DSPy: The framework for programming—not prompting—language models
Python
Clean files for PR
Isaac Millercc1f7023d11b
diff --git a/dsp/__init__.py b/dsp/__init__.pyindex f71e1bc2c..86b103f10 100644--- a/dsp/__init__.py+++ b/dsp/__init__.py@@ -2,7 +2,6 @@ from .modules import * # noqafrom .primitives import * # noqafrom .adapters import * # noqafrom .utils import settings # noqa-# from .deploy_dspy import *"""TODO:diff --git a/dsp/modules/__init__.py b/dsp/modules/__init__.pyindex 564f44475..10695435f 100644--- a/dsp/modules/__init__.py+++ b/dsp/modules/__init__.py@@ -29,8 +29,5 @@ from .sentence_vectorizer import *from .snowflake import *from .tensorrt_llm import TensorRTModelfrom .trainable_anyscale import TrainableAnyscale-# from .vllm_engine import VLLMOfflineEngine-# from .vllm_engine_2 import VLLMOfflineEngine2-# from .vllm_engine_3 import VLLMOfflineEngine3from .watsonx import *from .you import Youdiff --git a/dsp/modules/anthropic.py b/dsp/modules/anthropic.pyindex 57a7fccfd..cd3f2d4b7 100644--- a/dsp/modules/anthropic.py+++ b/dsp/modules/anthropic.py@@ -8,11 +8,9 @@ from dsp.utils.settings import settingsfrom dsp.modules.lm import LMtry:- # import anthropic- pass- # anthropic_rate_limit = anthropic.RateLimitError- anthropic_rate_limit = Exception+ import anthropic+ anthropic_rate_limit = anthropic.RateLimitErrorexcept ImportError:anthropic_rate_limit = Exceptiondiff --git a/dsp/modules/cache_utils.py b/dsp/modules/cache_utils.pyindex bb4b9fb16..a7e2bde2f 100644--- a/dsp/modules/cache_utils.py+++ b/dsp/modules/cache_utils.py@@ -6,7 +6,7 @@ from joblib import Memoryfrom dsp.utils import dotdict-cache_turn_on = os.environ.get('DSP_CACHEBOOL', 'true').lower() != 'false'+cache_turn_on = os.environ.get('DSP_CACHEBOOL', 'True').lower() != 'false'def noop_decorator(arg=None, *noop_args, **noop_kwargs):diff --git a/dsp/modules/gpt3.py b/dsp/modules/gpt3.pyindex 62d495692..b1696e645 100644--- a/dsp/modules/gpt3.py+++ b/dsp/modules/gpt3.py@@ -127,13 +127,13 @@ class GPT3(LM):max_tokens = kwargs.pop("max_tokens")kwargs["max_completion_tokens"] = max_tokenskwargs.pop("temperature")+if self.model_type == "chat":# caching mechanism requires hashable kwargsmessages = [{"role": "user", "content": prompt}]if self.system_prompt:messages.insert(0, {"role": "system", "content": self.system_prompt})kwargs["messages"] = messages- print("in basic_request", kwargs, "\n\n")kwargs = {"stringify_request": json.dumps(kwargs)}response = chat_request(**kwargs)@@ -250,9 +250,7 @@ def cached_gpt3_request_v2_wrapped(**kwargs):def _cached_gpt3_turbo_request_v2(**kwargs) -> OpenAIObject:if "stringify_request" in kwargs:kwargs = json.loads(kwargs["stringify_request"])- response = cast(OpenAIObject, openai.ChatCompletion.create(**kwargs))- print(response)- return response+ return cast(OpenAIObject, openai.ChatCompletion.create(**kwargs))@functools.lru_cache(maxsize=None if cache_turn_on else 0)@@ -276,9 +274,7 @@ def v1_cached_gpt3_request_v2_wrapped(**kwargs):def v1_cached_gpt3_turbo_request_v2(**kwargs):if "stringify_request" in kwargs:kwargs = json.loads(kwargs["stringify_request"])- response = openai.chat.completions.create(**kwargs)- print(response)- return response+ return openai.chat.completions.create(**kwargs)@functools.lru_cache(maxsize=None if cache_turn_on else 0)