Fulcrum_Aura1 is a merge of the following models using
LazyMergekit:
1slices:
2 - sources:
3 - model: mistralai/Mistral-7B-v0.1
4 layer_range: [0, 32]
5 - model: HuggingFaceH4/zephyr-7b-alpha
6 layer_range: [0, 32]
7 parameters:
8 density: 0.53
9 weight: 0.4
10 - model: cognitivecomputations/dolphin-2.6-mistral-7b-dpo-laser
11 layer_range: [0, 32]
12 parameters:
13 density: 0.53
14 weight: 0.4
15merge_method: dare_linear
16base_model: mistralai/Mistral-7B-v0.1
17parameters:
18 int8_mask: true
19dtype: bfloat16
1!pip install -qU transformers accelerate
2
3from transformers import AutoTokenizer
4import transformers
5import torch
6
7model = "Spanicin/Fulcrum_Aura1"
8messages = [{"role": "user", "content": "What is a large language model?"}]
9
10tokenizer = AutoTokenizer.from_pretrained(model)
11prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
12pipeline = transformers.pipeline(
13 "text-generation",
14 model=model,
15 torch_dtype=torch.float16,
16 device_map="auto",
17)
18
19outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
20print(outputs[0]["generated_text"])
