The AI Toolchain — issue -406, January 31, 2022

Haystack v1.1.0 adds model distillation, isolated pipeline eval, RCIReader for TableQA, ParsrConverter, and nDCG metrics.

$ git clone --branch v1.1.0 https://github.com/deepset-ai/haystack.git
# already have the repo? check out this version:
$ git checkout v1.1.0

# Step 1: augment training data
python augment_squad.py --squad_path squad2.json --output_path augmented_squad2.json --multiplication_factor 20

# Step 2: distil intermediate layers
student.distil_intermediate_layers_from(teacher, data_dir="dataset", train_filename="augmented_squad2.json")

# Step 3: distil prediction layer
student.distil_prediction_layer_from(teacher, data_dir="dataset", train_filename="squad2.json")

eval_result = pipeline.eval(labels=eval_labels, add_isolated_node_eval=True)
pipeline.print_eval_report(eval_result)

from haystack.nodes import RCIReader

reader = RCIReader(
    row_model_name_or_path="michaelrglass/albert-base-rci-wikisql-row",
    column_model_name_or_path="michaelrglass/albert-base-rci-wikisql-col"
)

• ›Adds student.distil_intermediate_layers_from(teacher, data_dir=..., train_filename=...) and student.distil_prediction_layer_from(teacher, data_dir=..., train_filename=...) methods to compress large reader models (teacher) into smaller models (student) via TinyBERT-style distillation, with a companion augment_squad.py --squad_path <your dataset> --output_path <output> --multiplication_factor 20 data-augmentation script.
• ›Adds add_isolated_node_eval=True parameter to pipeline.eval() and pipeline.print_eval_report() to expose per-node upper-bound metrics alongside integrated metrics, enabling bottleneck identification in pipelines such as ExtractiveQAPipeline.
• ›Adds nDCG to pipeline.eval()'s document metrics.
• ›Adds RCIReader(row_model_name_or_path=..., column_model_name_or_path=...) for TableQA using Row-Column-Intersection models, supporting larger tables and returning meaningful confidence scores unlike TableReader.
• ›Adds ParsrConverter (based on the open-source axa-group Parsr tool) for extracting text and tables from PDF and DOCX files in a format directly usable for TableQA.

Triton 2.18.0 adds implicit state management for ONNX/TensorRT, Inferentia batching, Jetson backend support, and ARM analyzer support.

$ git clone --branch v2.18.0 https://github.com/triton-inference-server/server.git
# already have the repo? check out this version:
$ git checkout v2.18.0

• ›Implicit state management is now supported for the ONNX Runtime and TensorRT backends.
• ›State initialization from a constant is now supported within implicit state management.
• ›PyTorch and TensorFlow models now support batching on Inferentia.
• ›PyTorch and Python backends are now supported on Jetson.
• ›ARM support added for the Performance Analyzer and Model Analyzer.

Zed v0.13 adds Go To Definition (F12) and navigation history shortcuts for Rust symbols.

$ git clone --branch v0.13 https://github.com/zed-industries/zed.git
# already have the repo? check out this version:
$ git checkout v0.13

• ›Adds Go To Definition for Rust symbols via F12, jumping to where any variable or function is declared.
• ›Adds navigation history with ctrl+- (back) and ctrl+shift+- (forward) to move between visited locations.
• ›Improves styling of the Project Diagnostics panel.

$ git clone --branch v0.12 https://github.com/zed-industries/zed.git
# already have the repo? check out this version:
$ git checkout v0.12

$ git clone --branch v0.11 https://github.com/zed-industries/zed.git
# already have the repo? check out this version:
$ git checkout v0.11

Weaviate v1.10.0 adds OpenAI embeddings, QnA reranking, HNSW EF boundaries, and a HEAD /v1/objects/{id} existence check.

$ git clone --branch v1.10.0 https://github.com/weaviate/weaviate.git
# already have the repo? check out this version:
$ git checkout v1.10.0

$ curl -s -o /dev/null -w "%{http_code}" -X HEAD http://localhost:8080/v1/objects/<id>

{
  "class": "Article",
  "vectorIndexConfig": {
    "dynamicEfMin": 100,
    "dynamicEfMax": 500,
    "dynamicEfFactor": 8
  }
}

• ›Adds HEAD /v1/objects/{id} endpoint that returns 204 when an object exists or 404 when it does not, without loading or unmarshaling the full object from disk.
• ›Adds ask: { rerank: true } to the QnA module so that multiple answer candidates are drawn from the top-n results and re-ranked by qna-specific score rather than always extracting from the single top semantic result.
• ›Adds dynamicEfMin (default 100), dynamicEfMax (default 500), and dynamicEfFactor (default 8) HNSW config parameters to bound and tune automatic ef derivation at query time.
• ›Adds the text2vec-openai module, enabling OpenAI embeddings as a vectorizer for both import and query inference with a valid OpenAI API key.
• ›Allows importing objects without a vector when vector indexing is enabled, so vectors can be added later via an update.