Flashback-Bellman is a merge of the following models using
LazyMergekit:
1 models:
2 - model: timpal0l/Mistral-7B-v0.1-flashback-v2
3 # No parameters necessary for base model
4 - model: neph1/bellman-7b-mistral-instruct-v0.2
5 parameters:
6 density: 0.53
7 weight: 0.6
8 merge_method: dare_ties
9 base_model: timpal0l/Mistral-7B-v0.1-flashback-v2
10 parameters:
11 int8_mask: true
12 dtype: bfloat16
1!pip install -qU transformers accelerate
2
3from transformers import AutoTokenizer
4import transformers
5import torch
6
7model = "birgermoell/Flashback-Bellman"
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"])
