ldv-cli 0.11.0__tar.gz → 0.12.0__tar.gz

ldv_cli-0.12.0/.env

@@ -0,0 +1 @@
+PYPI_TOKEN=pypi-AgEIcHlwaS5vcmcCJDM5ODc5ZGY0LWExOGQtNDM0MS1iMjcxLTQxN2E3OGE4NTE3NAACKlszLCIyNGFlYWVlMC1jZDg3LTQ0MWEtYjBlYS1iYTRjYjFjZjRmMWEiXQAABiDokA2L5mJtlb8YQ6WUYuo7v_AL_wua3b-JObZoZY1g_w

{ldv_cli-0.11.0 → ldv_cli-0.12.0}/.gitignore

@@ -13,5 +13,3 @@ venv/
 
 # local config
 ~/.lql/
-
-.env

{ldv_cli-0.11.0 → ldv_cli-0.12.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ldv-cli
-Version: 0.11.0
+Version: 0.12.0
 Summary: ldv — CLI for the Liquid DataViewer platform (formerly lql)
 Project-URL: Homepage, https://github.com/Liquid4All/lql
 Author: Liquid AI
@@ -225,7 +225,10 @@ commands are the data primitives for error analysis: they slice and summarize
 the dataset, and you do the reasoning over what they return.
 
 ```
-ldv eval list [--workspace <id>]                 List eval datasets only
+ldv eval list [--workspace <id>] [--runid <id>] [--taskid <id>]
+                                                 List eval datasets only. --runid/--taskid filter by
+                                                 run<id>/task<id> in the name or parquet storage path
+                                                 (e.g. run11213_task72284.parquet); they AND together.
                                                  Defaults to LDV_EVAL_WORKSPACE; without a
                                                  workspace, lists only evals you own.
 ldv eval correctness <id>                        Fast accuracy + correct/incorrect/missing counts

{ldv_cli-0.11.0 → ldv_cli-0.12.0}/README.md

@@ -209,7 +209,10 @@ commands are the data primitives for error analysis: they slice and summarize
 the dataset, and you do the reasoning over what they return.
 
 ```
-ldv eval list [--workspace <id>]                 List eval datasets only
+ldv eval list [--workspace <id>] [--runid <id>] [--taskid <id>]
+                                                 List eval datasets only. --runid/--taskid filter by
+                                                 run<id>/task<id> in the name or parquet storage path
+                                                 (e.g. run11213_task72284.parquet); they AND together.
                                                  Defaults to LDV_EVAL_WORKSPACE; without a
                                                  workspace, lists only evals you own.
 ldv eval correctness <id>                        Fast accuracy + correct/incorrect/missing counts

{ldv_cli-0.11.0 → ldv_cli-0.12.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "ldv-cli"
-version = "0.11.0"
+version = "0.12.0"
 description = "ldv — CLI for the Liquid DataViewer platform (formerly lql)"
 readme = "README.md"
 requires-python = ">=3.12"

{ldv_cli-0.11.0 → ldv_cli-0.12.0}/src/ldv/commands/evals.py

@@ -1,6 +1,7 @@
 import json
 import math
 import os
+import re
 import sys
 from typing import Annotated, List, Optional
 
@@ -46,9 +47,24 @@ def _fmt_accuracy(acc: object) -> str:
     return f"{n * 100:.1f}%"
 
 
+# Fields a run/task id may appear in: the human name and the storage path. The
+# parquet name (e.g. run11213_task72284.parquet) is the reliable signal.
+_ID_FIELDS = ("display_name", "name", "hf_bucket_key", "hf_bucket", "s3_object_key", "hf_repo_id")
+
+
+def _filter_by_id(items: list, prefix: str, num: str) -> list:
+    """Keep datasets whose name/storage path contains ``<prefix><num>`` — e.g.
+    prefix 'run' + '11213' matches 'run11213', 'run 11213', 'run-11213'. The
+    trailing-digit guard means 1121 doesn't match 11213."""
+    pat = re.compile(rf"(?i)(?<![A-Za-z]){prefix}[\s_-]?{re.escape(num)}(?!\d)")
+    return [d for d in items if any(pat.search(str(d.get(f) or "")) for f in _ID_FIELDS)]
+
+
 @app.command("list")
 def list_evals(
     workspace: Annotated[Optional[str], typer.Option("--workspace", help="Workspace (defaults to LDV_EVAL_WORKSPACE)")] = None,
+    runid: Annotated[Optional[str], typer.Option("--runid", help="Only evals whose name/storage path contains this run id (e.g. 11213 -> run11213)")] = None,
+    taskid: Annotated[Optional[str], typer.Option("--taskid", help="Only evals whose name/storage path contains this task id (e.g. 72284 -> task72284)")] = None,
     json_out: JsonOpt = False,
     profile: ProfileOpt = None,
     api_url: ApiUrlOpt = None,
@@ -65,6 +81,11 @@ def list_evals(
             "to list the shared eval workspace.\n"
         )
     items = client.get("/v1/datasets", params=params).json()
+    # --runid / --taskid AND together (run11213_task72284 matches both).
+    if runid:
+        items = _filter_by_id(items, r"run", runid)
+    if taskid:
+        items = _filter_by_id(items, r"task(?:[\s_-]?id)?", taskid)
     print_table(
         ["ID", "Name", "Rows", "Source"],
         [

{ldv_cli-0.11.0 → ldv_cli-0.12.0}/src/ldv/commands/instructions.py

@@ -149,8 +149,12 @@ Eval datasets (evaluation-run output: each row a sample with a model 'response'
 + a 'correct' verdict) are detected automatically. These commands are the data
 primitives for error analysis — YOU do the reasoning over what they return.
 
-  ldv eval list [--workspace <id>]     # Eval datasets only. Defaults to LDV_EVAL_WORKSPACE;
+  ldv eval list [--workspace <id>] [--runid <id>] [--taskid <id>]
+                                       # Eval datasets only. Defaults to LDV_EVAL_WORKSPACE;
                                        # without a workspace it lists only evals you own.
+                                       # --runid / --taskid filter to evals whose name OR storage
+                                       # path matches run<id> / task<id> (e.g.
+                                       # run11213_task72284.parquet). They AND together.
   ldv eval stats <id>                  # Accuracy + correctness counts + error-type
                                        # distribution + token stats (the distribution view)
   ldv eval correctness <id>            # Fast accuracy + correct/incorrect/missing counts

{ldv_cli-0.11.0 → ldv_cli-0.12.0}/uv.lock

@@ -173,7 +173,7 @@ wheels = [
 
 [[package]]
 name = "ldv-cli"
-version = "0.10.0"
+version = "0.9.0"
 source = { editable = "." }
 dependencies = [
     { name = "httpx" },