dspy

dspy copied to clipboard

Error loading program compiled using the Ensemble optimizer

4

The app crashes in below code with a keyerror: File "/home/drawal/miniconda3/envs/xray/lib/python3.11/site-packages/dspy/primitives/module.py", line 75, in load_state param.load_state(state[name]) This is because the load function expects the top level keys in the saved...

drawal1

fix(dsp): prevent cache instantiation when cache is disabled

1

The environment variable `DSP_CACHEBOOL` does not completely disable the filesystem cache. As a result, `joblib.Memory` is still instantiated, which will try to create a directory on the filesystem. In an...

stevenhoelscher

[Request] Support for Python3.8

2

Hi, Can DSPy support Python3.8? It currently supports >=3.9 according to the documentation.

kalpit-konverge

Remove top_p default value

3

Pratically, we should not use top_p and temperture at the same time, however, we may override temperture in some cases, but since DSPy predefine `top_p = 1`, it's impossible to...

xdite

Refactor Bedrock class to support Claude-3 model and improve error handling

10

This PR refactors the Bedrock class to support the Claude-3 model. The main changes include: 1. Added support for the Claude-3 model which uses a different API format with messages...

cta2106

Pydantic vision module to build off of and enable multimodal capabilities for all providers.

7

Hi all, This library is amazing! I added something similar to the pydantic Image class we use to abstract away all those messy conversions you need to do. Question, why...

sebbyjp

fix(dspy): together client response parsing bug + endpoint

10

I opened up a related issue just now https://github.com/stanfordnlp/dspy/issues/626 Basically I was getting errors when using the Together client. If I set use_chat_api to true, the response did not have...

adam-simple

enh(dspy): Move max_tokens explicitally into `GPT3` constructor

4

Makes the reason for output truncation clearer as per #730 Also adds a comment to the predict file where the max tokens are modified before the call is made.

harrysalmon

style(dspy): Add type checking and clearer types to BaseSentenceVectorizer

2

There's likely a clearer refactor of this, but as a starter for 10. It avoids python treating strings as itterables of characters and clarifies the possible input types of some...

harrysalmon

Fix multiple openai clients

3

Since the OpenAI module is setting parameters on the base openai import, it's not possible to use multiple different models that use an openai compatible api (e.g. both a local...

rawsh