feat(evaluation): adding examples and repetitions to create tracer sessions calls

[Shamik Karkhanis (shamikkarkhanis)]

Description

LSE-2266. Adds support for experiment loading state tracking by including num_examples and num_repetitions in session creation call to langsmith.

Release Notes

none

Test Plan

• cd python && uv run pytest tests/unit_tests/evaluation/test_runner.py -k "test_resolve_num_examples or test_evaluate_sends_num_examples" -v --PASSES
• cd js && pnpm test src/tests/evaluate_runner.test.ts -t "evaluate forwards num_examples" --PASSES

[github-actions]

JS perf benchmark

Lower is better. Noisy on shared runners — treat as a signal, not a gate.

Base64-heavy payload

Single large base64 string per message — the shape the worker-offload path is optimized for.
Payload: 2511.2 KB in / 5.2 KB out, 100 runs.

metric	main	this PR	delta
Wall time (ms)	2003.39	1797.26	-10.3%
createRun total (ms)	74.48	98.97	+32.9%
createRun p50 (ms)	0.59	0.61	+2.2%
createRun p95 (ms)	0.78	3.66	+366.3%
createRun p99 (ms)	27.80	33.62	+21.0%
createRun max (ms)	27.80	33.62	+21.0%
updateRun total (ms)	42.21	47.80	+13.2%
updateRun p95 (ms)	0.85	1.17	+36.8%
loop lag total (ms)	1040.95	949.42	-8.8%
loop lag p50 (ms)	0.09	0.10	+2.2%
loop lag p95 (ms)	4.21	4.85	+15.2%
loop lag p99 (ms)	12.71	18.93	+48.9%
loop lag max (ms)	217.85	113.44	-47.9%

Structural payload

Many small strings across a wide/nested object graph. Should bypass worker offload and use sync flush.
Payload: 1239.5 KB in / 13.3 KB out, 100 runs.

metric	main	this PR	delta
Wall time (ms)	1404.77	1396.33	-0.6%
createRun total (ms)	394.54	384.33	-2.6%
createRun p50 (ms)	3.41	3.35	-1.6%
createRun p95 (ms)	6.42	5.38	-16.3%
createRun p99 (ms)	7.20	14.56	+102.2%
createRun max (ms)	7.20	14.56	+102.2%
updateRun total (ms)	29.58	29.29	-1.0%
updateRun p95 (ms)	0.45	0.50	+11.5%
loop lag total (ms)	972.98	969.16	-0.4%
loop lag p50 (ms)	0.07	0.07	-0.2%
loop lag p95 (ms)	4.65	4.88	+4.9%
loop lag p99 (ms)	118.04	117.91	-0.1%
loop lag max (ms)	157.39	154.84	-1.6%