forked from phoenix-oss/llama-stack-mirror
Update the "InterleavedTextMedia" type (#635)
## What does this PR do? This is a long-pending change and particularly important to get done now. Specifically: - we cannot "localize" (aka download) any URLs from media attachments anywhere near our modeling code. it must be done within llama-stack. - `PIL.Image` is infesting all our APIs via `ImageMedia -> InterleavedTextMedia` and that cannot be right at all. Anything in the API surface must be "naturally serializable". We need a standard `{ type: "image", image_url: "<...>" }` which is more extensible - `UserMessage`, `SystemMessage`, etc. are moved completely to llama-stack from the llama-models repository. See https://github.com/meta-llama/llama-models/pull/244 for the corresponding PR in llama-models. ## Test Plan ```bash cd llama_stack/providers/tests pytest -s -v -k "fireworks or ollama or together" inference/test_vision_inference.py pytest -s -v -k "(fireworks or ollama or together) and llama_3b" inference/test_text_inference.py pytest -s -v -k chroma memory/test_memory.py \ --env EMBEDDING_DIMENSION=384 --env CHROMA_DB_PATH=/tmp/foobar pytest -s -v -k fireworks agents/test_agents.py \ --safety-shield=meta-llama/Llama-Guard-3-8B \ --inference-model=meta-llama/Llama-3.1-8B-Instruct ``` Updated the client sdk (see PR ...), installed the SDK in the same environment and then ran the SDK tests: ```bash cd tests/client-sdk LLAMA_STACK_CONFIG=together pytest -s -v agents/test_agents.py LLAMA_STACK_CONFIG=ollama pytest -s -v memory/test_memory.py # this one needed a bit of hacking in the run.yaml to ensure I could register the vision model correctly INFERENCE_MODEL=llama3.2-vision:latest LLAMA_STACK_CONFIG=ollama pytest -s -v inference/test_inference.py ```
This commit is contained in:
parent
10eb31badf
commit
8de8eb03c8
66 changed files with 1344 additions and 1801 deletions
|
@ -8,7 +8,7 @@ import base64
|
|||
import mimetypes
|
||||
import os
|
||||
|
||||
from llama_models.llama3.api.datatypes import URL
|
||||
from llama_stack.apis.common.content_types import URL
|
||||
|
||||
|
||||
def data_url_from_file(file_path: str) -> URL:
|
||||
|
|
|
@ -21,8 +21,13 @@ from pypdf import PdfReader
|
|||
from llama_models.llama3.api.datatypes import * # noqa: F403
|
||||
from llama_models.llama3.api.tokenizer import Tokenizer
|
||||
|
||||
from llama_stack.apis.common.content_types import InterleavedContent, TextContentItem
|
||||
from llama_stack.apis.memory import * # noqa: F403
|
||||
from llama_stack.apis.memory_banks import VectorMemoryBank
|
||||
from llama_stack.providers.datatypes import Api
|
||||
from llama_stack.providers.utils.inference.prompt_adapter import (
|
||||
interleaved_content_as_str,
|
||||
)
|
||||
|
||||
log = logging.getLogger(__name__)
|
||||
|
||||
|
@ -84,6 +89,26 @@ def content_from_data(data_url: str) -> str:
|
|||
return ""
|
||||
|
||||
|
||||
def concat_interleaved_content(content: List[InterleavedContent]) -> InterleavedContent:
|
||||
"""concatenate interleaved content into a single list. ensure that 'str's are converted to TextContentItem when in a list"""
|
||||
|
||||
ret = []
|
||||
|
||||
def _process(c):
|
||||
if isinstance(c, str):
|
||||
ret.append(TextContentItem(text=c))
|
||||
elif isinstance(c, list):
|
||||
for item in c:
|
||||
_process(item)
|
||||
else:
|
||||
ret.append(c)
|
||||
|
||||
for c in content:
|
||||
_process(c)
|
||||
|
||||
return ret
|
||||
|
||||
|
||||
async def content_from_doc(doc: MemoryBankDocument) -> str:
|
||||
if isinstance(doc.content, URL):
|
||||
if doc.content.uri.startswith("data:"):
|
||||
|
@ -108,7 +133,7 @@ async def content_from_doc(doc: MemoryBankDocument) -> str:
|
|||
else:
|
||||
return r.text
|
||||
|
||||
return interleaved_text_media_as_str(doc.content)
|
||||
return interleaved_content_as_str(doc.content)
|
||||
|
||||
|
||||
def make_overlapped_chunks(
|
||||
|
@ -121,6 +146,7 @@ def make_overlapped_chunks(
|
|||
for i in range(0, len(tokens), window_len - overlap_len):
|
||||
toks = tokens[i : i + window_len]
|
||||
chunk = tokenizer.decode(toks)
|
||||
# chunk is a string
|
||||
chunks.append(
|
||||
Chunk(content=chunk, token_count=len(toks), document_id=document_id)
|
||||
)
|
||||
|
@ -174,7 +200,7 @@ class BankWithIndex:
|
|||
|
||||
async def query_documents(
|
||||
self,
|
||||
query: InterleavedTextMedia,
|
||||
query: InterleavedContent,
|
||||
params: Optional[Dict[str, Any]] = None,
|
||||
) -> QueryDocumentsResponse:
|
||||
if params is None:
|
||||
|
|
Loading…
Add table
Add a link
Reference in a new issue