From 6b2fb51b35374e76703767d5382059e43b01162a Mon Sep 17 00:00:00 2001 From: kumapo Date: Sun, 24 Sep 2023 15:55:21 +0900 Subject: [PATCH] re-evaluate models with jsquad prompt with title --- .../result.jsquad-1.2.json | 6 +++--- .../result.jsquad-1.2.json | 4 ++-- .../result.jsquad-1.2.json | 4 ++-- 3 files changed, 7 insertions(+), 7 deletions(-) diff --git a/models/rinna/rinna-japanese-gpt-neox-3.6b-instruction-ppo/result.jsquad-1.2.json b/models/rinna/rinna-japanese-gpt-neox-3.6b-instruction-ppo/result.jsquad-1.2.json index 86254d4282..9e2f33cb7f 100644 --- a/models/rinna/rinna-japanese-gpt-neox-3.6b-instruction-ppo/result.jsquad-1.2.json +++ b/models/rinna/rinna-japanese-gpt-neox-3.6b-instruction-ppo/result.jsquad-1.2.json @@ -1,8 +1,8 @@ { "results": { "jsquad-1.2-0.4": { - "exact_match": 53.89464205312922, - "f1": 65.73194869643035 + "exact_match": 52.633948671769474, + "f1": 64.387511749343 } }, "versions": { @@ -19,4 +19,4 @@ "bootstrap_iters": 100000, "description_dict": {} } -} +} \ No newline at end of file diff --git a/models/rinna/rinna-japanese-gpt-neox-3.6b-instruction-sft-v2/result.jsquad-1.2.json b/models/rinna/rinna-japanese-gpt-neox-3.6b-instruction-sft-v2/result.jsquad-1.2.json index 92a360030a..5de18a40c1 100644 --- a/models/rinna/rinna-japanese-gpt-neox-3.6b-instruction-sft-v2/result.jsquad-1.2.json +++ b/models/rinna/rinna-japanese-gpt-neox-3.6b-instruction-sft-v2/result.jsquad-1.2.json @@ -1,8 +1,8 @@ { "results": { "jsquad-1.2-0.4": { - "exact_match": 47.90634849167042, - "f1": 62.1059309037734 + "exact_match": 47.54615038271049, + "f1": 61.633765369013354 } }, "versions": { diff --git a/models/rinna/rinna-japanese-gpt-neox-3.6b-instruction-sft/result.jsquad-1.2.json b/models/rinna/rinna-japanese-gpt-neox-3.6b-instruction-sft/result.jsquad-1.2.json index 77948c9677..d2001d0058 100644 --- a/models/rinna/rinna-japanese-gpt-neox-3.6b-instruction-sft/result.jsquad-1.2.json +++ b/models/rinna/rinna-japanese-gpt-neox-3.6b-instruction-sft/result.jsquad-1.2.json @@ -1,8 +1,8 @@ { "results": { "jsquad-1.2-0.4": { - "exact_match": 49.392165691130124, - "f1": 63.56485708061489 + "exact_match": 49.34714092751013, + "f1": 63.33718413567939 } }, "versions": {