Open
xiaobie-lhm
opened this issue 1 year ago
•
5 comments
Describe the bug
[Bug]:
raise ValueError("Columns must be same length as key")\nValueError: Columns must be same length as key\n", "source": "Columns must be same length as key", "details": null}
<h3>Steps to reproduce</h3>
<p><em>No response</em></p>
<h3>Expected Behavior</h3>
<p><em>No response</em></p>
<h3>GraphRAG Config Used</h3>
<p>encoding_model: cl100k_base
skip_workflows: []
llm:
api_key: ollama
type: openai_chat # or azure_openai_chat
model: gemma2
model_supports_json: true # recommended if this is available for your model.</p>
<h1>max_tokens: 4000</h1>
<h1>request_timeout: 180.0</h1>
<p>api_base: https://localhost:11434/v1</p>
<h1>api_version: 2024-02-15-preview</h1>
<h1>organization: <organization_id></h1>
<h1>deployment_name: <azure_model_deployment_name></h1>
<h1>tokens_per_minute: 150_000 # set a leaky bucket throttle</h1>
<h1>requests_per_minute: 10_000 # set a leaky bucket throttle</h1>
<h1>max_retries: 10</h1>
<h1>max_retry_wait: 10.0</h1>
<h1>sleep_on_rate_limit_recommendation: true # whether to sleep when azure suggests wait-times</h1>
<h1>concurrent_requests: 25 # the number of parallel inflight requests that may be made</h1>
<p>parallelization:
stagger: 0.3</p>
<h1>num_threads: 50 # the number of threads to use for parallel processing</h1>
<p>async_mode: threaded # or asyncio</p>
<p>embeddings:</p>
<h2>parallelization: override the global parallelization settings for embeddings</h2>
<p>async_mode: threaded # or asyncio
llm:
api_key: lm-studio
type: openai_embedding # or azure_openai_embedding
model: Publisher/Repository/nomic-embed-text-v1.5.Q5_K_M.gguf
api_base: http://localhost:1234/v1
# api_version: 2024-02-15-preview
# organization: <organization_id>
# deployment_name: <azure_model_deployment_name>
# tokens_per_minute: 150_000 # set a leaky bucket throttle
# requests_per_minute: 10_000 # set a leaky bucket throttle
# max_retries: 10
# max_retry_wait: 10.0
# sleep_on_rate_limit_recommendation: true # whether to sleep when azure suggests wait-times
# concurrent_requests: 25 # the number of parallel inflight requests that may be made
# batch_size: 16 # the number of documents to send in a single request
# batch_max_tokens: 8191 # the maximum number of tokens to send in a single request
# target: required # or optional</p>
<p>chunks:
size: 300
overlap: 100
group_by_columns: [id] # by default, we don't allow chunks to cross documents</p>
<p>input:
type: file # or blob
file_type: text # or csv
base_dir: "input"
file_encoding: utf-8
file_pattern: ".*\.txt$"</p>
<p>cache:
type: file # or blob
base_dir: "cache"</p>
<h1>connection_string: <azure_blob_storage_connection_string></h1>
<h1>container_name: <azure_blob_storage_container_name></h1>
<p>storage:
type: file # or blob
base_dir: "output/${timestamp}/artifacts"</p>
<h1>connection_string: <azure_blob_storage_connection_string></h1>
<h1>container_name: <azure_blob_storage_container_name></h1>
<p>reporting:
type: file # or console, blob
base_dir: "output/${timestamp}/reports"</p>
<h1>connection_string: <azure_blob_storage_connection_string></h1>
<h1>container_name: <azure_blob_storage_container_name></h1>
<p>entity_extraction:</p>
<h2>llm: override the global llm settings for this task</h2>
<h2>parallelization: override the global parallelization settings for this task</h2>
<h2>async_mode: override the global async_mode settings for this task</h2>
<p>prompt: "prompts/entity_extraction.txt"
entity_types: [organization,person,geo,event]
max_gleanings: 0</p>
<p>summarize_descriptions:</p>
<h2>llm: override the global llm settings for this task</h2>
<h2>parallelization: override the global parallelization settings for this task</h2>
<h2>async_mode: override the global async_mode settings for this task</h2>
<p>prompt: "prompts/summarize_descriptions.txt"
max_length: 500</p>
<p>claim_extraction:</p>
<h2>llm: override the global llm settings for this task</h2>
<h2>parallelization: override the global parallelization settings for this task</h2>
<h2>async_mode: override the global async_mode settings for this task</h2>
<h1>enabled: true</h1>
<p>prompt: "prompts/claim_extraction.txt"
description: "Any claims or facts that could be relevant to information discovery."
max_gleanings: 0</p>
<p>community_report:</p>
<h2>llm: override the global llm settings for this task</h2>
<h2>parallelization: override the global parallelization settings for this task</h2>
<h2>async_mode: override the global async_mode settings for this task</h2>
<p>prompt: "prompts/community_report.txt"
max_length: 2000
max_input_length: 8000</p>
<p>cluster_graph:
max_cluster_size: 10</p>
<p>embed_graph:
enabled: false # if true, will generate node2vec embeddings for nodes</p>
<h1>num_walks: 10</h1>
<h1>walk_length: 40</h1>
<h1>window_size: 2</h1>
<h1>iterations: 3</h1>
<h1>random_seed: 597832</h1>
<p>umap:
enabled: false # if true, will generate UMAP embeddings for nodes</p>
<p>snapshots:
graphml: false
raw_entities: false
top_level_nodes: false</p>
<p>local_search:</p>
<h1>text_unit_prop: 0.5</h1>
<h1>community_prop: 0.1</h1>
<h1>conversation_history_max_turns: 5</h1>
<h1>top_k_mapped_entities: 10</h1>
<h1>top_k_relationships: 10</h1>
<h1>max_tokens: 12000</h1>
<p>global_search:</p>
<h1>max_tokens: 12000</h1>
<h1>data_max_tokens: 12000</h1>
<h1>map_max_tokens: 1000</h1>
<h1>reduce_max_tokens: 2000</h1>
<h1>concurrency: 32</h1>
<h3>Logs and screenshots</h3>
<p><em>No response</em></p>
<h3>Additional Information</h3>
<ul>
<li>GraphRAG Version:</li>
<li>Operating System:</li>
<li>Python Version:</li>
<li>Related Issues:</li>
</ul>
I uses the same configuration as yours, but there is the same bug.
pip install graphrag
mkdir -p ./ragtest/input
curl https://www.gutenberg.org/cache/epub/24022/pg24022.txt > ./ragtest/input/book.txt
python -m graphrag.index --init --root ./ragtest
change .env and setting.yaml file, the same content as yours
python -m graphrag.index --root ./ragtest
raise ValueError("Columns must be same length as key")\nValueError: Columns must be same length as key\n"
Using Xinference for models will solve this problem.
what is Xinference and how to use it?
Refer to: https://inference.readthedocs.io/en/latest/.
Xinference is a large language model inference framework that supports LLM and embedding models, and can seamlessly call OpenAI interfaces.
microsoft
