From 864716d6666efd19543040c12cc8748827d15770 Mon Sep 17 00:00:00 2001
From: Logan <logan.markewich@live.com>
Date: Mon, 8 Apr 2024 13:30:41 -0600
Subject: [PATCH] Fix many GitHub issues (#12655)
fixes
---
.../core/node_parser/text/sentence.py | 2 +-
.../core/response_synthesizers/__init__.py | 6 +++++-
.../response_synthesizers/simple_summarize.py | 16 ++++++++--------
.../llama_index/llms/openllm/base.py | 2 +-
.../llms/llama-index-llms-openllm/pyproject.toml | 2 +-
.../pyproject.toml | 4 ++--
.../llama_index/readers/file/docs/base.py | 3 +++
.../llama-index-readers-file/pyproject.toml | 2 +-
.../packs/llama_dataset_metadata/base.py | 9 ++++-----
.../pyproject.toml | 2 +-
10 files changed, 27 insertions(+), 21 deletions(-)
diff --git a/llama-index-core/llama_index/core/node_parser/text/sentence.py b/llama-index-core/llama_index/core/node_parser/text/sentence.py
index 550e0d146..79c5dd5e0 100644
--- a/llama-index-core/llama_index/core/node_parser/text/sentence.py
+++ b/llama-index-core/llama_index/core/node_parser/text/sentence.py
@@ -199,7 +199,7 @@ class SentenceSplitter(MetadataAwareTextSplitter):
"""
token_size = self._token_size(text)
- if self._token_size(text) <= chunk_size:
+ if token_size <= chunk_size:
return [_Split(text, is_sentence=True, token_size=token_size)]
text_splits_by_fns, is_sentence = self._get_splits_by_fns(text)
diff --git a/llama-index-core/llama_index/core/response_synthesizers/__init__.py b/llama-index-core/llama_index/core/response_synthesizers/__init__.py
index 35d10e1df..4c488b70d 100644
--- a/llama-index-core/llama_index/core/response_synthesizers/__init__.py
+++ b/llama-index-core/llama_index/core/response_synthesizers/__init__.py
@@ -1,7 +1,10 @@
"""Init file."""
from llama_index.core.response_synthesizers.accumulate import Accumulate
-from llama_index.core.response_synthesizers.base import BaseSynthesizer
+from llama_index.core.response_synthesizers.base import (
+ BaseSynthesizer,
+ SynthesizerComponent,
+)
from llama_index.core.response_synthesizers.compact_and_refine import (
CompactAndRefine,
)
@@ -15,6 +18,7 @@ from llama_index.core.response_synthesizers.type import ResponseMode
__all__ = [
"ResponseMode",
"BaseSynthesizer",
+ "SynthesizerComponent",
"Refine",
"SimpleSummarize",
"TreeSummarize",
diff --git a/llama-index-core/llama_index/core/response_synthesizers/simple_summarize.py b/llama-index-core/llama_index/core/response_synthesizers/simple_summarize.py
index 152459df1..27d92e151 100644
--- a/llama-index-core/llama_index/core/response_synthesizers/simple_summarize.py
+++ b/llama-index-core/llama_index/core/response_synthesizers/simple_summarize.py
@@ -52,23 +52,23 @@ class SimpleSummarize(BaseSynthesizer):
**response_kwargs: Any,
) -> RESPONSE_TEXT_TYPE:
text_qa_template = self._text_qa_template.partial_format(query_str=query_str)
+ single_text_chunk = "\n".join(text_chunks)
truncated_chunks = self._prompt_helper.truncate(
prompt=text_qa_template,
- text_chunks=text_chunks,
+ text_chunks=[single_text_chunk],
)
- node_text = "\n".join(truncated_chunks)
response: RESPONSE_TEXT_TYPE
if not self._streaming:
response = await self._llm.apredict(
text_qa_template,
- context_str=node_text,
+ context_str=truncated_chunks,
**response_kwargs,
)
else:
response = self._llm.stream(
text_qa_template,
- context_str=node_text,
+ context_str=truncated_chunks,
**response_kwargs,
)
@@ -86,23 +86,23 @@ class SimpleSummarize(BaseSynthesizer):
**kwargs: Any,
) -> RESPONSE_TEXT_TYPE:
text_qa_template = self._text_qa_template.partial_format(query_str=query_str)
+ single_text_chunk = "\n".join(text_chunks)
truncated_chunks = self._prompt_helper.truncate(
prompt=text_qa_template,
- text_chunks=text_chunks,
+ text_chunks=[single_text_chunk],
)
- node_text = "\n".join(truncated_chunks)
response: RESPONSE_TEXT_TYPE
if not self._streaming:
response = self._llm.predict(
text_qa_template,
- context_str=node_text,
+ context_str=truncated_chunks,
**kwargs,
)
else:
response = self._llm.stream(
text_qa_template,
- context_str=node_text,
+ context_str=truncated_chunks,
**kwargs,
)
diff --git a/llama-index-integrations/llms/llama-index-llms-openllm/llama_index/llms/openllm/base.py b/llama-index-integrations/llms/llama-index-llms-openllm/llama_index/llms/openllm/base.py
index 1781326a6..e0923d268 100644
--- a/llama-index-integrations/llms/llama-index-llms-openllm/llama_index/llms/openllm/base.py
+++ b/llama-index-integrations/llms/llama-index-llms-openllm/llama_index/llms/openllm/base.py
@@ -94,7 +94,7 @@ class OpenLLM(LLM):
trust_remote_code: bool = Field(
description="Optional flag to trust remote code. This is synonymous to Transformers' `trust_remote_code`. Default to False."
)
- _llm: openllm.LLM[Any, Any]
+ _llm: openllm.LLM[Any, Any] = PrivateAttr()
def __init__(
self,
diff --git a/llama-index-integrations/llms/llama-index-llms-openllm/pyproject.toml b/llama-index-integrations/llms/llama-index-llms-openllm/pyproject.toml
index 3d3bb0c08..518e778e0 100644
--- a/llama-index-integrations/llms/llama-index-llms-openllm/pyproject.toml
+++ b/llama-index-integrations/llms/llama-index-llms-openllm/pyproject.toml
@@ -28,7 +28,7 @@ exclude = ["**/BUILD"]
license = "MIT"
name = "llama-index-llms-openllm"
readme = "README.md"
-version = "0.1.3"
+version = "0.1.4"
[tool.poetry.dependencies]
python = ">=3.8.1,<4.0"
diff --git a/llama-index-integrations/multi_modal_llms/llama-index-multi-modal-llms-anthropic/pyproject.toml b/llama-index-integrations/multi_modal_llms/llama-index-multi-modal-llms-anthropic/pyproject.toml
index d367dada5..53742840a 100644
--- a/llama-index-integrations/multi_modal_llms/llama-index-multi-modal-llms-anthropic/pyproject.toml
+++ b/llama-index-integrations/multi_modal_llms/llama-index-multi-modal-llms-anthropic/pyproject.toml
@@ -27,12 +27,12 @@ exclude = ["**/BUILD"]
license = "MIT"
name = "llama-index-multi-modal-llms-anthropic"
readme = "README.md"
-version = "0.1.3"
+version = "0.1.4"
[tool.poetry.dependencies]
python = ">=3.8.1,<4.0"
llama-index-core = "^0.10.1"
-anthropic = "0.20.0"
+anthropic = "^0.23.1"
[tool.poetry.group.dev.dependencies]
ipython = "8.10.0"
diff --git a/llama-index-integrations/readers/llama-index-readers-file/llama_index/readers/file/docs/base.py b/llama-index-integrations/readers/llama-index-readers-file/llama_index/readers/file/docs/base.py
index a3c168d61..441ee5e05 100644
--- a/llama-index-integrations/readers/llama-index-readers-file/llama_index/readers/file/docs/base.py
+++ b/llama-index-integrations/readers/llama-index-readers-file/llama_index/readers/file/docs/base.py
@@ -34,6 +34,9 @@ class PDFReader(BaseReader):
fs: Optional[AbstractFileSystem] = None,
) -> List[Document]:
"""Parse file."""
+ if not isinstance(file, Path):
+ file = Path(file)
+
try:
import pypdf
except ImportError:
diff --git a/llama-index-integrations/readers/llama-index-readers-file/pyproject.toml b/llama-index-integrations/readers/llama-index-readers-file/pyproject.toml
index b00813109..354f59e22 100644
--- a/llama-index-integrations/readers/llama-index-readers-file/pyproject.toml
+++ b/llama-index-integrations/readers/llama-index-readers-file/pyproject.toml
@@ -50,7 +50,7 @@ license = "MIT"
maintainers = ["FarisHijazi", "Haowjy", "ephe-meral", "hursh-desai", "iamarunbrahma", "jon-chuang", "mmaatouk", "ravi03071991", "sangwongenip", "thejessezhang"]
name = "llama-index-readers-file"
readme = "README.md"
-version = "0.1.13"
+version = "0.1.14"
[tool.poetry.dependencies]
python = ">=3.8.1,<4.0"
diff --git a/llama-index-packs/llama-index-packs-llama-dataset-metadata/llama_index/packs/llama_dataset_metadata/base.py b/llama-index-packs/llama-index-packs-llama-dataset-metadata/llama_index/packs/llama_dataset_metadata/base.py
index 66cd6e4cb..09ade2600 100644
--- a/llama-index-packs/llama-index-packs-llama-dataset-metadata/llama_index/packs/llama_dataset_metadata/base.py
+++ b/llama-index-packs/llama-index-packs-llama-dataset-metadata/llama_index/packs/llama_dataset_metadata/base.py
@@ -7,6 +7,7 @@ from llama_index.core.download.module import LLAMA_HUB_URL
from llama_index.core.download.utils import get_file_content
from llama_index.core.indices.base import BaseIndex
from llama_index.core.llama_pack.base import BaseLlamaPack
+from llama_index.core.settings import Settings
if TYPE_CHECKING:
from llama_index.core.llama_dataset import LabelledRagDataset
@@ -143,11 +144,9 @@ class DatasetCard(BaseMetadata):
)
# extract baseline config info from index
- llm = index.service_context.llm.model
- embed_model = index.as_retriever().get_service_context().embed_model.model_name
- chunk_size = (
- index.as_retriever().get_service_context().transformations[0].chunk_size
- )
+ llm = Settings.llm.metadata.model_name
+ embed_model = Settings.embed_model.model_name
+ chunk_size = index._transformations[0].chunk_size
similarity_top_k = index.as_retriever()._similarity_top_k
baseline_config = BaselineConfig(
llm=llm,
diff --git a/llama-index-packs/llama-index-packs-llama-dataset-metadata/pyproject.toml b/llama-index-packs/llama-index-packs-llama-dataset-metadata/pyproject.toml
index fcbc6ff08..9d3d880b3 100644
--- a/llama-index-packs/llama-index-packs-llama-dataset-metadata/pyproject.toml
+++ b/llama-index-packs/llama-index-packs-llama-dataset-metadata/pyproject.toml
@@ -29,7 +29,7 @@ license = "MIT"
maintainers = ["nerdai"]
name = "llama-index-packs-llama-dataset-metadata"
readme = "README.md"
-version = "0.1.3"
+version = "0.1.4"
[tool.poetry.dependencies]
python = ">=3.8.1,<4.0"
--
GitLab