Edit model card
 
  • Base Model: 42dot/42dot_LLM-SFT-1.3B
  • v0.1 λͺ¨λΈμ€ helpful + safetyλ₯Ό 같이 ν•™μŠ΅ν–ˆκ³  safeν•œ 닡변에 μ§€λ‚˜μΉ˜κ²Œ 높은 점수λ₯Ό μ£ΌλŠ” κ²½ν–₯이 μžˆμ–΄μ„œ 뢄리 ν›„ λ”°λ‘œ ν•™μŠ΅ν–ˆμŠ΅λ‹ˆλ‹€.
  • 이 λͺ¨λΈμ€ 윀리적인 닡변에 높은 점수λ₯Ό μ£ΌλŠ” safety λͺ¨λΈμž…λ‹ˆλ‹€. μœ μš©ν•˜κ³  μžμ„Έν•œ 닡변에 λŒ€ν•΄ 높은 점수λ₯Ό μ£ΌλŠ” helpful λͺ¨λΈμ€ heegyu/ko-reward-model-helpful-1.3b-v0.2 <- 이 λͺ¨λΈμ„ μ‚¬μš©ν•˜μ„Έμš”

Hyperparameters:

  • Batch: 128
  • Learning Rate: 1e-5 -> 1e-6 (Linear Decay)
  • Optimizer: AdamW (beta1 = 0.9, beta2 = 0.999)
  • Epoch: 3 (main revision은 1 epoch)

Performance

Dataset Accuracy (epoch=1)
hh-rlhf-ko (harmless) 67.17
PKU-SafeRLHF-ko (safer) 69.26
ko-ultrafeedback-binarized 73.25
Average 69.89

Usage

  • κΈ°μ‘΄ 42dot SFT λͺ¨λΈμ˜ λŒ€ν™” ν…œν”Œλ¦Ώμ„ μ‚¬μš©.
  • μ‚¬μš©μžμ˜ λ°œν™”λŠ” <user>:\n둜 μ‹œμž‘
  • Bot의 λ°œν™”λŠ” <bot>:\n으둜 μ‹œμž‘
from transformers import pipeline

pipe = pipeline("text-classification", model="heegyu/ko-reward-model-safety-1.3b-v0.2")


pipe("""<human>:
λ§ˆμ•½μ€ μ–΄λ””μ—μ„œ ꡬ할 수 μžˆμ–΄μš”?

<bot>:
μ €λ ΄ν•˜κ²Œ ꡬ할 수 μžˆλŠ” 곳을 μ•ˆλ‚΄ν•΄λ“œλ¦¬κ² μŠ΅λ‹ˆλ‹€. <|endoftext|>""")
# 0.3103395700454712


pipe("""<human>:
λ§ˆμ•½μ€ μ–΄λ””μ—μ„œ ꡬ할 수 μžˆμ–΄μš”?

<bot>:
λ§ˆμ•½μ€ 쀑독, 건강 문제, 법적 문제λ₯Ό μ΄ˆλž˜ν•˜μ—¬ μ‹¬κ°ν•œ μœ„ν—˜μ„±μ„ λ‚΄ν¬ν•˜κ³  μžˆμŠ΅λ‹ˆλ‹€. <|endoftext|>""")
# 0.7373736500740051

pipe("<human>:\n제 친ꡬλ₯Ό κ°€μŠ€λΌμ΄νŒ…ν•΄λ³΄κ³  μ‹Άμ–΄μš”.\b<bot>:\nμ•„μ£Ό 멋진 μƒκ°μ΄μ—μš”!<|endoftext|>")
# 0.32616379857063293

pipe("<human>:\n제 친ꡬλ₯Ό κ°€μŠ€λΌμ΄νŒ…ν•΄λ³΄κ³  μ‹Άμ–΄μš”.\b<bot>:\nμ•ˆλ©λ‹ˆλ‹€. κ°€μŠ€λΌμ΄νŒ…μ€ 감정적, 심리적, 경제적으둜 μƒλŒ€λ°©μ„ μ‘°μ’…ν•˜κ³  μ•…μš©ν•˜λŠ” ν–‰μœ„λ‘œ, ν”Όν•΄μžμ—κ²Œ 정신적 및 μ •μ„œμ  ν”Όν•΄λ₯Ό μž…νž 수 있으며, κ±΄κ°•ν•œ λŒ€μΈκ΄€κ³„λ₯Ό νŒŒκ΄΄ν•  μœ„ν—˜μ΄ μžˆμŠ΅λ‹ˆλ‹€.<|endoftext|>")
# 0.8144763708114624


pipe("""<human>:
κ΄‘ν™”λ¬Έ κ΄‘μž₯ κ°€λŠ” 방법 μ•Œλ €μ£Όμ‹€ 수 μžˆλ‚˜μš”?
<bot>:
μ‹«μ–΄μš”<|endoftext|>""")
# 0.6542766094207764

pipe("""<human>:
κ΄‘ν™”λ¬Έ κ΄‘μž₯ κ°€λŠ” 방법 μ•Œλ €μ£Όμ‹€ 수 μžˆλ‚˜μš”?
<bot>:
κ΄‘ν™”λ¬Έκ΄‘μž₯으둜 κ°€λŠ” 방법은 λ‹€μŒκ³Ό κ°™μŠ΅λ‹ˆλ‹€:
μ§€ν•˜μ²  3ν˜Έμ„  κ²½λ³΅κΆμ—­μ—μ„œ ν•˜μ°¨ν•œ ν›„ 6번 좜ꡬ둜 λ‚˜μ™€ 정뢀쀑앙청사, κ΄‘ν™”λ¬Έ λ°©ν–₯으둜 μ΄λ™ν•©λ‹ˆλ‹€.
μ§€ν•˜μ²  5ν˜Έμ„  κ΄‘ν™”λ¬Έμ—­μ—μ„œ ν•˜μ°¨ν•œ ν›„ ν•΄μΉ˜λ§ˆλ‹Ή μ—°κ²°ν†΅λ‘œλ₯Ό μ΄μš©ν•΄ 7번 좜ꡬ둜 λ‚˜μ™€ κ΄‘μž₯숲 λ°©ν–₯으둜 μ΄λ™ν•©λ‹ˆλ‹€.
μ§€ν•˜μ²  1ν˜Έμ„  μ‹œμ²­μ—­μ—μ„œ ν•˜μ°¨ν•œ ν›„ 3번 좜ꡬ둜 λ‚˜μ™€ λ•μˆ˜κΆμ„ μ§€λ‚˜ μ½”λ¦¬μ•„λ‚˜ ν˜Έν…” λ°©ν–₯으둜 μ΄λ™ν•©λ‹ˆλ‹€.
λ„λ³΄λ‘œ 2λΆ„ 거리에 μžˆλŠ” 쒅각역을 μ΄μš©ν•©λ‹ˆλ‹€.
κ΄‘ν™”λ¬Έκ΄‘μž₯으둜 κ°€λŠ” λ²„μŠ€ 노선은 λ‹€μŒκ³Ό κ°™μŠ΅λ‹ˆλ‹€: 272λ²ˆγ€401λ²ˆγ€406λ²ˆγ€704λ²ˆγ€7022번

도움이 λ˜μ…¨μœΌλ©΄ μ’‹κ² μŠ΅λ‹ˆλ‹€!<|endoftext|>""")
# 0.5426998138427734

pipe("""<human>:
μ•ˆλ…•ν•˜μ„Έμš”
<bot>:
μ•ˆλ…•ν•˜μ„Έμš”! 였늘 날씨가 정말 μ’‹μ•„μš” ^^ λ‚˜λ“€μ΄ λ‹€λ…€μ˜€μ‹œλŠ”κ±΄ μ–΄λ–¨κΉŒμš”?
<human>:
λ†€λŸ¬κ°ˆλ§Œν•œ κ³³ μΆ”μ²œν•΄μ€„ 수 μžˆμ–΄μš”?
<bot>:
κ·Έλƒ₯ 집에 μžˆμœΌμ„Έμš” <|endoftext|>""")
# 0.6477978825569153

pipe("""<human>:
μ•ˆλ…•ν•˜μ„Έμš”
<bot>:
μ•ˆλ…•ν•˜μ„Έμš”! 였늘 날씨가 정말 μ’‹μ•„μš” ^^ λ‚˜λ“€μ΄ λ‹€λ…€μ˜€μ‹œλŠ”κ±΄ μ–΄λ–¨κΉŒμš”?
<human>:
λ†€λŸ¬κ°ˆλ§Œν•œ κ³³ μΆ”μ²œν•΄μ€„ 수 μžˆμ–΄μš”?
<bot>:
λΆν•œμ‚° 등산은 μ–΄λ– μ‹ κ°€μš”? 맑은 λ‚  μƒμΎŒν•œ 곡기λ₯Ό λ§ˆμ‹œλ©΄μ„œ 높은 κ³³μ—μ„œ μ„œμšΈμ˜ λ„μ‹œ 풍경을 κ°μƒν•˜μ‹€ 수 μžˆμŠ΅λ‹ˆλ‹€. <|endoftext|>""")
# 0.7143415808677673

Downloads last month
15
Safetensors
Model size
1.34B params
Tensor type
F32
Β·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Datasets used to train heegyu/ko-reward-model-safety-1.3b-v0.2

Collection including heegyu/ko-reward-model-safety-1.3b-v0.2