-
|
Hi, I've some difficulties to use not openai embeddings. In reference with the example provided in chat-multi-extract-3.py at the beginning of the code I added this definition: then I've added this: just after the original code: The program runs until it needs to create the embeddings the this happens Following the original code with the added lines |
Beta Was this translation helpful? Give feedback.
Answered by
shelaik
Mar 26, 2024
Replies: 1 comment 2 replies
-
|
Thanks for trying Langroid. The problem is that this line: does not actually change the vecdb, since the |
Beta Was this translation helpful? Give feedback.
2 replies
-
|
Thanks you for your reply, I've changed the DocChatAgentConfig in that way adding the LanceDBConfig: doc_agent = DocChatAgent(
DocChatAgentConfig(
llm=llm_cfg,
n_neighbor_chunks=2,
parsing=ParsingConfig(
chunk_size=50,
overlap=10,
n_similar_docs=3,
n_neighbor_ids=4,
),
cross_encoder_reranking_model="",
vecdb = lr.vector_store.LanceDBConfig(
collection_name="docqa-chat-multi-extract", replace_collection=True, embedding=embed_cfg
),
)after defining this I've also set a dummy Also I've commented this line However now I'm getting this error: ╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮
│ G:\envs\langroid\chat-multi-extract-3.py:152 in main │
│ │
│ 149 │ │ │ │ n_neighbor_ids=4, │
│ 150 │ │ │ ), │
│ 151 │ │ │ cross_encoder_reranking_model="", │
│ ❱ 152 │ │ │ vecdb = lr.vector_store.LanceDBConfig( │
│ 153 │ │ │ │ collection_name="docqa-chat-multi-extract", replace_collection=True, emb │
│ 154 │ │ │ │ ), │
│ 155 │ │ ) │
│ │
│ ╭─────────────────────────────────── locals ───────────────────────────────────╮ │
│ │ debug = False │ │
│ │ llm_cfg = OpenAIGPTConfig( │ │
│ │ │ type='openai', │ │
│ │ │ streamer=<function noop_fn at 0x000001AA847DF880>, │ │
│ │ │ api_base=None, │ │
│ │ │ formatter=None, │ │
│ │ │ timeout=45, │ │
│ │ │ chat_model='ollama/llama2:13b-chat-q8_0', │ │
│ │ │ completion_model='gpt-3.5-turbo-instruct', │ │
│ │ │ temperature=0.0, │ │
│ │ │ chat_context_length=16000, │ │
│ │ │ completion_context_length=8000, │ │
│ │ │ max_output_tokens=1024, │ │
│ │ │ min_output_tokens=1, │ │
│ │ │ use_completion_for_chat=False, │ │
│ │ │ use_chat_for_completion=True, │ │
│ │ │ stream=True, │ │
│ │ │ cache_config=None, │ │
│ │ │ chat_cost_per_1k_tokens=(0.0, 0.0), │ │
│ │ │ completion_cost_per_1k_tokens=(0.0, 0.0), │ │
│ │ │ api_key='sk-111111111111111111111111111111111111111111111111', │ │
│ │ │ organization='', │ │
│ │ │ litellm=False, │ │
│ │ │ ollama=False, │ │
│ │ │ seed=42, │ │
│ │ │ params=None, │ │
│ │ │ run_on_first_use=<function noop at 0x000001AA85F98680>, │ │
│ │ │ hf_formatter=None │ │
│ │ ) │ │
│ │ model = 'ollama/llama2:13b-chat-q8_0' │ │
│ │ nocache = False │ │
│ ╰──────────────────────────────────────────────────────────────────────────────╯ │
│ │
│ in pydantic.env_settings.BaseSettings.__init__:40 │
│ │
│ in pydantic.main.BaseModel.__init__:341 │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
ValidationError: 1 validation error for LanceDBConfig
embedding
value is not a valid dict (type=type_error.dict)Not sure where I'm wrong, thanks for the help. |
Beta Was this translation helpful? Give feedback.
-
|
solved: Thanks. |
Beta Was this translation helpful? Give feedback.
Answer selected by
shelaik
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
solved:
I changed this:
embed_cfg = SentenceTransformerEmbeddings()in
embed_cfg = SentenceTransformerEmbeddingsConfig()Thanks.