Update ExtractBench results (shared 32k-budget fairness re-run)

#5
Files changed (1) hide show
  1. .eval_results/extractbench.yaml +12 -12
.eval_results/extractbench.yaml CHANGED
@@ -1,40 +1,40 @@
1
  - dataset:
2
  id: llamaindex/ExtractBench
3
  task_id: mean
4
- value: 82.35
5
- date: '2026-08-20'
6
  source:
7
  url: https://huggingface.co/datasets/llamaindex/ExtractBench
8
  name: ExtractBench
9
  user: boyang-runllama
10
- notes: "Pipeline name: nuextract3_extract; served checkpoint numind/NuExtract3 on vLLM, one-shot template-guided extraction (the JSON Schema is converted to a NuExtract template)"
11
  - dataset:
12
  id: llamaindex/ExtractBench
13
  task_id: short
14
- value: 88.06
15
- date: '2026-08-20'
16
  source:
17
  url: https://huggingface.co/datasets/llamaindex/ExtractBench
18
  name: ExtractBench
19
  user: boyang-runllama
20
- notes: "Pipeline name: nuextract3_extract; served checkpoint numind/NuExtract3 on vLLM, one-shot template-guided extraction (the JSON Schema is converted to a NuExtract template)"
21
  - dataset:
22
  id: llamaindex/ExtractBench
23
  task_id: medium
24
- value: 76.76
25
- date: '2026-08-20'
26
  source:
27
  url: https://huggingface.co/datasets/llamaindex/ExtractBench
28
  name: ExtractBench
29
  user: boyang-runllama
30
- notes: "Pipeline name: nuextract3_extract; served checkpoint numind/NuExtract3 on vLLM, one-shot template-guided extraction (the JSON Schema is converted to a NuExtract template)"
31
  - dataset:
32
  id: llamaindex/ExtractBench
33
  task_id: long
34
- value: 37.72
35
- date: '2026-08-20'
36
  source:
37
  url: https://huggingface.co/datasets/llamaindex/ExtractBench
38
  name: ExtractBench
39
  user: boyang-runllama
40
- notes: "Pipeline name: nuextract3_extract; served checkpoint numind/NuExtract3 on vLLM, one-shot template-guided extraction (the JSON Schema is converted to a NuExtract template)"
 
1
  - dataset:
2
  id: llamaindex/ExtractBench
3
  task_id: mean
4
+ value: 81.02
5
+ date: '2026-08-26'
6
  source:
7
  url: https://huggingface.co/datasets/llamaindex/ExtractBench
8
  name: ExtractBench
9
  user: boyang-runllama
10
+ notes: "Pipeline name: nuextract3_extract; self-hosted on vLLM, JSON Schema converted to NuExtract template"
11
  - dataset:
12
  id: llamaindex/ExtractBench
13
  task_id: short
14
+ value: 86.41
15
+ date: '2026-08-26'
16
  source:
17
  url: https://huggingface.co/datasets/llamaindex/ExtractBench
18
  name: ExtractBench
19
  user: boyang-runllama
20
+ notes: "Pipeline name: nuextract3_extract; self-hosted on vLLM, JSON Schema converted to NuExtract template"
21
  - dataset:
22
  id: llamaindex/ExtractBench
23
  task_id: medium
24
+ value: 77.01
25
+ date: '2026-08-26'
26
  source:
27
  url: https://huggingface.co/datasets/llamaindex/ExtractBench
28
  name: ExtractBench
29
  user: boyang-runllama
30
+ notes: "Pipeline name: nuextract3_extract; self-hosted on vLLM, JSON Schema converted to NuExtract template"
31
  - dataset:
32
  id: llamaindex/ExtractBench
33
  task_id: long
34
+ value: 32.71
35
+ date: '2026-08-26'
36
  source:
37
  url: https://huggingface.co/datasets/llamaindex/ExtractBench
38
  name: ExtractBench
39
  user: boyang-runllama
40
+ notes: "Pipeline name: nuextract3_extract; self-hosted on vLLM, JSON Schema converted to NuExtract template"