lewtun HF Staff commited on
Commit
16b3acc
·
verified ·
1 Parent(s): f81d1c0

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +15 -0
README.md CHANGED
@@ -48,6 +48,21 @@ tags:
48
  win rates. Built to compare the rule-based chatbot from the run4 agent swarm
49
  (`cmpatino/chatbot-swarm`) at different commits, and small LMs, against one fixed reference.
50
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
51
  ## Construction
52
 
53
  - Source: the `everyday-conversations` subset of [`HuggingFaceTB/smoltalk`](https://huggingface.co/datasets/HuggingFaceTB/smoltalk)
 
48
  win rates. Built to compare the rule-based chatbot from the run4 agent swarm
49
  (`cmpatino/chatbot-swarm`) at different commits, and small LMs, against one fixed reference.
50
 
51
+ ```python
52
+ from datasets import load_dataset
53
+
54
+ ds = load_dataset("HuggingFaceSwarms/SmolTalkBench", "gold", split="test")
55
+ ```
56
+
57
+ ## Configs
58
+
59
+ | config | reference replies (`output`) |
60
+ |---|---|
61
+ | `gold` (default) | the source conversations' replies, by `meta-llama/Meta-Llama-3.1-70B-Instruct` |
62
+
63
+ Variants with other reference models go in their own configs, keeping the same `id`s and
64
+ instructions so that results compare across them.
65
+
66
  ## Construction
67
 
68
  - Source: the `everyday-conversations` subset of [`HuggingFaceTB/smoltalk`](https://huggingface.co/datasets/HuggingFaceTB/smoltalk)