqlalchemy.dialects.sqlite.aiosqlite.AsyncAdapt_aiosqlite_cursor object │ │
│ │ at 0x7fc081312dc0> │ │
│ ╰───────────────────────────────────────────────────────────────────────────────────────╯ │
│ │
│ /home/arnle/miniconda3/lib/python3.9/site-packages/sqlalchemy/dialects/sqlite/aiosqlite.p │
│ y:228 in _handle_exception │
│ │
│ 225 │ │ │ │ from_=error, │
│ 226 │ │ │ ) │
│ 227 │ │ else: │
│ ❱ 228 │ │ │ raise error │
│ 229 │
│ 230 │
│ 231 class AsyncAdaptFallback_aiosqlite_connection(AsyncAdapt_aiosqlite_connection): │
│ │
│ ╭────────────────────────────────────── locals ──────────────────────────────────────╮ │
│ │ error = OperationalError('no such column: bot_config.created_at') │ │
│ │ self = <AdaptedConnection <Connection(Thread-6, started daemon 140464776869440)>> │ │
│ ╰────────────────────────────────────────────────────────────────────────────────────╯ │
│ │
│ /home/arnle/miniconda3/lib/python3.9/site-packages/sqlalchemy/dialects/sqlite/aiosqlite.p │
│ y:82 in execute │
│ │
│ 79 │ │ │ if parameters is None: │
│ 80 │ │ │ │ self.await_(_cursor.execute(operation)) │
│ 81 │ │ │ else: │
│ ❱ 82 │ │ │ │ self.await_(_cursor.execute(operation, parameters)) │
│ 83 │ │ │ │
│ 84 │ │ │ if _cursor.description: │
│ 85 │ │ │ │ self.description = _cursor.description │
│ │
│ ╭─────────────────────────────────────── locals ────────────────────────────────────────╮ │
│ │ _cursor = <aiosqlite.cursor.Cursor object at 0x7fc0812df730> │ │
│ │ operation = 'SELECT bot_config.created_at, bot_config.updated_at, │ │
│ │ bot_config.plugins, bot_con'+251 │ │
│ │ parameters = ('Marvin', 1, 0) │ │
│ │ self = <sqlalchemy.dialects.sqlite.aiosqlite.AsyncAdapt_aiosqlite_cursor object │ │
│ │ at 0x7fc081312dc0> │ │
│ ╰───────────────────────────────────────────────────────────────────────────────────────╯ │
│ │
│ /home/arnle/miniconda3/lib/python3.9/site-packages/sqlalchemy/util/_concurrency_py3k.py:6 │
│ 8 in await_only │
│ │
│ 65 │ # a coroutine to run. Once the awaitable is done, the driver greenlet │
│ 66 │ # switches back to this greenlet with the result of awaitable that is │
│ 67 │ # then returned to the caller (or raised as error) │
│ ❱ 68 │ return current.driver.switch(awaitable) │
│ 69 │
│ 70 │
│ 71 def await_fallback(awaitable: Coroutine) -> Any: │
│ │
│ ╭────────────────────────────────────── locals ──────────────────────────────────────╮ │
│ │ awaitable = <coroutine object Cursor.execute at 0x7fc08138ad40> │ │
│ │ current = <_AsyncIoGreenlet object at 0x7fc08126d2c0 (otid=0x7fc0a6d45330) dead> │ │
│ ╰────────────────────────────────────────────────────────────────────────────────────╯ │
│ │
│ /home/arnle/miniconda3/lib/python3.9/site-packages/sqlalchemy/util/_concurrency_py3k.py:1 │
│ 21 in greenlet_spawn │
│ │
│ 118 │ │ │ try: │
│ 119 │ │ │ │ # wait for a coroutine from await_only and then return its │
│ 120 │ │ │ │ # result back to it. │
│ ❱ 121 │ │ │ │ value = await result │
│ 122 │ │ │ except BaseException: │
│ 123 │ │ │ │ # this allows an exception to be raised within │
│ 124 │ │ │ │ # the moderated greenlet so that it can continue │
│ │
│ ╭─────────────────────────────────────── locals ────────────────────────────────────────╮ │
│ │ _require_await = False │ │
│ │ args = (<sqlalchemy.sql.selectable.Select object at 0x7fc081bf5c70>,) │ │
│ │ context = <_AsyncIoGreenlet object at 0x7fc08126d2c0 (otid=0x7fc0a6d45330) │ │
│ │ dead> │ │
│ │ fn = <bound method Session.execute of <sqlmodel.orm.session.Session │ │
│ │ object at 0x7fc081c7fa90>> │ │
│ │ kwargs = { │ │
│ │ │ 'params': None, │ │
│ │ │ 'execution_options': immutabledict({'prebuffer_rows': True}), │ │
│ │ │ 'bind_arguments': None │ │
│ │ } │ │
│ │ result = <coroutine object Cursor.execute at 0x7fc08138ad40> │ │
│ │ switch_occurred = True │ │
│ │ value = <aiosqlite.cursor.Cursor object at 0x7fc0812df730> │ │
│ ╰───────────────────────────────────────────────────────────────────────────────────────╯ │
│ │
│ /home/arnle/miniconda3/lib/python3.9/site-packages/aiosqlite/cursor.py:37 in execute │
│ │
│ 34 │ │ """Execute the given query.""" │
│ 35 │ │ if parameters is None: │
│ 36 │ │ │ parameters = [] │
│ ❱ 37 │ │ await self._execute(self._cursor.execute, sql, parameters) │
│ 38 │ │ return self │
│ 39 │ │
│ 40 │ async def executemany( │
│ │
│ ╭─────────────────────────────────────── locals ────────────────────────────────────────╮ │
│ │ parameters = ('Marvin', 1, 0) │ │
│ │ self = <aiosqlite.cursor.Cursor object at 0x7fc0812df730> │ │
│ │ sql = 'SELECT bot_config.created_at, bot_config.updated_at, │ │
│ │ bot_config.plugins, bot_con'+251 │ │
│ ╰───────────────────────────────────────────────────────────────────────────────────────╯ │
│ │
│ /home/arnle/miniconda3/lib/python3.9/site-packages/aiosqlite/cursor.py:31 in _execute │
│ │
│ 28 │ │
│ 29 │ async def _execute(self, fn, *args, **kwargs): │
│ 30 │ │ """Execute the given function on the shared connection's thread.""" │
│ ❱ 31 │ │ return await self._conn._execute(fn, *args, **kwargs) │
│ 32 │ │
│ 33 │ async def execute(self, sql: str, parameters: Iterable[Any] = None) -> "Cursor" │
│ 34 │ │ """Execute the given query.""" │
│ │
│ ╭─────────────────────────────────────── locals ────────────────────────────────────────╮ │
│ │ args = ( │ │
│ │ │ 'SELECT bot_config.created_at, bot_config.updated_at, │ │
│ │ bot_config.plugins, bot_con'+251, │ │
│ │ │ ('Marvin', 1, 0) │ │
│ │ ) │ │
│ │ fn = <built-in method execute of sqlite3.Cursor object at 0x7fc08137d180> │ │
│ │ kwargs = {} │ │
│ │ self = <aiosqlite.cursor.Cursor object at 0x7fc0812df730> │ │
│ ╰───────────────────────────────────────────────────────────────────────────────────────╯ │
│ │
│ /home/arnle/miniconda3/lib/python3.9/site-packages/aiosqlite/core.py:137 in _execute │
│ │
│ 134 │ │ │
│ 135 │ │ self._tx.put_nowait((future, function)) │
│ 136 │ │ │
│ ❱ 137 │ │ return await future │
│ 138 │ │
│ 139 │ async def _connect(self) -> "Connection": │
│ 140 │ │ """Connect to the actual sqlite database.""" │
│ │
│ ╭─────────────────────────────────────── locals ────────────────────────────────────────╮ │
│ │ args = ( │ │
│ │ │ 'SELECT bot_config.created_at, bot_config.updated_at, │ │
│ │ bot_config.plugins, bot_con'+251, │ │
│ │ │ ('Marvin', 1, 0) │ │
│ │ ) │ │
│ │ fn = <built-in method execute of sqlite3.Cursor object at 0x7fc08137d180> │ │
│ │ function = functools.partial(<built-in method execute of sqlite3.Cursor object at │ │
│ │ 0x7fc08137d180>, 'SELECT bot_config.created_at, bot_config.updated_at, │ │
│ │ bot_config.plugins, bot_config.input_transformers, bot_config.id, │ │
│ │ bot_config.name, bot_config.personality, bot_config.instructions, │ │
│ │ bot_config.instructions_template, bot_config.description, │ │
│ │ bot_config.profile_picture \nFROM bot_config \nWHERE bot_config.name = ?\n │ │
│ │ LIMIT ? OFFSET ?', ('Marvin', 1, 0)) │ │
│ │ future = <Future finished exception=OperationalError('no such column: │ │
│ │ bot_config.created_at')> │ │
│ │ kwargs = {} │ │
│ │ self = <Connection(Thread-6, started daemon 140464776869440)> │ │
│ ╰───────────────────────────────────────────────────────────────────────────────────────╯ │
│ │
│ /home/arnle/miniconda3/lib/python3.9/asyncio/futures.py:284 in __await__ │
│ │
│ 281 │ def __await__(self): │
│ 282 │ │ if not self.done(): │
│ 283 │ │ │ self._asyncio_future_blocking = True │
│ ❱ 284 │ │ │ yield self # This tells Task to wait for completion. │
│ 285 │ │ if not self.done(): │
│ 286 │ │ │ raise RuntimeError("await wasn't used with future") │
│ 287 │ │ return self.result() # May raise too. │
│ │
│ ╭─────────────────────────────────────── locals ────────────────────────────────────────╮ │
│ │ self = <Future finished exception=OperationalError('no such column: │ │
│ │ bot_config.created_at')> │ │
│ ╰───────────────────────────────────────────────────────────────────────────────────────╯ │
│ │
│ /home/arnle/miniconda3/lib/python3.9/asyncio/tasks.py:328 in __wakeup │
│ │
│ 325 │ │
│ 326 │ def __wakeup(self, future): │
│ 327 │ │ try: │
│ ❱ 328 │ │ │ future.result() │
│ 329 │ │ except BaseException as exc: │
│ 330 │ │ │ # This may also be a cancellation. │
│ 331 │ │ │ self.__step(exc) │
│ │
│ ╭─────────────────────────────────────── locals ────────────────────────────────────────╮ │
│ │ future = <Future finished exception=OperationalError('no such column: │ │
│ │ bot_config.created_at')> │ │
│ │ self = None │ │
│ ╰───────────────────────────────────────────────────────────────────────────────────────╯ │
│ │
│ /home/arnle/miniconda3/lib/python3.9/asyncio/futures.py:201 in result │
│ │
│ 198 │ │ │ raise exceptions.InvalidStateError('Result is not ready.') │
│ 199 │ │ self.__log_traceback = False │
│ 200 │ │ if self._exception is not None: │
│ ❱ 201 │ │ │ raise self._exception │
│ 202 │ │ return self._result │
│ 203 │ │
│ 204 │ def exception(self): │
│ │
│ ╭─────────────────────────────────────── locals ────────────────────────────────────────╮ │
│ │ self = <Future finished exception=OperationalError('no such column: │ │
│ │ bot_config.created_at')> │ │
│ ╰───────────────────────────────────────────────────────────────────────────────────────╯ │
│ │
│ /home/arnle/miniconda3/lib/python3.9/site-packages/aiosqlite/core.py:110 in run │
│ │
│ 107 │ │ │ │ break │
│ 108 │ │ │ try: │
│ 109 │ │ │ │ LOG.debug("executing %s", function) │
│ ❱ 110 │ │ │ │ result = function() │
│ 111 │ │ │ │ LOG.debug("operation %s completed", function) │
│ 112 │ │ │ │ │
│ 113 │ │ │ │ def set_result(fut, result): │
│ │
│ ╭─────────────────────────────────────── locals ────────────────────────────────────────╮ │
│ │ function = functools.partial(<built-in method close of sqlite3.Connection object │ │
│ │ at 0x7fc081ca56c0>) │ │
│ │ future = <Future finished result=None> │ │
│ │ result = None │ │
│ │ self = <Connection(Thread-6, started daemon 140464776869440)> │ │
│ │ set_exception = <function Connection.run.<locals>.set_exception at 0x7fc08128b430> │ │
│ │ set_result = <function Connection.run.<locals>.set_result at 0x7fc08128b0d0> │ │
│ ╰───────────────────────────────────────────────────────────────────────────────────────╯ │
╰───────────────────────────────────────────────────────────────────────────────────────────╯
OperationalError: (sqlite3.OperationalError) no such column: bot_config.created_at
[SQL: SELECT bot_config.created_at, bot_config.updated_at, bot_config.plugins,
bot_config.input_transformers, bot_config.id, bot_config.name, bot_config.personality,
bot_config.instructions, bot_config.instructions_template, bot_config.description,
bot_config.profile_picture
FROM bot_config
WHERE bot_config.name = ?
LIMIT ? OFFSET ?]
[parameters: ('Marvin', 1, 0)]
(Background on this error at: https://sqlalche.me/e/14/e3q8)
### Versions
```Text
Same error when running `marvin version`
Additional context
Pop OS 22.04
Store plugin outputs in a loader
First check
- I added a descriptive title to this issue.
- I used the GitHub search to look for a similar issue and didn't find it.
- I searched the Marvin documentation for this feature.
Describe the current behavior
When you create a plugin and the output is too big for the engine, it will just fail.
It would be nice that the output is stored as a vector instead, so that it can be used for the next steps, instead of failing.
I tried to find similar issues, I did find similar ones but they did not convey this exactly I think.
A similar issue is #162
Describe the proposed behavior
When you declare a plugin, instead of feeding the whole output as a message, have a way of storing the result instead, so it can be used as a knowledge base instead of directly being consumed and potentially failing because of its length.
Example Use
No response
Additional context
No response
Error while opening the server.
error on commit ba4fe3b
(base) PS C:\Users\jalba\Sage> marvin server start
[04/07/23 16:33:58] INFO marvin.marvin: Using OpenAI model "gpt-3.5-turbo" logging.py:50
INFO:marvin.marvin:[default on default]Using OpenAI model "gpt-3.5-turbo"[/]
INFO: Started server process [1952]
INFO: Waiting for application startup.
INFO: Application startup complete.
INFO: Uvicorn running on http://127.0.0.1:4200 (Press CTRL+C to quit)
INFO: 127.0.0.1:56318 - "GET / HTTP/1.1" 307 Temporary Redirect
INFO: 127.0.0.1:56318 - "GET /docs HTTP/1.1" 200 OK
INFO: 127.0.0.1:56318 - "GET /openapi.json HTTP/1.1" 500 Internal Server Error
ERROR: Exception in ASGI application
Traceback (most recent call last):
File "C:\Users\jalba\miniconda3\lib\site-packages\uvicorn\protocols\http\httptools_impl.py", line 436, in run_asgi
result = await app( # type: ignore[func-returns-value]
File "C:\Users\jalba\miniconda3\lib\site-packages\uvicorn\middleware\proxy_headers.py", line 78, in __call__
return await self.app(scope, receive, send)
File "C:\Users\jalba\miniconda3\lib\site-packages\fastapi\applications.py", line 276, in __call__
await super().__call__(scope, receive, send)
File "C:\Users\jalba\miniconda3\lib\site-packages\starlette\applications.py", line 122, in __call__
await self.middleware_stack(scope, receive, send)
File "C:\Users\jalba\miniconda3\lib\site-packages\starlette\middleware\errors.py", line 184, in __call__
raise exc
File "C:\Users\jalba\miniconda3\lib\site-packages\starlette\middleware\errors.py", line 162, in __call__
await self.app(scope, receive, _send)
File "C:\Users\jalba\miniconda3\lib\site-packages\starlette\middleware\cors.py", line 84, in __call__
await self.app(scope, receive, send)
File "C:\Users\jalba\miniconda3\lib\site-packages\starlette\middleware\exceptions.py", line 79, in __call__
raise exc
File "C:\Users\jalba\miniconda3\lib\site-packages\starlette\middleware\exceptions.py", line 68, in __call__
await self.app(scope, receive, sender)
File "C:\Users\jalba\miniconda3\lib\site-packages\fastapi\middleware\asyncexitstack.py", line 21, in __call__
raise e
File "C:\Users\jalba\miniconda3\lib\site-packages\fastapi\middleware\asyncexitstack.py", line 18, in __call__
await self.app(scope, receive, send)
File "C:\Users\jalba\miniconda3\lib\site-packages\starlette\routing.py", line 718, in __call__
await route.handle(scope, receive, send)
File "C:\Users\jalba\miniconda3\lib\site-packages\starlette\routing.py", line 276, in handle
await self.app(scope, receive, send)
File "C:\Users\jalba\miniconda3\lib\site-packages\starlette\routing.py", line 66, in app
response = await func(request)
File "C:\Users\jalba\miniconda3\lib\site-packages\fastapi\applications.py", line 231, in openapi
return JSONResponse(self.openapi())
File "C:\Users\jalba\miniconda3\lib\site-packages\fastapi\applications.py", line 206, in openapi
self.openapi_schema = get_openapi(
File "C:\Users\jalba\miniconda3\lib\site-packages\fastapi\openapi\utils.py", line 423, in get_openapi
definitions = get_model_definitions(
File "C:\Users\jalba\miniconda3\lib\site-packages\fastapi\utils.py", line 44, in get_model_definitions
m_schema, m_definitions, m_nested_models = model_process_schema(
File "pydantic\schema.py", line 581, in pydantic.schema.model_process_schema
File "pydantic\schema.py", line 622, in pydantic.schema.model_type_schema
File "pydantic\schema.py", line 255, in pydantic.schema.field_schema
File "pydantic\schema.py", line 462, in pydantic.schema.field_type_schema
File "pydantic\schema.py", line 850, in pydantic.schema.field_singleton_schema
File "pydantic\schema.py", line 701, in pydantic.schema.field_singleton_sub_fields_schema
File "pydantic\schema.py", line 527, in pydantic.schema.field_type_schema
File "pydantic\schema.py", line 924, in pydantic.schema.field_singleton_schema
File "C:\Users\jalba\miniconda3\lib\abc.py", line 123, in __subclasscheck__
return _abc_subclasscheck(cls, subclass)
TypeError: issubclass() arg 1 must be a class
Add pre/post message formatting hooks
In order to avoid situations where bots like utility bots misinterpret a message as an instruction, rather than text to process, introduce processing hooks to manipulate messages. For example, "The text to process is: {message}" could be the modification applied to utility bot inputs.
I don't have the ability to visit external websites
Homepage says the following about bot plugins.
Plugins¶
Plugins allow bots to access new information and functionality. By default, bots have plugins that let them browse the internet, visit URLs, and run simple calculations.
When I am trying to ask bot to access a website and ask it questions, it tells me following:
"
│ I'm sorry for the confusion earlier. As an AI language model, I don't have the ability to visit external websites. │
│ However, I can still provide some feedback based on what you tell me about your website or any other website you're │
│ interested in.
"
Is bots allowed to browse the internet and visit URLs?
Creating a custom bot and then using in the TUI fails
Context:
marvin version 0.7.4
Created bot based on example in the docs.
Script:
from marvin import Bot
import asyncio
async def create_ford_bot():
ford_bot = Bot(
name="Ford",
personality="Can't get the hang of Thursdays",
instructions="Always responds as if researching an article for the Hitchhiker's Guide to the Galaxy",
)
await ford_bot.save()
asyncio.run(create_ford_bot())Run the file.
Bot is found in list.
Then: marvin chat
Results in error:
│
│ 19 │ │ RenderError: If the object can not be rendered. │
│ 20 │ """ │
│ 21 │ if not is_renderable(renderable): │
│ ❱ 22 │ │ raise RenderError( │
│ 23 │ │ │ f"unable to render {renderable!r}; a string, Text, or other Rich renderable │
│ 24 │ │ ) │
│ 25 │
│ │
│ ╭───── locals ──────╮ │
│ │ renderable = None │ │
│ ╰───────────────────╯ │
╰─────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
RenderError: unable to render None; a string, Text, or other Rich renderable is required
Marvin pip install issues in windows 11
I tried installing marvin as per the instructions(in a new virtual environment). During the installation I got some errors before pip installation was somewhat successfully completed:
But when I run "marvin chat" command in the command prompt of my VS Code then nothing happens and I don't receive an error either.
I know this is not much information but this is all I have.
Thank you in advance for any assistance :)
marvin setup-openai fails with "cannot find context for fork"
Traceback (most recent call last):
File "C:\Users\ccrow\AppData\Local\Programs\Python\Python310\lib\runpy.py", line 196, in run_module_as_main
return run_code(code, main_globals, None,
File "C:\Users\ccrow\AppData\Local\Programs\Python\Python310\lib\runpy.py", line 86, in run_code
exec(code, run_globals)
File "C:\Users\ccrow\AppData\Local\Programs\Python\Python310\Scripts\marvin.exe_main.py", line 4, in
File "C:\Users\ccrow\AppData\Local\Programs\Python\Python310\lib\site-packages\marvin_init.py", line 18, in
from marvin.utilities.logging import get_logger
File "C:\Users\ccrow\AppData\Local\Programs\Python\Python310\lib\site-packages\marvin\utilities_init.py", line 1, in
from . import logging, async_utils, types, strings, collections, models
File "C:\Users\ccrow\AppData\Local\Programs\Python\Python310\lib\site-packages\marvin\utilities\async_utils.py", line 10, in
process_pool = concurrent.futures.ProcessPoolExecutor(mp_context=mp.get_context("fork"))
File "C:\Users\ccrow\AppData\Local\Programs\Python\Python310\lib\multiprocessing\context.py", line 243, in get_context
return super().get_context(method)
File "C:\Users\ccrow\AppData\Local\Programs\Python\Python310\lib\multiprocessing\context.py", line 193, in get_context
raise ValueError('cannot find context for %r' % method) from None
ValueError: cannot find context for 'fork'
JS/TS version & a few questions
This is phenomenal 🔥
I'd love to help add TS support for the core ai_fn decorator.
- What are your internal plans for JS/TS support?
- How much is tied to Marvin the company & platform vs being an agnostic tool like LangChain that can call into any hosted or local LLM provider?
- Any advice on where to start either in this monorepo or separately on a JS/TS port of the
ai_fnfunctionality in particular?
Thanks && congrats on your public launch! 🙏
Confusing error message when trying to use GPT-4 without access
“InvalidRequestError: The model: gpt-4 does not exist”.
We could trap this and provide a more informative message.
Get marvin to import correctly when running other python processes on multiple threads
When import marvin into a streamlit app process, it kept conflicting with the main thread running the streamlit app. I worked around it by setting thread conditions within marvin to workaround this. Would love this to be a default.
# __init__.py
from importlib.metadata import version as _get_version
# load nest_asyncio
import nest_asyncio as _nest_asyncio
import asyncio as _asyncio
import threading
import asyncio as _asyncio
import threading
if threading.current_thread() == threading.main_thread():
_nest_asyncio.apply()
else:
_asyncio.set_event_loop(_asyncio.new_event_loop())
_nest_asyncio.apply()Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.