Comments (14)
The default model for gpt-migrate is gpt-4-32k
. That was the model throwing the error, because I don't have access to it. If that's what you mean?
I'm not sure about the underlying code; I haven't checked the source myself. But this repo in combination with the stack trace I provided should be enough to figure out what exactly was throwing the error...I think? Let me know if there's something more you need.
from gpt-migrate.
Sounds good @BeMain - will follow up
from gpt-migrate.
Same issue here. I think you need to upgrade your API plan to use GPT-4 model. Readme file says GPT-4-32K is preferred
from gpt-migrate.
Getting the same problem, I only have access to gpt-4 and not gpt-4-32k and no way to upgrade from what I can see.
from gpt-migrate.
I was able to at least get this running now. but there is a rate limit error and the execution stops abruptly 😞
from gpt-migrate.
Oh, so that's what the error message means; that the model isn't available? That might be it, as using gpt-3.5-turbo or gpt-4 doesn't cause the error (but instead complains about token length). Strange, I tried this a few weeks back and think I got the error that the model isn't available (when trying to use gpt-4-32k) back then too, only I remember the error message as being much clearer. Maybe things have changed recently.
I too only have access to gpt-4, so perhaps the only option is waiting for #2 then?
from gpt-migrate.
I was able to at least get this running now. but there is a rate limit error and the execution stops abruptly 😞
What did you do different to get it to work?
from gpt-migrate.
I was able to at least get this running now. but there is a rate limit error and the execution stops abruptly 😞
What did you do different to get it to work?
used gpt-4 model. Since there is a rate limit for gpt-4 unlike gpt-4-32K, I used try and catch for all API calls in utils.py file to retry the request.
def llm_run(prompt,waiting_message,success_message,globals):
output = ""
with yaspin(text=waiting_message, spinner="dots") as spinner:
try:
output = globals.ai.run(prompt)
except openai.error.RateLimitError as e:
retry_time = e.retry_after if hasattr(e, 'retry_after') else 30
time.sleep(retry_time)
output = globals.ai.run(prompt) # Retry the API request
spinner.ok("✅ ")
if success_message:
success_text = typer.style(success_message, fg=typer.colors.GREEN)
typer.echo(success_text)
return output
from gpt-migrate.
Hey @BeMain @mpmnath @christian-ek
I'm the co-maintainer of litellm. We should have returned a better error message here. That error means that no valid model was found.
Can you show me the code to repro this problem? What was the model name that got passed in
from gpt-migrate.
Was there any update to this? It doesn't like gpt-4 at all, but gpt-4-turbo-preview seems to work but the tokens are incorrect.
gpt-4
openai.error.InvalidRequestError: The model `gpt-4` does not exist or you do not have access to it. Learn more: https://help.openai.com/en/articles/7102672-how-can-i-access-gpt-4.
gpt-4-turbo-preview
ValueError: No valid completion model args passed in - {'model': 'gpt-4-turbo-preview', 'messages': [{'role': 'user', 'content': '...'}], 'functions': [], 'function_call': '', 'temperature': 0.0, 'top_p': 1, 'n': 1, 'stream': False, 'stop': None, 'max_tokens': 10000, 'presence_penalty': 0, 'frequency_penalty': 0, 'logit_bias': {}, 'user': '', 'force_timeout': 60, 'azure': False, 'logger_fn': None, 'verbose': False, 'optional_params': {'temperature': 0.0, 'max_tokens': 10000}}
I'll start updating args based on the openai API.
from gpt-migrate.
I just downloaded this last week and starting setting up. I'm using gpt-4-32k and currently stuck on this error. How do I fix it?
from gpt-migrate.
Hey - we should have this fixed on our end (i.e. raising the right error here). I also don't see this line ValueError: No valid completion model args passed in
in litellm's codebase anymore - https://github.com/search?q=repo%3ABerriAI%2Flitellm+%22ValueError%3A+No+valid+completion+model+args+passed+in%22&type=code
from gpt-migrate.
Interesting. I wonder if this is coming from one of the python libraries. Since I just downloaded the gpt migrate last week, I should have the latest code.
│ 270 │ │ raise InvalidRequestError(f"CohereException - {error_str}", f"{model}") │
│ 271 │ │ elif "CohereConnectionError" in exception_type: # cohere seems to fire these err │
│ 272 │ │ raise RateLimitError(f"CohereException - {original_exception.message}") │
│ > 273 │ raise original_exception # base case - return the original exception │
│ 274 │ else: │
│ 275 │ raise original_exception │
│ 276 │
│ │
│ ┌─────────────────────────────────────────── locals ───────────────────────────────────────────┐ │
│ │ error_str = "No valid completion model args passed in - {'model': │ │
│ │ 'openrouter/openai/gpt-4-32"+3349 │ │
│ │ exception_type = 'ValueError' │ │
│ │ model = 'openrouter/openai/gpt-4-32k' │ │
│ │ original_exception = ValueError('No valid completion model args passed in - {'model':
from gpt-migrate.
that looks like litellm but i don't think it's our latest version
Based on the requirements.txt i can see it's on a pretty old version -
Can you try this snippet on our latest version (v1.34.35
) and let me know if you see it
#pip install litellm==1.34.35
from litellm import completion
import os
## set ENV variables
os.environ["OPENAI_API_KEY"] = "your-openai-key"
messages = [{ "content": "Hello, how are you?","role": "user"}]
# openai call
response = completion(model="gpt-4-32k", messages=messages)
print(response)
from gpt-migrate.
Related Issues (20)
- Target repository doesn't make use of internal functions HOT 3
- Demanding 320k tokens [ISSUE STILL NOT RESOLVED BUT CLOSED] HOT 6
- Add azure support <Feature request> HOT 5
- Issue with GPT-3.5 for Python to PHP
- [FEATURE REQUEST]: Using this for more than migration.
- Sweep Github App
- Use poetry for installation HOT 4
- Support for libraries HOT 1
- Support for OpenRouter HOT 4
- LiteLLM Improvements HOT 3
- Use my own OpenAI keys instead of Open Router
- Invalid response object from API: 'Request Entity Too Large' (HTTP response code was 413) with openrouter gpt-4-32k HOT 1
- Confused about dependency handling
- Openai error "ModuleNotFoundError: No module named 'openai.error'" HOT 4
- Vector embeddings for different languages?
- TypeError: can only concatenate str (not "NoneType") to str HOT 4
- Langchain / OpenAI Dependencies
- `json.decoder.JSONDecodeError`
- Add distributed inference
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.
from gpt-migrate.