New Features

The main changes in this release are:

• AutoPDL: automatic optimization of PDL programs
• Calling PDL functions from Jinja and Python
• Map/reduce block

AutoPDL: automatic optimization of PDL programs

Given a PDL program with some free variables, a search space for these free variable, a loss function, and a training set, you can optimize the PDL program using the pdl-optimize command. This allows to optimize any part of a PDL program such as the textual prompts, the few-shots examples, or the prompting patterns. This work is based on the paper AutoPDL: Automatic Prompt Optimization for LLM Agents. A new section of the manual is describing how to use the optimizer.

Calling PDL functions from Jinja and Python

It is now possible to call a function defined in PDL inside a Jinja expression or a Python code block. Here is an example where the function translate is called in the Jinja expression ${ translate("Hello", language="French") }:

description: Calling a PDL function from Jinja
defs:
  translate:
    function:
      sentence: string
      language: string
    return:
      lastOf:
      - |
        Translate the sentence '${ sentence }' to ${ language }.
        Only give the result of the translation.
      - model: ollama_chat/granite3.2:2b
text: |
  The way to say hello in French is ${ translate("Hello", language="French") }.

Map/reduce block

Similar to the repeat loops, PDL now offers a map block. The difference with the repeat block is that the context is not accumulated between iterations, each iteration is executed with the same context. Here is an example of map block:

lastOf:
- "Hello, "
- for:
    name: [Alice, Bob, Charlie]
  map:
    lastOf:
    - my name is ${ name }
    - model: ollama/granite3.2:2b
  join:
    as: array

What's Changed

• chore: Configure Renovate by @renovate[bot] in chore: Configure Renovate #944
• Dataset preprocessing scripts for AutoPDL by @claudiosv in Dataset preprocessing scripts for AutoPDL #994
• fix: avoid side effects in PdlContext by @mandel in fix: avoid side effects in PdlContext #1054
• tests: update react example by @mandel in tests: update react example #1001
• tests: update calc example to use .signature by @mandel in tests: update calc example to use .signature #1067
• fix: output streaming in case of fallback by @mandel in fix: output streaming in case of fallback #1071
• feat: make PDL functions callable as Python and jinja functions by @mandel in feat: make PDL functions callable as Python and jinja functions #1070
• feat: allow unnamed arguments when calling a PDL function in Jinja by @mandel in feat: allow unnamed arguments when calling a PDL function in Jinja #1076
• feat: map/reduce by @mandel in feat: map/reduce #1010
• refactor: refactor AST to generate simpler JSON Schema by @mandel in refactor: refactor AST to generate simpler JSON Schema #1079
• Fix for exception in optimizer by @vazirim in Fix for exception in optimizer #1084
• docs: update tutorial by @mandel in docs: update tutorial #1080
• Requirements implementation by @vazirim in Requirements implementation #1097
• fix: update AST in the UI by @mandel in fix: update AST in the UI #1100
• fix: ensure that PazyApply.result returns a non-lazy value by @mandel in fix: ensure that PazyApply.result returns a non-lazy value #1101
• feat: parallel map reduce by @mandel in feat: parallel map reduce #1102
• Fix for llm message post-processing by @vazirim in Fix for llm message post-processing #1106
• AutoPDL Simplification by @claudiosv in AutoPDL Simplification #1088
• tests: map-based gsm8k by @mandel in tests: map-based gsm8k #1107
• optimizer CLI and doc by @vazirim in optimizer CLI and doc #1151

Full Changelog: v0.7.1...v0.8.0