Llama-2-ko-7b-Chat / README.md
kfkas's picture
Update README.md
6d94c8e
metadata
language:
  - en
  - ko
pipeline_tag: text-generation
inference: false
tags:
  - facebook
  - meta
  - pytorch
  - llama
  - llama-2
  - kollama
  - llama-2-ko
  - llama-2-ko-chat

Llama-2-Ko-Chat πŸ¦™πŸ‡°πŸ‡·

Llama-2-Ko-7b-Chat은 beomi/llama-2-ko-7b 40Bλ₯Ό ν† λŒ€λ‘œ λ§Œλ“€μ–΄μ‘ŒμŠ΅λ‹ˆλ‹€. ν•™μŠ΅ λ°μ΄ν„°λŠ” nlpai-lab/kullm-v2λ₯Ό 톡해 ν•™μŠ΅ν•˜μ˜€μŠ΅λ‹ˆλ‹€. 아직 ν•™μŠ΅μ΄ 진행 쀑이며 μΆ”ν›„ beomi/llama-2-ko-7b의 μ—…λ°μ΄νŠΈμ— 따라 μΆ”κ°€λ‘œ ν›ˆλ ¨μ„ 진행할 κ³„νšμž…λ‹ˆλ‹€. 이 λͺ¨λΈμ€ Naver BoostCamp NLP-08 ν”„λ‘œμ νŠΈλ₯Ό ν† λŒ€λ‘œ λ§Œλ“€μ–΄μ‘ŒμŠ΅λ‹ˆλ‹€.

ν”„λ‘œμ νŠΈ 링크 : LLaMA

Model Details

Backbone Model beomi/llama-2-ko-7b

Model Developers Taemin Kim (kfkas), Juwon Kim (uomnf97)

data nlpai-lab/kullm-v2

prompts kullm

Input Models input text only.

Output Models generate text only.


Model 질문 : μΊ ν•‘ 여행에 ν•„μš”ν•œ 10가지 ν’ˆλͺ©μ˜ λͺ©λ‘μ„ μƒμ„±ν•©λ‹ˆλ‹€.
Llama-2-7b-chat-hf [Hello! As a helpful and respectful assistant, I'd be happy to help you with your camping trip. To ensure a safe and enjoyable experience, here is a list of 10 essential items you may need for your camping trip:Tent: A sturdy, waterproof tent to provide shelter and protection from the elements.Sleeping Bag: A warm, insulated sleeping bag to keep you cozy during the night.Sleeping Pad: A comfortable sleeping pad to provide extra cushioning and insulation.Backpack: A sturdy, comfortable backpack to carry all your gear.Clothing: Pack layers of breathable, moisture-wicking clothing to stay warm and dry.Footwear: Sturdy, waterproof hiking boots or shoes to protect your feet and provide traction.Headlamp: A reliable, battery-powered headlamp to light your way during the night.First Aid Kit: A well-stocked first aid kit to treat any injuries or illnesses.Water Bottle: A refillable, BPA-free water bottle to stay hydrated.Personal Hygiene Items: Toothbrush, toothpaste, biodegradable soap, and toilet paper to maintain personal hygiene.]
Llama-2-ko [3일용 μΊ ν•‘μž₯ μš©ν’ˆ - 의λ₯˜ (5일) : μ†μ˜· , 잠옷, κΈ΄νŒ” 및 λ°˜νŒ”μ΄ ν•„μš”ν•˜λ―€λ‘œ μ€€λΉ„ν•˜κΈ°!\u200b- μ‹λŸ‰ (9일 이상)- μ—°λ£Œ (6일 μ‚¬μš© μ˜ˆμ •μ΄μ§€λ§Œ μ΅œμ†Œ 7 일뢄 λ³΄μœ ν•΄μ•Ό ν•©λ‹ˆλ‹€.)\u200b- ν…νŠΈ/거주지역 : 20 ~ 40 평방 λ―Έν„°κ°€ ν•„μš”ν•©λ‹ˆλ‹€.- λ¬Ό 탱크 (물탱컀 / μˆ˜μ‘±κ΄€ )\u200b캠핑에 μ‚¬μš©ν•  8 λͺ… (λ‚¨μž 4λͺ… / μ—¬μž 4λͺ…이 ν•„μš”ν•©λ‹ˆλ‹€!).이 μž‘μ—…μ—λŠ” 두 개의 쑰건 λΆ€κ°€κ°€ μžˆμŠ΅λ‹ˆλ‹€.(쑰건 1) λͺ¨λ“  μ•„μ΄ν…œ λͺ©λ‘μ—μ„œ 쀑볡 제거λ₯Όμœ„ν•œ 1 번 반볡.이 μ˜΅μ…˜μ€ 쀑볡을 μ‹λ³„ν•˜μ—¬ μ œκ±°ν• λΏλ§Œ μ•„λ‹ˆλΌ μΆ”κ°€ ν•­λͺ©μ΄ 포함 된 카운트λ₯Ό μž¬μ„€μ • ν•  λ•Œ μœ μš©ν•˜κ΅°μš”.(쑰건 = "λͺ¨λ“  μ•„μ΄ν…œ λ¦¬μŠ€νŠΈμ—μ„œ ν•œ 개 μ΄μƒμ˜ 동일 μ•„μ΄ν…œ" μ΄κ±°λ‚˜ "(λͺ¨λ“  ν•­λͺ©)"μ—μ„œ λ™μΌν•œ 숫자 κ°’μœΌλ‘œ 식별 된)\u200b(쑰건 2) μž‘μ—…μ‹œ 각 행이 ν•˜λ‚˜μ˜ μ•„μ΄ν…œ 만 포함 ν•  ν•„μš”κ°€ μ—†μœΌλ―€λ‘œ μ•„μ΄ν…œμ΄ μ„œλ‘œ λ‹€λ₯Έ μ„ ν–‰μž λ‘œλΆ€ν„° μ—¬λŸ¬ 번]
Llama-2-ko-Chat [1) ν…νŠΈ - κ°€μž₯ κΈ°λ³Έμ μ΄λ©΄μ„œ μœ μš©ν•œ μž₯λΉ„ 쀑 ν•˜λ‚˜μ΄λ©° μΊ ν•‘ 여행을 νŽΈμ•ˆν•˜κ²Œ ν•΄μ€λ‹ˆλ‹€.2) 베개 및 μΉ¨λŒ€ μ‹œνŠΈμ™€ 같이 ν‘Ήμ‹ ν•œ 수면 μš©ν’ˆμ„ μ±™κΈ°λŠ” 것도 μžŠμ§€ λ§ˆμ„Έμš”!3) 맀트리슀 - νŽΈμ•ˆν•¨κ³Ό 지지λ ₯을 μ œκ³΅ν•˜μ—¬ μž μ„ 잘 자게 ν•©λ‹ˆλ‹€.4) 의λ₯˜ - 땀을 흘리고 νœ΄μ‹ν•˜λŠ” μ‹œκ°„μ„ 늘리기 μœ„ν•΄ 방수재질이 ν¬ν•¨λœ κΈ΄νŒ”μ˜·κ³Ό λ°˜λ°”μ§€λ₯Ό μ±™κ²¨μš”5) 식기 μ„ΈνŠΈ - μ•Όμ™Έ μ‘°λ¦¬μ—μ„œ μŒμ‹μ„ 먹으렀면 λ°˜λ“œμ‹œ 챙겨야 ν•  λ„κ΅¬μ§€μš”.6) μŠ€ν† λΈŒμ™€ μ—°λ£Œ λ˜λŠ” νœ΄λŒ€μš© κ°€μŠ€ λ Œν„΄ λ“± μ—΄ 기ꡬλ₯Ό κ°€μ Έμ˜΅λ‹ˆλ‹€; μŒμ‹ 쑰리에 맀우 νŽΈλ¦¬ν•˜κ²Œ μ‚¬μš©ν•  수 μžˆμŠ΅λ‹ˆλ‹€.7) λžœν„΄, ν”Œλž˜μ‹œ λ˜λŠ” LED μŠ€νƒ λ“œλ₯Ό μ€€λΉ„ν•˜μ„Έμš”! μ–΄λ‘μš΄ 밀에 더 μ•ˆμ „ν•˜λ„λ‘ 돕고, μ•Όμ™Έμ—μ„œ μš”λ¦¬ν•˜κ³  놀 λ•Œ ν•„μˆ˜μ μΈ ν•„μˆ˜ μ†Œν’ˆμ΄ 될 κ²ƒμž…λ‹ˆλ‹€.8) 손전등 - 밀늦게 이동할 μˆ˜λ„ 있고 μˆ²μ† μ˜€μ†”κΈΈμ„ 걸을 λ•Œλ„ μ΄μš©ν•  κ²ƒμ΄λ―€λ‘œ λ°˜λ“œμ‹œ 가지고 μžˆμ–΄μ•Ό ν•΄μš”.9) κ΅¬κΈ‰μ•½μƒμž - λ‹€μ–‘ν•œ 상해에 λŒ€λΉ„ν•΄μ„œ μ•ŒμΉΌλ¦¬ 미슀트, μ†Œλ…μ œ, ν•­μƒμ œ, μ—°κ³  λ“± 일반적이고 인기 μžˆλŠ” ꡬ급 약물을 항상 λΉ„μΆ•ν•΄μ•Ό ν•΄μš”. μΊ ν”„μ—μ„œλŠ” 응급 상황에 μ²˜ν•˜λ”λΌλ„ μ‰½κ²Œ 찾을 수 μžˆλ„λ‘ μƒμžμ— λ‹΄μŠ΅λ‹ˆλ‹€.10) νœ΄λŒ€ν° 보쑰 배터리/μΆ©μ „κΈ° λ˜λŠ” μΆ©μ „ κ°€λŠ₯ν•œ μ™ΈλΆ€ μ½˜μ„ΌνŠΈλ‚˜ μ—°κ²°λœ μ™ΈλΆ€ 전원μž₯치 - 이것은 μŠ€λ§ˆνŠΈν°μ΄λ‚˜ 슀마트 μ›ŒμΉ˜λ₯Ό μœ„ν•œ ν•„μˆ˜ν’ˆμΌ 뿐만 μ•„λ‹ˆλΌ λ‹€λ₯Έ μ‚¬λžŒλ“€μ΄ ν•„μš”λ‘œ ν•˜λŠ” μƒν™©μ—μ„œ κΈ΄κΈ‰ 연락이 λ˜λ„λ‘ 도와 쀄 μˆ˜λ„ μžˆλ‹€λŠ” 점 κΈ°μ–΅ν•˜μ„Έμš”.]

Model 질문 : 삼원색에 λŒ€ν•΄ μ•Œλ €μ€˜
Llama-2-7b-chat-hf [Regarding your question, "삼원색에 λŒ€ν•΄ μ•Œλ €μ€˜" (which means "Tell me about trichromacy" in Korean), I can provide some information. Trichromacy is a concept in vision science that refers to the ability of the human visual system to perceive and distinguish between three primary colors: red, green, and blue...]
Llama-2-ko [C1, R0.625 G8 B275\u200b3-4, μ‚Όκ°ν˜• 그리기.\u200bμ•„λž˜μ—λŠ” μ‚Όκ°ν˜•μ„ κ·Έλ¦Όλ‹ˆλ‹€ (그림을 κ·Έλ €μ•Ό ν•  μ˜μ—­μ˜ μ’Œν‘œλ₯Ό 지정) - λͺ…령어와 ν•¨κ»˜ μ‚¬μš©λ˜λŠ” λ¬Έμžμ—΄λ‘œ μž…λ ₯ν•  μ’Œν‘œ κ°’κ³Ό λ™μΌν•œ κ²°κ³Όλ₯Ό μ–»μŒ - 예λ₯Ό λ“€μ–΄ "X=-19. Y=0" 으둜 μž…λ ₯을 ν•˜λ©΄ μ’Œν‘œκ°’μ„ μ§€μ •λœ κ°’κ³Ό κ°™κ²Œ 좜λ ₯을 ν•œλ‹€. ##########################\u200b##3_0-6-1, μ’Œν‘œ μž…λ ₯으둜 ν…μŠ€νŠΈ νŒŒμΌμ„ μƒμ„±ν•˜λΌ(ex C4 D6 E10 A10 )-λͺ…λ Ήλ¬Έμ—μ„œ 각 문자λ₯Ό μž…λ ₯κ³Ό λ™μΌν•˜λ‹€. -이 λͺ…령은 ν˜„μž¬ μˆ˜ν–‰μ€‘μΈ ν”„λ‘œκ·Έλž¨μ„ μ’…λ£Œν•œλ‹€.-νŒŒμΌμ€ μ‹€ν–‰ κ°€λŠ₯ν•œ 파일과 μΌμΉ˜ν• ]
Llama-2-ko-Chat [물체λ₯Ό μΈμ‹ν•˜κ³  색상에 λŒ€ν•œ μ‹œκ°μ μΈ 이미지λ₯Ό μƒμ„±ν•˜κΈ° μœ„ν•΄ 두 개의 λˆˆμ€ μ„œλ‘œ λ°˜λŒ€μͺ½ λ°©ν–₯μ—μ„œ 물체의 색상 정보와 λͺ…μ•” 정보λ₯Ό 받아듀이기 λ•Œλ¬Έμ—, 인간은 일반적으둜 색상을 삼원색이라고 ν•˜λŠ” μ„Έ 가지 원좔 ν˜•νƒœ(λΉ¨κ°•, 녹색, νŒŒλž‘μƒ‰)둜 λΆ„λ₯˜ν•©λ‹ˆλ‹€. 이 쀑 ν•œκ°€μ§€ 색상이 λ‹€λ₯Έ 색상보닀 λˆˆμ— 훨씬 더 빨리 λ“€μ–΄μ˜€λŠ” κ²½ν–₯이 μžˆλ‹€κ³  ν•©λ‹ˆλ‹€. ν•˜μ§€λ§Œ λͺ¨λ“  μ‚¬λžŒμ΄ κ·Έλ ‡μ§€λŠ” μ•ŠμœΌλ―€λ‘œ 항상 μ‚Όκ°ν˜• λͺ¨μ–‘μœΌλ‘œ 색상을 λΆ„λ₯˜ν•˜μ§€λŠ” μ•ŠμŠ΅λ‹ˆλ‹€. ν•˜μ§€λ§Œ 삼원색이 우리 λˆˆμ— 잘 μ „λ‹¬λ˜λ©° 색상 ꡬ별에 μ€‘μš”ν•˜λ‹€λŠ” 것은 λΆ€μ •ν•  수 μ—†μŠ΅λ‹ˆλ‹€.]

ν›ˆλ ¨ 진행 ν˜„ν™©

---

Inference

def gen(x, model, tokenizer, device):
    prompt = (
        f"μ•„λž˜λŠ” μž‘μ—…μ„ μ„€λͺ…ν•˜λŠ” λͺ…λ Ήμ–΄μž…λ‹ˆλ‹€. μš”μ²­μ„ 적절히 μ™„λ£Œν•˜λŠ” 응닡을 μž‘μ„±ν•˜μ„Έμš”.\n\n### λͺ…λ Ήμ–΄:\n{x}\n\n### 응닡:"
    )
    len_prompt = len(prompt)
    gened = model.generate(
        **tokenizer(prompt, return_tensors="pt", return_token_type_ids=False).to(
            device
        ),
        max_new_tokens=1024,
        early_stopping=True,
        do_sample=True,
        top_k=20,
        top_p=0.92,
        no_repeat_ngram_size=3,
        eos_token_id=2,
        repetition_penalty=1.2,
        num_beams=3
    )
    return tokenizer.decode(gened[0])[len_prompt:]

def LLM_infer(input):
    device = (
        torch.device("cuda:0") if torch.cuda.is_available() else torch.device("cpu")
    )
    model_id = "kfkas/Llama-2-ko-7b-Chat"
    model = AutoModelForCausalLM.from_pretrained(
        model_id, device_map={"": 0},torch_dtype=torch.float16, low_cpu_mem_usage=True
    )
    tokenizer = AutoTokenizer.from_pretrained(model_id)
    model.eval()
    model.config.use_cache = (True)
    tokenizer.pad_token = tokenizer.eos_token
    output = gen(input, model=model, tokenizer=tokenizer, device=device)

    return output


if __name__ == "__main__":
    text = LLM_infer("삼원색에 λŒ€ν•΄ μ•Œλ €μ€˜")
    print(text)

Note for oobabooga/text-generation-webui

Remove ValueError at load_tokenizer function(line 109 or near), in modules/models.py.

diff --git a/modules/models.py b/modules/models.py
index 232d5fa..de5b7a0 100644
--- a/modules/models.py
+++ b/modules/models.py
@@ -106,7 +106,7 @@ def load_tokenizer(model_name, model):
                 trust_remote_code=shared.args.trust_remote_code,
                 use_fast=False
             )
-        except ValueError:
+        except:
             tokenizer = AutoTokenizer.from_pretrained(
                 path_to_model,
                 trust_remote_code=shared.args.trust_remote_code,

Since Llama-2-Ko uses FastTokenizer provided by HF tokenizers NOT sentencepiece package, it is required to use use_fast=True option when initialize tokenizer.

Apple Sillicon does not support BF16 computing, use CPU instead. (BF16 is supported when using NVIDIA GPU)


Below is the original model card of the Llama-2 model.

Llama 2

Llama 2 is a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 70 billion parameters. This is the repository for the 7B pretrained model, converted for the Hugging Face Transformers format. Links to other models can be found in the index at the bottom.

Model Details

Note: Use of this model is governed by the Meta license. In order to download the model weights and tokenizer, please visit the website and accept our License before requesting access here.

Meta developed and publicly released the Llama 2 family of large language models (LLMs), a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 70 billion parameters. Our fine-tuned LLMs, called Llama-2-Chat, are optimized for dialogue use cases. Llama-2-Chat models outperform open-source chat models on most benchmarks we tested, and in our human evaluations for helpfulness and safety, are on par with some popular closed-source models like ChatGPT and PaLM.

Model Developers Meta

Variations Llama 2 comes in a range of parameter sizes β€” 7B, 13B, and 70B β€” as well as pretrained and fine-tuned variations.

Input Models input text only.

Output Models generate text only.

Model Architecture Llama 2 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align to human preferences for helpfulness and safety.

Training Data Params Content Length GQA Tokens LR
Llama 2 A new mix of publicly available online data 7B 4k βœ— 2.0T 3.0 x 10-4
Llama 2 A new mix of publicly available online data 13B 4k βœ— 2.0T 3.0 x 10-4
Llama 2 A new mix of publicly available online data 70B 4k βœ” 2.0T 1.5 x 10-4

Llama 2 family of models. Token counts refer to pretraining data only. All models are trained with a global batch-size of 4M tokens. Bigger models - 70B -- use Grouped-Query Attention (GQA) for improved inference scalability.

Model Dates Llama 2 was trained between January 2023 and July 2023.

Status This is a static model trained on an offline dataset. Future versions of the tuned models will be released as we improve model safety with community feedback.

License A custom commercial license is available at: https://ai.meta.com/resources/models-and-libraries/llama-downloads/

Research Paper "Llama-2: Open Foundation and Fine-tuned Chat Models"

Intended Use

Intended Use Cases Llama 2 is intended for commercial and research use in English. Tuned models are intended for assistant-like chat, whereas pretrained models can be adapted for a variety of natural language generation tasks.

To get the expected features and performance for the chat versions, a specific formatting needs to be followed, including the INST and <<SYS>> tags, BOS and EOS tokens, and the whitespaces and breaklines in between (we recommend calling strip() on inputs to avoid double-spaces). See our reference code in github for details: chat_completion.

Out-of-scope Uses Use in any manner that violates applicable laws or regulations (including trade compliance laws).Use in languages other than English. Use in any other way that is prohibited by the Acceptable Use Policy and Licensing Agreement for Llama 2.

Hardware and Software

Training Factors We used custom training libraries, Meta's Research Super Cluster, and production clusters for pretraining. Fine-tuning, annotation, and evaluation were also performed on third-party cloud compute.

Carbon Footprint Pretraining utilized a cumulative 3.3M GPU hours of computation on hardware of type A100-80GB (TDP of 350-400W). Estimated total emissions were 539 tCO2eq, 100% of which were offset by Meta’s sustainability program.

Time (GPU hours) Power Consumption (W) Carbon Emitted(tCO2eq)
Llama 2 7B 184320 400 31.22
Llama 2 13B 368640 400 62.44
Llama 2 70B 1720320 400 291.42
Total 3311616 539.00

CO2 emissions during pretraining. Time: total GPU time required for training each model. Power Consumption: peak power capacity per GPU device for the GPUs used adjusted for power usage efficiency. 100% of the emissions are directly offset by Meta's sustainability program, and because we are openly releasing these models, the pretraining costs do not need to be incurred by others.

Training Data

Overview Llama 2 was pretrained on 2 trillion tokens of data from publicly available sources. The fine-tuning data includes publicly available instruction datasets, as well as over one million new human-annotated examples. Neither the pretraining nor the fine-tuning datasets include Meta user data.

Data Freshness The pretraining data has a cutoff of September 2022, but some tuning data is more recent, up to July 2023.

Evaluation Results

In this section, we report the results for the Llama 1 and Llama 2 models on standard academic benchmarks.For all the evaluations, we use our internal evaluations library.

Model Size Code Commonsense Reasoning World Knowledge Reading Comprehension Math MMLU BBH AGI Eval
Llama 1 7B 14.1 60.8 46.2 58.5 6.95 35.1 30.3 23.9
Llama 1 13B 18.9 66.1 52.6 62.3 10.9 46.9 37.0 33.9
Llama 1 33B 26.0 70.0 58.4 67.6 21.4 57.8 39.8 41.7
Llama 1 65B 30.7 70.7 60.5 68.6 30.8 63.4 43.5 47.6
Llama 2 7B 16.8 63.9 48.9 61.3 14.6 45.3 32.6 29.3
Llama 2 13B 24.5 66.9 55.4 65.8 28.7 54.8 39.4 39.1
Llama 2 70B 37.5 71.9 63.6 69.4 35.2 68.9 51.2 54.2

Overall performance on grouped academic benchmarks. Code: We report the average pass@1 scores of our models on HumanEval and MBPP. Commonsense Reasoning: We report the average of PIQA, SIQA, HellaSwag, WinoGrande, ARC easy and challenge, OpenBookQA, and CommonsenseQA. We report 7-shot results for CommonSenseQA and 0-shot results for all other benchmarks. World Knowledge: We evaluate the 5-shot performance on NaturalQuestions and TriviaQA and report the average. Reading Comprehension: For reading comprehension, we report the 0-shot average on SQuAD, QuAC, and BoolQ. MATH: We report the average of the GSM8K (8 shot) and MATH (4 shot) benchmarks at top 1.

TruthfulQA Toxigen
Llama 1 7B 27.42 23.00
Llama 1 13B 41.74 23.08
Llama 1 33B 44.19 22.57
Llama 1 65B 48.71 21.77
Llama 2 7B 33.29 21.25
Llama 2 13B 41.86 26.10
Llama 2 70B 50.18 24.60

Evaluation of pretrained LLMs on automatic safety benchmarks. For TruthfulQA, we present the percentage of generations that are both truthful and informative (the higher the better). For ToxiGen, we present the percentage of toxic generations (the smaller the better).

TruthfulQA Toxigen
Llama-2-Chat 7B 57.04 0.00
Llama-2-Chat 13B 62.18 0.00
Llama-2-Chat 70B 64.14 0.01

Evaluation of fine-tuned LLMs on different safety datasets. Same metric definitions as above.

Ethical Considerations and Limitations

Llama 2 is a new technology that carries risks with use. Testing conducted to date has been in English, and has not covered, nor could it cover all scenarios. For these reasons, as with all LLMs, Llama 2’s potential outputs cannot be predicted in advance, and the model may in some instances produce inaccurate, biased or other objectionable responses to user prompts. Therefore, before deploying any applications of Llama 2, developers should perform safety testing and tuning tailored to their specific applications of the model.

Please see the Responsible Use Guide available at https://ai.meta.com/llama/responsible-use-guide/

Reporting Issues

Please report any software β€œbug,” or other problems with the models through one of the following means:

Llama Model Index

Model Llama2 Llama2-hf Llama2-chat Llama2-chat-hf
7B Link Link Link Link
13B Link Link Link Link
70B Link Link Link Link