v0.1.34

🎉 Introduction to new functions of GPTCache

1. Add support for Qdrant Vector Store
2. Add support for Mongodb Cache Store
3. Fix bug about the redis vector and onnx similarity evaluation

What's Changed

• Correct the wrong search return value in the Redis vector store. by @SimFG in #452
• [Feature] Cache consistency check for Chroma & Milvus by @wybryan in #448
• Fix the pylint error and add the chromadb test by @SimFG in #457
• [add] support for mongodb storage by @a9raag in #454
• Fix the wrong return value of onnx similarity evaluation by @SimFG in #460

New Contributors

• @a9raag made their first contribution in #454

Full Changelog: 0.1.33...0.1.34

v0.1.33

🎉 Introduction to new functions of GPTCache

1. Make some improvements to the code by fixing a few bugs. For further information, please refer to the new pull request list.
2. Add How to better configure your cache document

What's Changed

• Updated link to langchain instructions by @technicalpickles in #434
• Fix the eviction error by @SimFG in #440
• [Feature] search only operation support by @wybryan in #445
• T10H-85 - VectorBase change for namespace allocation by @jacktempo7 in #449
• Add How to better configure your cache document by @SimFG in #450

New Contributors

• @technicalpickles made their first contribution in #434
• @wybryan made their first contribution in #445
• @jacktempo7 made their first contribution in #449

Full Changelog: 0.1.32...0.1.33

v0.1.32

🎉 Introduction to new functions of GPTCache

1. Support the redis as vector store

from gptcache.manager import VectorBase

vector_base = VectorBase("redis", dimension=10)

2. Fix the context len config bug

What's Changed

• Fix context_len in config by @zc277584121 in #430
• Fix sequence match example by @zc277584121 in #431
• Add the Redis vector store by @SimFG in #432

New Contributors

• @zc277584121 made their first contribution in #430

Full Changelog: 0.1.31...0.1.32

v0.1.31

🎉 Introduction to new functions of GPTCache

1. To improve the precision of cache hits, four similarity evaluation methods were added

• SBERT CrossEncoder Evaluation
• Cohere rerank api (Free accounts can make up to 100 calls per minute.)
• Multi-round dialog similarity weight matching
• Time Evaluation. For the cached answer, first check the time dimension, such as only using the generated cache for the past day

2. Fix some bugs

• OpenAI exceptions type #416
• LangChainChat does work with _agenerate function #400

more details: https://github.com/zilliztech/GPTCache/blob/main/docs/release_note.md

What's Changed

• Raise the same type's error for the openai by @SimFG in #421
• Add sequence match evaluation. by @wxywb in #420
• Add the Time Evaluation by @SimFG in #423
• Improve SequenceMatchEvaluation for several cases. by @wxywb in #424
• Change the evaluation score of sequence evaluation to be larger as th… by @wxywb in #425
• LangchainChat support _agenerate function by @SimFG in #426
• Add SBERT CrossEncoder evaluation. by @wxywb in #428
• Update the version to 0.1.31 by @SimFG in #429

Full Changelog: 0.1.30...0.1.31

v0.1.30

🎉 Introduction to new functions of GPTCache

1. Support to use the cohere rerank api to evaluate the similarity

from gptcache.similarity_evaluation import CohereRerankEvaluation

evaluation = CohereRerankEvaluation()
score = evaluation.evaluation(
    {
        'question': 'What is the color of sky?'
    },
    {
        'answer': 'the color of sky is blue'
    }
)

2. Improve the gptcache server api, refer to the "/docs" path after starting the server
3. Fix the bug about the langchain track token usage

What's Changed

• Add input summarization. by @wxywb in #404
• Langchain track token usage by @SimFG in #409
• Support to download the cache files by @SimFG in #410
• Support to use the cohere rerank api to evaluate the similarity by @SimFG in #412

Full Changelog: 0.1.29...0.1.30

v0.1.29

🎉 Introduction to new functions of GPTCache

1. Improve the GPTCache server by using FASTAPI

NOTE: The api struct has been optimized, details: Use GPTCache server

2. Add the usearch vector store

from gptcache.manager import manager_factory

data_manager = manager_factory("sqlite,usearch", vector_params={"dimension": 10})

What's Changed

• Improve the unit test flow by @SimFG in #397
• Add: USearch vector search engine by @VoVoR in #399
• Add the saved token report, auto flush data by @SimFG in #401
• Use the fastapi to improve the GPTCache server by @SimFG in #405
• Update the version to 0.1.29 by @SimFG in #406

New Contributors

• @VoVoR made their first contribution in #399

Full Changelog: 0.1.28...0.1.29

v0.1.28

🎉 Introduction to new functions of GPTCache

To handle a large prompt, there are currently two options available:

1. Increase the column size of CacheStorage.

from gptcache.manager import manager_factory

data_manager = manager_factory(
    "sqlite,faiss", scalar_params={"table_len_config": {"question_question": 5000}}
)

More Details:

• 'question_question': the question column size in the question table, default to 3000.
• 'answer_answer': the answer column size in the answer table, default to 3000.
• 'session_id': the session id column size in the session table, default to 1000.
• 'dep_name': the name column size in the dep table, default to 1000.
• 'dep_data': the data column size in the dep table, default to 3000.

2. When using a template, use the dynamic value in the template as the cache key instead of using the entire template as the key.

• str template

from gptcache import Config
from gptcache.processor.pre import last_content_without_template

template_obj = "tell me a joke about {subject}"
prompt = template_obj.format(subject="animal")
value = last_content_without_template(
    data={"messages": [{"content": prompt}]}, cache_config=Config(template=template_obj)
)
print(value)
# ['animal']

• langchain prompt template

from langchain import PromptTemplate

from gptcache import Config
from gptcache.processor.pre import last_content_without_template

template_obj = PromptTemplate.from_template("tell me a joke about {subject}")
prompt = template_obj.format(subject="animal")

value = last_content_without_template(
    data={"messages": [{"content": prompt}]},
    cache_config=Config(template=template_obj.template),
)
print(value)
# ['animal']

3. Wrap the openai object, reference: BaseCacheLLM

import random

from gptcache import Cache
from gptcache.adapter import openai
from gptcache.adapter.api import init_similar_cache
from gptcache.processor.pre import last_content

cache_obj = Cache()
init_similar_cache(
    data_dir=str(random.random()), pre_func=last_content, cache_obj=cache_obj
)


def proxy_openai_chat_complete(*args, **kwargs):
    nonlocal is_proxy
    is_proxy = True
    import openai as real_openai

    return real_openai.ChatCompletion.create(*args, **kwargs)


openai.ChatCompletion.llm = proxy_openai_chat_complete
openai.ChatCompletion.cache_args = {"cache_obj": cache_obj}

openai.ChatCompletion.create(
    model="gpt-3.5-turbo",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "What's GitHub"},
    ],
)

What's Changed

• Add the BaseCacheLLM abstract class to wrap the llm by @SimFG in #394
• Add the pre-function of handling long prompt and Update context doc by @SimFG in #395
• Support to config the context pre-process by the yaml file by @SimFG in #396

Full Changelog: 0.1.27...0.1.28

v0.1.27

🎉 Introduction to new functions of GPTCache

1. Support the uform embedding, which can be used the bilingual (english + chinese) language

thanks @ashvardanian 's contribution

from gptcache.embedding import UForm

test_sentence = 'Hello, world.'
encoder = UForm(model='unum-cloud/uform-vl-english')
embed = encoder.to_embeddings(test_sentence)

test_sentence = '什么是Github'
encoder = UForm(model='unum-cloud/uform-vl-multilingual')
embed = encoder.to_embeddings(test_sentence)

What's Changed

• Fix the wrong LangChainChat comment by @SimFG in #381
• Add UForm multi-modal embedding by @SimFG in #382
• Support to config the cache storage data size by @SimFG in #383
• Update the protobuf version in the doc by @SimFG in #387
• Update the version to 0.1.27 by @SimFG in #389

Full Changelog: 0.1.26...0.1.27

v0.1.26

🎉 Introduction to new functions of GPTCache

1. Support the paddlenlp embedding @vax521

from gptcache.embedding import PaddleNLP

test_sentence = 'Hello, world.'
encoder = PaddleNLP(model='ernie-3.0-medium-zh')
embed = encoder.to_embeddings(test_sentence)

2. Support the openai Moderation api

from gptcache.adapter import openai
from gptcache.adapter.api import init_similar_cache
from gptcache.processor.pre import get_openai_moderation_input

init_similar_cache(pre_func=get_openai_moderation_input)
openai.Moderation.create(
    input="hello, world",
)

3. Add the llama_index bootcamp, through which you can learn how GPTCache works with llama index

details: WebPage QA

What's Changed

• Replace summarization test model. by @wxywb in #368
• Add the llama index bootcamp by @SimFG in #371
• Update the llama index example url by @SimFG in #372
• Support the openai moderation adapter by @SimFG in #376
• Paddlenlp embedding support by @SimFG in #377
• Update the cache config template file and example directory by @SimFG in #380

Full Changelog: 0.1.25...0.1.26

v0.1.25

🎉 Introduction to new functions of GPTCache

1. Support the DocArray vector database

from gptcache.manager import manager_factory

data_manager = manager_factory("sqlite,docarray")

2. Add rwkv model for embedding

from gptcache.embedding import Rwkv

test_sentence = 'Hello, world.'
encoder = Rwkv(model='sgugger/rwkv-430M-pile')
embed = encoder.to_embeddings(test_sentence)

What's Changed

• [skip ci]Add workflow to publish release image by @Bennu-Li in #345
• Update the doc directory by @SimFG in #348
• Add the docker image doc by @SimFG in #349
• DocArray as a vectorstore by @jupyterjazz in #351
• Fix the doc generation failure by @SimFG in #352
• Replace base image and simplify dockerfile by @Chiiizzzy in #353
• Example with DocArray by @jupyterjazz in #354
• Add comments for session by @shiyu22 in #355
• DocArray example adjustment by @jupyterjazz in #356
• Improve the generation doc script by @SimFG in #358
• Change the model test cases as L2 cases by @SimFG in #362
• Add concat_context. by @wxywb in #365
• Add the pre/report/session docs by @SimFG in #364
• Add rwkv model for embedding. by @wxywb in #363
• Update the version to 0.1.25 by @SimFG in #367

New Contributors

• @jupyterjazz made their first contribution in #351

Full Changelog: 0.1.24...0.1.25