Text Generation
Transformers
Safetensors
Korean
llama
korean
causal-lm
instruction-tuned
from-scratch
kawk
conversational
text-generation-inference
Instructions to use Infinity08/KAWK-500M-Korean-Instruct-v1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Infinity08/KAWK-500M-Korean-Instruct-v1 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="Infinity08/KAWK-500M-Korean-Instruct-v1") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("Infinity08/KAWK-500M-Korean-Instruct-v1") model = AutoModelForCausalLM.from_pretrained("Infinity08/KAWK-500M-Korean-Instruct-v1", device_map="auto") messages = [ {"role": "user", "content": "Who are you?"}, ] inputs = tokenizer.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", ).to(model.device) outputs = model.generate(**inputs, max_new_tokens=40) print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use Infinity08/KAWK-500M-Korean-Instruct-v1 with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "Infinity08/KAWK-500M-Korean-Instruct-v1" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Infinity08/KAWK-500M-Korean-Instruct-v1", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/Infinity08/KAWK-500M-Korean-Instruct-v1
- SGLang
How to use Infinity08/KAWK-500M-Korean-Instruct-v1 with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "Infinity08/KAWK-500M-Korean-Instruct-v1" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Infinity08/KAWK-500M-Korean-Instruct-v1", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "Infinity08/KAWK-500M-Korean-Instruct-v1" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Infinity08/KAWK-500M-Korean-Instruct-v1", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use Infinity08/KAWK-500M-Korean-Instruct-v1 with Docker Model Runner:
docker model run hf.co/Infinity08/KAWK-500M-Korean-Instruct-v1
File size: 10,822 Bytes
9a1b839 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 | model,section,task,metric,value,stderr
kawk500m-instruct-v1,groups,kobest,sample_len,4561.0,
kawk500m-instruct-v1,groups,kobest,"acc,none",0.5231308923481692,0.007297298748221897
kawk500m-instruct-v1,groups,kobest,"acc_norm,none",0.474,0.022352791650914174
kawk500m-instruct-v1,groups,kobest,"f1,none",0.46754121249081004,
kawk500m-instruct-v1,groups,kmmlu_stem,sample_len,9900.0,
kawk500m-instruct-v1,groups,kmmlu_stem,"acc,none",0.13797979797979798,0.003422447552202874
kawk500m-instruct-v1,groups,kmmlu_other,sample_len,8400.0,
kawk500m-instruct-v1,groups,kmmlu_other,"acc,none",0.14952380952380953,0.0038749065104715114
kawk500m-instruct-v1,groups,kmmlu_applied_science,sample_len,11600.0,
kawk500m-instruct-v1,groups,kmmlu_applied_science,"acc,none",0.12637931034482758,0.0030704612706282304
kawk500m-instruct-v1,groups,kmmlu_humss,sample_len,5130.0,
kawk500m-instruct-v1,groups,kmmlu_humss,"acc,none",0.2155945419103314,0.005734774245353025
kawk500m-instruct-v1,groups,kmmlu,sample_len,35030.0,
kawk500m-instruct-v1,groups,kmmlu,"acc,none",0.1482729089351984,0.001880917345299668
kawk500m-instruct-v1,results,kobest_boolq,sample_len,1404.0,
kawk500m-instruct-v1,results,kobest_boolq,"acc,none",0.5135327635327636,0.013343877331261893
kawk500m-instruct-v1,results,kobest_boolq,"f1,none",0.4553178437236408,
kawk500m-instruct-v1,results,kobest_copa,sample_len,1000.0,
kawk500m-instruct-v1,results,kobest_copa,"acc,none",0.635,0.015231776226264848
kawk500m-instruct-v1,results,kobest_copa,"f1,none",0.6335455422572189,
kawk500m-instruct-v1,results,kobest_hellaswag,sample_len,500.0,
kawk500m-instruct-v1,results,kobest_hellaswag,"acc,none",0.354,0.02140758204791649
kawk500m-instruct-v1,results,kobest_hellaswag,"f1,none",0.3513485918054787,
kawk500m-instruct-v1,results,kobest_hellaswag,"acc_norm,none",0.474,0.022352791650914174
kawk500m-instruct-v1,results,kobest_sentineg,sample_len,397.0,
kawk500m-instruct-v1,results,kobest_sentineg,"acc,none",0.6045340050377834,0.024570689701367988
kawk500m-instruct-v1,results,kobest_sentineg,"f1,none",0.592580971990718,
kawk500m-instruct-v1,results,kobest_wic,sample_len,1260.0,
kawk500m-instruct-v1,results,kobest_wic,"acc,none",0.4865079365079365,0.014086365971849238
kawk500m-instruct-v1,results,kobest_wic,"f1,none",0.3561228043986665,
kawk500m-instruct-v1,results,kmmlu_biology,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_biology,"acc,none",0.204,0.012749374359024311
kawk500m-instruct-v1,results,kmmlu_chemical_engineering,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_chemical_engineering,"acc,none",0.215,0.01299784381903183
kawk500m-instruct-v1,results,kmmlu_chemistry,sample_len,600.0,
kawk500m-instruct-v1,results,kmmlu_chemistry,"acc,none",0.21666666666666667,0.016832783728500004
kawk500m-instruct-v1,results,kmmlu_civil_engineering,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_civil_engineering,"acc,none",0.084,0.008776162089491113
kawk500m-instruct-v1,results,kmmlu_computer_science,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_computer_science,"acc,none",0.086,0.008870325962594761
kawk500m-instruct-v1,results,kmmlu_ecology,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_ecology,"acc,none",0.122,0.010354864712936776
kawk500m-instruct-v1,results,kmmlu_electrical_engineering,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_electrical_engineering,"acc,none",0.082,0.008680515615523732
kawk500m-instruct-v1,results,kmmlu_information_technology,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_information_technology,"acc,none",0.104,0.009658016218524242
kawk500m-instruct-v1,results,kmmlu_materials_engineering,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_materials_engineering,"acc,none",0.157,0.011510146979230262
kawk500m-instruct-v1,results,kmmlu_math,sample_len,300.0,
kawk500m-instruct-v1,results,kmmlu_math,"acc,none",0.26666666666666666,0.02557404853322572
kawk500m-instruct-v1,results,kmmlu_mechanical_engineering,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_mechanical_engineering,"acc,none",0.102,0.009575368801653944
kawk500m-instruct-v1,results,kmmlu_agricultural_sciences,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_agricultural_sciences,"acc,none",0.174,0.011994493230973449
kawk500m-instruct-v1,results,kmmlu_construction,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_construction,"acc,none",0.111,0.009938701010583716
kawk500m-instruct-v1,results,kmmlu_fashion,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_fashion,"acc,none",0.186,0.012310790208412926
kawk500m-instruct-v1,results,kmmlu_food_processing,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_food_processing,"acc,none",0.176,0.012048616898597498
kawk500m-instruct-v1,results,kmmlu_health,sample_len,100.0,
kawk500m-instruct-v1,results,kmmlu_health,"acc,none",0.25,0.04351941398892446
kawk500m-instruct-v1,results,kmmlu_interior_architecture_and_design,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_interior_architecture_and_design,"acc,none",0.111,0.009938701010583716
kawk500m-instruct-v1,results,kmmlu_marketing,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_marketing,"acc,none",0.131,0.0106748748448379
kawk500m-instruct-v1,results,kmmlu_patent,sample_len,100.0,
kawk500m-instruct-v1,results,kmmlu_patent,"acc,none",0.25,0.04351941398892446
kawk500m-instruct-v1,results,kmmlu_public_safety,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_public_safety,"acc,none",0.116,0.010131468138756924
kawk500m-instruct-v1,results,kmmlu_real_estate,sample_len,200.0,
kawk500m-instruct-v1,results,kmmlu_real_estate,"acc,none",0.22,0.02936514188266327
kawk500m-instruct-v1,results,kmmlu_refrigerating_machinery,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_refrigerating_machinery,"acc,none",0.157,0.011510146979230262
kawk500m-instruct-v1,results,kmmlu_aviation_engineering_and_maintenance,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_aviation_engineering_and_maintenance,"acc,none",0.144,0.01110798754893916
kawk500m-instruct-v1,results,kmmlu_electronics_engineering,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_electronics_engineering,"acc,none",0.092,0.009144376393151129
kawk500m-instruct-v1,results,kmmlu_energy_management,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_energy_management,"acc,none",0.207,0.012818553557844009
kawk500m-instruct-v1,results,kmmlu_environmental_science,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_environmental_science,"acc,none",0.101,0.009533618929341046
kawk500m-instruct-v1,results,kmmlu_gas_technology_and_engineering,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_gas_technology_and_engineering,"acc,none",0.118,0.010206869264381718
kawk500m-instruct-v1,results,kmmlu_geomatics,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_geomatics,"acc,none",0.125,0.010463483381956722
kawk500m-instruct-v1,results,kmmlu_industrial_engineer,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_industrial_engineer,"acc,none",0.071,0.008125578442487959
kawk500m-instruct-v1,results,kmmlu_machine_design_and_manufacturing,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_machine_design_and_manufacturing,"acc,none",0.123,0.010391293421849803
kawk500m-instruct-v1,results,kmmlu_maritime_engineering,sample_len,600.0,
kawk500m-instruct-v1,results,kmmlu_maritime_engineering,"acc,none",0.15666666666666668,0.014851643766757317
kawk500m-instruct-v1,results,kmmlu_nondestructive_testing,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_nondestructive_testing,"acc,none",0.15,0.011297239823409416
kawk500m-instruct-v1,results,kmmlu_railway_and_automotive_engineering,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_railway_and_automotive_engineering,"acc,none",0.136,0.01084535023047304
kawk500m-instruct-v1,results,kmmlu_telecommunications_and_wireless_technology,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_telecommunications_and_wireless_technology,"acc,none",0.105,0.00969892102602496
kawk500m-instruct-v1,results,kmmlu_accounting,sample_len,100.0,
kawk500m-instruct-v1,results,kmmlu_accounting,"acc,none",0.24,0.04292346959909278
kawk500m-instruct-v1,results,kmmlu_criminal_law,sample_len,200.0,
kawk500m-instruct-v1,results,kmmlu_criminal_law,"acc,none",0.195,0.028085923439997246
kawk500m-instruct-v1,results,kmmlu_economics,sample_len,130.0,
kawk500m-instruct-v1,results,kmmlu_economics,"acc,none",0.27692307692307694,0.03939825345266472
kawk500m-instruct-v1,results,kmmlu_education,sample_len,100.0,
kawk500m-instruct-v1,results,kmmlu_education,"acc,none",0.28,0.045126085985421296
kawk500m-instruct-v1,results,kmmlu_korean_history,sample_len,100.0,
kawk500m-instruct-v1,results,kmmlu_korean_history,"acc,none",0.23,0.04229525846816507
kawk500m-instruct-v1,results,kmmlu_law,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_law,"acc,none",0.24,0.013512312258920847
kawk500m-instruct-v1,results,kmmlu_management,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_management,"acc,none",0.204,0.012749374359024311
kawk500m-instruct-v1,results,kmmlu_political_science_and_sociology,sample_len,300.0,
kawk500m-instruct-v1,results,kmmlu_political_science_and_sociology,"acc,none",0.23,0.024337372337779037
kawk500m-instruct-v1,results,kmmlu_psychology,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_psychology,"acc,none",0.224,0.013190830072364589
kawk500m-instruct-v1,results,kmmlu_social_welfare,sample_len,1000.0,
kawk500m-instruct-v1,results,kmmlu_social_welfare,"acc,none",0.172,0.011939788882495308
kawk500m-instruct-v1,results,kmmlu_taxation,sample_len,200.0,
kawk500m-instruct-v1,results,kmmlu_taxation,"acc,none",0.235,0.03005647949775547
kawk500m-instruct-v1,results,kobest,sample_len,4561.0,
kawk500m-instruct-v1,results,kobest,"acc,none",0.5231308923481692,0.007297298748221897
kawk500m-instruct-v1,results,kobest,"acc_norm,none",0.474,0.022352791650914174
kawk500m-instruct-v1,results,kobest,"f1,none",0.46754121249081004,
kawk500m-instruct-v1,results,kmmlu_stem,sample_len,9900.0,
kawk500m-instruct-v1,results,kmmlu_stem,"acc,none",0.13797979797979798,0.003422447552202874
kawk500m-instruct-v1,results,kmmlu_other,sample_len,8400.0,
kawk500m-instruct-v1,results,kmmlu_other,"acc,none",0.14952380952380953,0.0038749065104715114
kawk500m-instruct-v1,results,kmmlu_applied_science,sample_len,11600.0,
kawk500m-instruct-v1,results,kmmlu_applied_science,"acc,none",0.12637931034482758,0.0030704612706282304
kawk500m-instruct-v1,results,kmmlu_humss,sample_len,5130.0,
kawk500m-instruct-v1,results,kmmlu_humss,"acc,none",0.2155945419103314,0.005734774245353025
kawk500m-instruct-v1,results,kmmlu,sample_len,35030.0,
kawk500m-instruct-v1,results,kmmlu,"acc,none",0.1482729089351984,0.001880917345299668
|