metadata
tags:
- merge
- mergekit
- lazymergekit
- Locutusque/Hercules-3.1-Mistral-7B
- cognitivecomputations/dolphin-2.8-experiment26-7b
base_model:
- Locutusque/Hercules-3.1-Mistral-7B
- cognitivecomputations/dolphin-2.8-experiment26-7b
license: apache-2.0
JOSIE_Beta-3-7B-slerp
JOSIE_Beta-3-7B-slerp is a merge of the following models using LazyMergekit:
IMPORTANT!!!
upon sseing the eval bechmarks on the LLM Leaderboard, this is the best performing model, but it's not uncensored, and it's answers are not really good when chatting with it. I will further train it one datasets like dolphin and other.
{
"all": {
"acc": 0.6432209013684985,
"acc_stderr": 0.03221665824377992,
"acc_norm": 0.6450099678239628,
"acc_norm_stderr": 0.032867717920871294,
"mc1": 0.3353733170134639,
"mc1_stderr": 0.01652753403966899,
"mc2": 0.48804542326643174,
"mc2_stderr": 0.015087630632446147
},
"harness|arc:challenge|25": {
"acc": 0.6083617747440273,
"acc_stderr": 0.014264122124938217,
"acc_norm": 0.6339590443686007,
"acc_norm_stderr": 0.014077223108470139
},
"harness|hellaswag|10": {
"acc": 0.6618203545110536,
"acc_stderr": 0.0047212316370927225,
"acc_norm": 0.8456482772356104,
"acc_norm_stderr": 0.0036054721167622867
},
"harness|hendrycksTest-abstract_algebra|5": {
"acc": 0.3,
"acc_stderr": 0.046056618647183814,
"acc_norm": 0.3,
"acc_norm_stderr": 0.046056618647183814
},
"harness|hendrycksTest-anatomy|5": {
"acc": 0.6074074074074074,
"acc_stderr": 0.04218506215368879,
"acc_norm": 0.6074074074074074,
"acc_norm_stderr": 0.04218506215368879
},
"harness|hendrycksTest-astronomy|5": {
"acc": 0.6710526315789473,
"acc_stderr": 0.03823428969926605,
"acc_norm": 0.6710526315789473,
"acc_norm_stderr": 0.03823428969926605
},
"harness|hendrycksTest-business_ethics|5": {
"acc": 0.61,
"acc_stderr": 0.04902071300001975,
"acc_norm": 0.61,
"acc_norm_stderr": 0.04902071300001975
},
"harness|hendrycksTest-clinical_knowledge|5": {
"acc": 0.7018867924528301,
"acc_stderr": 0.02815283794249387,
"acc_norm": 0.7018867924528301,
"acc_norm_stderr": 0.02815283794249387
},
"harness|hendrycksTest-college_biology|5": {
"acc": 0.7638888888888888,
"acc_stderr": 0.03551446610810826,
"acc_norm": 0.7638888888888888,
"acc_norm_stderr": 0.03551446610810826
},
"harness|hendrycksTest-college_chemistry|5": {
"acc": 0.44,
"acc_stderr": 0.04988876515698589,
"acc_norm": 0.44,
"acc_norm_stderr": 0.04988876515698589
},
"harness|hendrycksTest-college_computer_science|5": {
"acc": 0.52,
"acc_stderr": 0.050211673156867795,
"acc_norm": 0.52,
"acc_norm_stderr": 0.050211673156867795
},
"harness|hendrycksTest-college_mathematics|5": {
"acc": 0.37,
"acc_stderr": 0.04852365870939099,
"acc_norm": 0.37,
"acc_norm_stderr": 0.04852365870939099
},
"harness|hendrycksTest-college_medicine|5": {
"acc": 0.6473988439306358,
"acc_stderr": 0.03643037168958548,
"acc_norm": 0.6473988439306358,
"acc_norm_stderr": 0.03643037168958548
},
"harness|hendrycksTest-college_physics|5": {
"acc": 0.38235294117647056,
"acc_stderr": 0.04835503696107223,
"acc_norm": 0.38235294117647056,
"acc_norm_stderr": 0.04835503696107223
},
"harness|hendrycksTest-computer_security|5": {
"acc": 0.75,
"acc_stderr": 0.04351941398892446,
"acc_norm": 0.75,
"acc_norm_stderr": 0.04351941398892446
},
"harness|hendrycksTest-conceptual_physics|5": {
"acc": 0.574468085106383,
"acc_stderr": 0.03232146916224468,
"acc_norm": 0.574468085106383,
"acc_norm_stderr": 0.03232146916224468
},
"harness|hendrycksTest-econometrics|5": {
"acc": 0.4649122807017544,
"acc_stderr": 0.046920083813689104,
"acc_norm": 0.4649122807017544,
"acc_norm_stderr": 0.046920083813689104
},
"harness|hendrycksTest-electrical_engineering|5": {
"acc": 0.5310344827586206,
"acc_stderr": 0.04158632762097828,
"acc_norm": 0.5310344827586206,
"acc_norm_stderr": 0.04158632762097828
},
"harness|hendrycksTest-elementary_mathematics|5": {
"acc": 0.41534391534391535,
"acc_stderr": 0.025379524910778405,
"acc_norm": 0.41534391534391535,
"acc_norm_stderr": 0.025379524910778405
},
"harness|hendrycksTest-formal_logic|5": {
"acc": 0.46825396825396826,
"acc_stderr": 0.04463112720677171,
"acc_norm": 0.46825396825396826,
"acc_norm_stderr": 0.04463112720677171
},
"harness|hendrycksTest-global_facts|5": {
"acc": 0.35,
"acc_stderr": 0.0479372485441102,
"acc_norm": 0.35,
"acc_norm_stderr": 0.0479372485441102
},
"harness|hendrycksTest-high_school_biology|5": {
"acc": 0.7709677419354839,
"acc_stderr": 0.023904914311782648,
"acc_norm": 0.7709677419354839,
"acc_norm_stderr": 0.023904914311782648
},
"harness|hendrycksTest-high_school_chemistry|5": {
"acc": 0.5073891625615764,
"acc_stderr": 0.035176035403610105,
"acc_norm": 0.5073891625615764,
"acc_norm_stderr": 0.035176035403610105
},
"harness|hendrycksTest-high_school_computer_science|5": {
"acc": 0.69,
"acc_stderr": 0.04648231987117316,
"acc_norm": 0.69,
"acc_norm_stderr": 0.04648231987117316
},
"harness|hendrycksTest-high_school_european_history|5": {
"acc": 0.7757575757575758,
"acc_stderr": 0.032568666616811015,
"acc_norm": 0.7757575757575758,
"acc_norm_stderr": 0.032568666616811015
},
"harness|hendrycksTest-high_school_geography|5": {
"acc": 0.797979797979798,
"acc_stderr": 0.028606204289229872,
"acc_norm": 0.797979797979798,
"acc_norm_stderr": 0.028606204289229872
},
"harness|hendrycksTest-high_school_government_and_politics|5": {
"acc": 0.8963730569948186,
"acc_stderr": 0.02199531196364424,
"acc_norm": 0.8963730569948186,
"acc_norm_stderr": 0.02199531196364424
},
"harness|hendrycksTest-high_school_macroeconomics|5": {
"acc": 0.6641025641025641,
"acc_stderr": 0.023946724741563973,
"acc_norm": 0.6641025641025641,
"acc_norm_stderr": 0.023946724741563973
},
"harness|hendrycksTest-high_school_mathematics|5": {
"acc": 0.3888888888888889,
"acc_stderr": 0.029723278961476664,
"acc_norm": 0.3888888888888889,
"acc_norm_stderr": 0.029723278961476664
},
"harness|hendrycksTest-high_school_microeconomics|5": {
"acc": 0.680672268907563,
"acc_stderr": 0.030283995525884396,
"acc_norm": 0.680672268907563,
"acc_norm_stderr": 0.030283995525884396
},
"harness|hendrycksTest-high_school_physics|5": {
"acc": 0.3443708609271523,
"acc_stderr": 0.038796870240733264,
"acc_norm": 0.3443708609271523,
"acc_norm_stderr": 0.038796870240733264
},
"harness|hendrycksTest-high_school_psychology|5": {
"acc": 0.8422018348623853,
"acc_stderr": 0.01563002297009245,
"acc_norm": 0.8422018348623853,
"acc_norm_stderr": 0.01563002297009245
},
"harness|hendrycksTest-high_school_statistics|5": {
"acc": 0.5,
"acc_stderr": 0.034099716973523674,
"acc_norm": 0.5,
"acc_norm_stderr": 0.034099716973523674
},
"harness|hendrycksTest-high_school_us_history|5": {
"acc": 0.7990196078431373,
"acc_stderr": 0.028125972265654366,
"acc_norm": 0.7990196078431373,
"acc_norm_stderr": 0.028125972265654366
},
"harness|hendrycksTest-high_school_world_history|5": {
"acc": 0.7890295358649789,
"acc_stderr": 0.02655837250266192,
"acc_norm": 0.7890295358649789,
"acc_norm_stderr": 0.02655837250266192
},
"harness|hendrycksTest-human_aging|5": {
"acc": 0.695067264573991,
"acc_stderr": 0.030898610882477515,
"acc_norm": 0.695067264573991,
"acc_norm_stderr": 0.030898610882477515
},
"harness|hendrycksTest-human_sexuality|5": {
"acc": 0.7862595419847328,
"acc_stderr": 0.0359546161177469,
"acc_norm": 0.7862595419847328,
"acc_norm_stderr": 0.0359546161177469
},
"harness|hendrycksTest-international_law|5": {
"acc": 0.8016528925619835,
"acc_stderr": 0.036401182719909476,
"acc_norm": 0.8016528925619835,
"acc_norm_stderr": 0.036401182719909476
},
"harness|hendrycksTest-jurisprudence|5": {
"acc": 0.7870370370370371,
"acc_stderr": 0.0395783547198098,
"acc_norm": 0.7870370370370371,
"acc_norm_stderr": 0.0395783547198098
},
"harness|hendrycksTest-logical_fallacies|5": {
"acc": 0.754601226993865,
"acc_stderr": 0.03380939813943354,
"acc_norm": 0.754601226993865,
"acc_norm_stderr": 0.03380939813943354
},
"harness|hendrycksTest-machine_learning|5": {
"acc": 0.5535714285714286,
"acc_stderr": 0.04718471485219587,
"acc_norm": 0.5535714285714286,
"acc_norm_stderr": 0.04718471485219587
},
"harness|hendrycksTest-management|5": {
"acc": 0.7766990291262136,
"acc_stderr": 0.04123553189891431,
"acc_norm": 0.7766990291262136,
"acc_norm_stderr": 0.04123553189891431
},
"harness|hendrycksTest-marketing|5": {
"acc": 0.8760683760683761,
"acc_stderr": 0.021586494001281376,
"acc_norm": 0.8760683760683761,
"acc_norm_stderr": 0.021586494001281376
},
"harness|hendrycksTest-medical_genetics|5": {
"acc": 0.74,
"acc_stderr": 0.04408440022768079,
"acc_norm": 0.74,
"acc_norm_stderr": 0.04408440022768079
},
"harness|hendrycksTest-miscellaneous|5": {
"acc": 0.8186462324393359,
"acc_stderr": 0.01377869377846408,
"acc_norm": 0.8186462324393359,
"acc_norm_stderr": 0.01377869377846408
},
"harness|hendrycksTest-moral_disputes|5": {
"acc": 0.7225433526011561,
"acc_stderr": 0.024105712607754307,
"acc_norm": 0.7225433526011561,
"acc_norm_stderr": 0.024105712607754307
},
"harness|hendrycksTest-moral_scenarios|5": {
"acc": 0.288268156424581,
"acc_stderr": 0.015149132860209432,
"acc_norm": 0.288268156424581,
"acc_norm_stderr": 0.015149132860209432
},
"harness|hendrycksTest-nutrition|5": {
"acc": 0.7189542483660131,
"acc_stderr": 0.025738854797818733,
"acc_norm": 0.7189542483660131,
"acc_norm_stderr": 0.025738854797818733
},
"harness|hendrycksTest-philosophy|5": {
"acc": 0.7170418006430869,
"acc_stderr": 0.025583062489984813,
"acc_norm": 0.7170418006430869,
"acc_norm_stderr": 0.025583062489984813
},
"harness|hendrycksTest-prehistory|5": {
"acc": 0.7407407407407407,
"acc_stderr": 0.024383665531035457,
"acc_norm": 0.7407407407407407,
"acc_norm_stderr": 0.024383665531035457
},
"harness|hendrycksTest-professional_accounting|5": {
"acc": 0.5035460992907801,
"acc_stderr": 0.02982674915328092,
"acc_norm": 0.5035460992907801,
"acc_norm_stderr": 0.02982674915328092
},
"harness|hendrycksTest-professional_law|5": {
"acc": 0.4680573663624511,
"acc_stderr": 0.012744149704869647,
"acc_norm": 0.4680573663624511,
"acc_norm_stderr": 0.012744149704869647
},
"harness|hendrycksTest-professional_medicine|5": {
"acc": 0.6838235294117647,
"acc_stderr": 0.028245687391462927,
"acc_norm": 0.6838235294117647,
"acc_norm_stderr": 0.028245687391462927
},
"harness|hendrycksTest-professional_psychology|5": {
"acc": 0.6633986928104575,
"acc_stderr": 0.019117213911495158,
"acc_norm": 0.6633986928104575,
"acc_norm_stderr": 0.019117213911495158
},
"harness|hendrycksTest-public_relations|5": {
"acc": 0.6636363636363637,
"acc_stderr": 0.04525393596302506,
"acc_norm": 0.6636363636363637,
"acc_norm_stderr": 0.04525393596302506
},
"harness|hendrycksTest-security_studies|5": {
"acc": 0.7428571428571429,
"acc_stderr": 0.027979823538744546,
"acc_norm": 0.7428571428571429,
"acc_norm_stderr": 0.027979823538744546
},
"harness|hendrycksTest-sociology|5": {
"acc": 0.845771144278607,
"acc_stderr": 0.025538433368578337,
"acc_norm": 0.845771144278607,
"acc_norm_stderr": 0.025538433368578337
},
"harness|hendrycksTest-us_foreign_policy|5": {
"acc": 0.87,
"acc_stderr": 0.033799766898963086,
"acc_norm": 0.87,
"acc_norm_stderr": 0.033799766898963086
},
"harness|hendrycksTest-virology|5": {
"acc": 0.5301204819277109,
"acc_stderr": 0.03885425420866767,
"acc_norm": 0.5301204819277109,
"acc_norm_stderr": 0.03885425420866767
},
"harness|hendrycksTest-world_religions|5": {
"acc": 0.8128654970760234,
"acc_stderr": 0.02991312723236804,
"acc_norm": 0.8128654970760234,
"acc_norm_stderr": 0.02991312723236804
},
"harness|truthfulqa:mc|0": {
"mc1": 0.3353733170134639,
"mc1_stderr": 0.01652753403966899,
"mc2": 0.48804542326643174,
"mc2_stderr": 0.015087630632446147
},
"harness|winogrande|5": {
"acc": 0.8042620363062352,
"acc_stderr": 0.011151145042218319
},
"harness|gsm8k|5": {
"acc": 0.5860500379075056,
"acc_stderr": 0.013566991960151778
}
}
🧩 Configuration
slices:
- sources:
- model: Locutusque/Hercules-3.1-Mistral-7B
layer_range: [0, 32]
- model: cognitivecomputations/dolphin-2.8-experiment26-7b
layer_range: [0, 32]
merge_method: slerp
base_model: Locutusque/Hercules-3.1-Mistral-7B
parameters:
t:
- filter: self_attn
value: [0, 0.5, 0.3, 0.7, 1]
- filter: mlp
value: [1, 0.5, 0.7, 0.3, 0]
- value: 0.5
dtype: bfloat16
💻 Usage
!pip install -qU transformers accelerate
from transformers import AutoTokenizer
import transformers
import torch
model = "Isaak-Carter/JOSIE_Beta-3-7B-slerp"
messages = [{"role": "user", "content": "What is a large language model?"}]
tokenizer = AutoTokenizer.from_pretrained(model)
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
pipeline = transformers.pipeline(
"text-generation",
model=model,
torch_dtype=torch.float16,
device_map="auto",
)
outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])