DSPy: The framework for programming—not prompting—language models
Python
Cleanup for PR
Isaac Miller9b7d2a3d9f7d
diff --git a/dsp/modules/lm.py b/dsp/modules/lm.pyindex 88fbe284b..e06962a63 100644--- a/dsp/modules/lm.py+++ b/dsp/modules/lm.py@@ -1,11 +1,4 @@from abc import ABC, abstractmethod-from concurrent.futures import ThreadPoolExecutor, Future-from copy import deepcopy-from enum import Enum-from typing import Optional, Union, List-import ujson--class LM(ABC):@@ -55,7 +48,7 @@ class LM(ABC):printed = []n = n + skip- for x in reversed(self.history):+ for x in reversed(self.history[-100:]):prompt = x["prompt"]if prompt != last_prompt:@@ -85,10 +78,6 @@ class LM(ABC):printed.append((prompt, x))elif provider == "you.com":printed.append((prompt, x["response"]["answer"]))- elif provider == "vllm_offline":- printed.append((prompt, x["response"].outputs))- elif provider == "vllm":- printed.append((prompt, x["response"]["choices"]))else:printed.append((prompt, x["response"]["choices"]))@@ -128,10 +117,6 @@ class LM(ABC):text = choices[0].message.contentelif provider == "cloudflare":text = choices[0]- elif provider == "vllm":- text = choices[0]["message"]["content"]- elif provider == "vllm_offline":- text = choices[0].textelse:text = choices[0]["text"]printing_value += self.print_green(text, end="") if color_format else text@@ -169,107 +154,5 @@ class LM(ABC):"""Returns a copy of the language model with the same parameters."""kwargs = {**self.kwargs, **kwargs}model = kwargs.pop("model")- return self.__class__(model, **kwargs)--#--------------------------------------------------------------------------------# Classes for finetuning LMs-#---------------------------------------------------------------------------------class TrainingMethod(Enum):- # TODO: add docstring- SFT = "SFT"- Preference = "Preference"---"""Dictionary mapping training methods to the data keys they require."""-TRAINING_METHOD_TO_DATA_KEYS = {- TrainingMethod.SFT: ["prompt", "completion"],- TrainingMethod.Preference: ["prompt", "chosen", "rejected"],-}---def verify_training_method_data_format(- method: TrainingMethod,- data_path: str-) -> Optional[AssertionError]:- # TODO: add docstring- expected_keys = TRAINING_METHOD_TO_DATA_KEYS[method]- data = ujson.load(open(data_path))- for ind, data_dict in enumerate(data):- err_msg = f"The datapoint at index {ind} is missing the keys required for {method} training."- err_msg = f"\n Expected: {expected_keys}"- err_msg = f"\n Found: {data_dict.keys()}"- assert all([key in data_dict for key in expected_keys]), err_msg---class TrainableLM(LM, ABC):- # TODO: add docstring-- """The training methods supported by this class, should be overwritten."""- SUPPORTED_TRAINING_METHODS: List[TrainingMethod] = []-- def get_finetune(- self,- method: TrainingMethod,- train_path: str,- eval_path: Optional[str],- **kwargs- ) -> Union[Future['TrainableLM'], Exception]:- # TODO: add docstring-- # Input verification- assert method in self.SUPPORTED_TRAINING_METHODS- # Note: Commenting this out because we do not get our data in prompt completion format- # verify_training_method_data_format(method, data_path=train_path)- # if eval_path:- # verify_training_method_data_format(method, data_path=eval_path)-- # Create new model that will eventually be obtained through- # future.result() when its "start_training" method completes its- # execution- future: Future['TrainableLM'] = Future()- new_lm = deepcopy(self)-- # Capture the current instance in the closure and start the training- # process asynchronously- executor = ThreadPoolExecutor(max_workers=1)- executor.submit(- new_lm.start_training,- future,- method=method,- train_path=train_path,- eval_path=eval_path,- **kwargs- )- executor.shutdown(wait=False)-- return future-- @abstractmethod- def start_training(- self,- future: Future['TrainableLM'],- train_path: str,- eval_path: Optional[str],- method: TrainingMethod,- **kwargs):- # TODO: add docstring- lname = self.__class__.__name__- err_msg = f"{lname} does not implement the 'start_training' method."- raise NotImplementedError(err_msg)-- @abstractmethod- def stop_training(self):- # TODO: add docstring- lname = self.__class__.__name__- err_msg = f"{lname} does not implement the 'stop_training' method."- raise NotImplementedError(err_msg)--def read_jsonl(filename):- with open(filename, "r") as f:- return [ujson.loads(line) for line in f]-def write_jsonl(filename, data):- with open(filename, "w") as f:- for item in data:- f.write(ujson.dumps(item) + "\n")\ No newline at end of file+ return self.__class__(model=model, **kwargs)diff --git a/dspy/adapters/chat_adapter.py b/dspy/adapters/chat_adapter.pyindex 87d25b404..03f1236ca 100644--- a/dspy/adapters/chat_adapter.py+++ b/dspy/adapters/chat_adapter.py@@ -63,7 +63,6 @@ class ChatAdapter(Adapter):)if list(fields.keys()) != list(signature.output_fields.keys()):- # print("Expected", list(signature.output_fields.keys()), "but got", fields.keys(), "from", completion)raise ValueError(f"Expected {list(signature.output_fields.keys())} but got {fields.keys()}")return fieldsdiff --git a/dspy/datasets/gsm8k.py b/dspy/datasets/gsm8k.pyindex e1fc6b0d1..328d1aee8 100644--- a/dspy/datasets/gsm8k.py+++ b/dspy/datasets/gsm8k.py@@ -44,8 +44,8 @@ class GSM8K:rng = random.Random(0)rng.shuffle(official_test)- trainset = official_train- devset = official_train[-200:]+ trainset = official_train[:200]+ devset = official_train[200:500]testset = official_test[:]import dspydiff --git a/dspy/predict/predict.py b/dspy/predict/predict.pyindex dc8ae3437..0e3610b86 100644--- a/dspy/predict/predict.py+++ b/dspy/predict/predict.py@@ -288,4 +288,3 @@ def v2_5_generate(lm, lm_kwargs, signature, demos, inputs, _parse_values=True):# # TODO: FIXME: Hmm, I guess expected behavior is that contexts can# affect execution. Well, we need to determine whether context dominates, __init__ demoninates, or forward dominates.# Generally, unless overwritten, we'd see n=None, temperature=None.-# That will eventually mean we have to learn them.diff --git a/requirements.txt b/requirements.txtindex 29f419764..f4b28d788 100644--- a/requirements.txt+++ b/requirements.txt@@ -1,4 +1,3 @@-# DSPy requirementsbackoffdatasetsjoblib~=1.3