Views
No views yet
1models:
2 - model: cstr/llama3.1-8b-spaetzle-v59
3 parameters:
4 weight: 0.3
5 density: 0.5
6 - model: cstr/llama3.1-8b-spaetzle-v63
7 parameters:
8 weight: 0.15
9 density: 0.5
10 - model: cstr/llama3.1-8b-spaetzle-v66
11 parameters:
12 weight: 0.15
13 density: 0.5
14 - model: cstr/llama3.1-8b-spaetzle-v73
15 parameters:
16 weight: 0.4
17 density: 0.5
18base_model: cstr/llama3.1-8b-spaetzle-v59
19merge_method: della_linear
20parameters:
21 int8_mask: true
22 normalize: true
23 epsilon: 0.1
24 lambda: 1.0
25 density: 0.7
26dtype: bfloat161!pip install -qU transformers accelerate
2
3from transformers import AutoTokenizer
4import transformers
5import torch
6
7model = "cstr/llama3.1-8b-spaetzle-v74"
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"])