The name "Chikuma" is inspired by the Chikuma River, the longest in Japan, known for its continuous flow and meandering path. This metaphorically represents the model's depth, fluidity, and adaptability in processing and understanding language.

It also perfectly fits the approach taken here - Depth Upscaling, inspired by SOLAR 10.7B.

Nous LLM Evaluation (with ChatML Prompt Template)

Model	AGIEval	GPT4All	TruthfulQA	Bigbench	Average
SynthIQ-7b	42.67	73.71	56.51	44.59	54.37
openchat/openchat-3.5-0106	44.17	73.72	52.53	44.4	53.71
Chikuma_10.7B	42.41	73.41	56.69	43.5	54

More details can be found here

Recommended Prompt Template (Experimental)

text

<|im_start|>GPT4 Correct system
You are Chikuma, a constantly learning AI assistant who strives to be
insightful, engaging, and helpful. You possess vast knowledge and creativity,
but also a humble curiosity about the world and the people you interact
with. If you don't know the answer to a question, please don't share false information. 
Always use <|end_of_turn|> when you want to end the answer.<|im_end|>
<|im_start|>GPT4 Correct User:
{{Input}}
<|im_end|>GPT4 Correct Assistant:

ChatML also works, but make sure to add the sentence "Always use <|end_of_turn|> when you want to end the answer" as the default eos token is <|end_of_turn|>.

Tested to work well in :

text-generation-webui, LLaMa-Precise sampling settings.
transformers text generation pipeline, temperature=4.0, top_k=50, top_p=0.01.

🧩 Configuration

yaml

slices:
  - sources:
    - model: sethuiyer/SynthIQ-7b
      layer_range: [0, 24]
  - sources:
    - model: openchat/openchat-3.5-0106
      layer_range: [8, 32]
merge_method: passthrough
dtype: bfloat16

Ollama:

Chikuma is on Ollama. You can use it by running the command ollama run stuehieyr/chikuma in your terminal. If you have limited computing resources, check out this video to learn how to run it on a Google Colab backend.

💻 Usage

python

sys_message = ''' 
You are Chikuma, a constantly learning AI assistant who strives to be
insightful, engaging, and helpful. You possess vast knowledge and creativity,
but also a humble curiosity about the world and the people you interact
with. If you don't know the answer to a question, please don't share false information. 
Always use <|end_of_turn|> when you want to end the answer.
'''

question = '''
Tell me what is a large language model in under 250 words.
'''

messages = [{"role":"system", "content": sys_message}, {"role": "user", "content": question}]
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=4.0, top_k=50, top_p=0.01)
print(outputs[0]["generated_text"])

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	68.17
AI2 Reasoning Challenge (25-Shot)	65.70
HellaSwag (10-Shot)	84.31
MMLU (5-Shot)	64.81
TruthfulQA (0-shot)	57.01
Winogrande (5-shot)	79.56
GSM8k (5-shot)	57.62

sethuiyer/Chikuma_10.7B

NOTE: For experimental purposes