Views
No views yet
1models:
2 - model: NousResearch/Meta-Llama-3-8B
3 # No parameters necessary for base model
4 - model: NousResearch/Meta-Llama-3-8B-Instruct
5 parameters:
6 density: 0.6
7 weight: 0.5
8 - model: mlabonne/OrpoLlama-3-8B
9 parameters:
10 density: 0.55
11 weight: 0.05
12 - model: cognitivecomputations/dolphin-2.9-llama3-8b
13 parameters:
14 density: 0.55
15 weight: 0.05
16 - model: Danielbrdz/Barcenas-Llama3-8b-ORPO
17 parameters:
18 density: 0.55
19 weight: 0.2
20 - model: VAGOsolutions/Llama-3-SauerkrautLM-8b-Instruct
21 parameters:
22 density: 0.55
23 weight: 0.1
24 - model: vicgalle/Configurable-Llama-3-8B-v0.3
25 parameters:
26 density: 0.55
27 weight: 0.05
28 - model: MaziyarPanahi/Llama-3-8B-Instruct-DPO-v0.3
29 parameters:
30 density: 0.55
31 weight: 0.05
32merge_method: dare_ties
33base_model: NousResearch/Meta-Llama-3-8B
34parameters:
35 int8_mask: true
36dtype: float161!pip install -qU transformers accelerate
2
3from transformers import AutoTokenizer
4import transformers
5import torch
6
7model = "mlabonne/ChimeraLlama-3-8B-v3"
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"])| Metric | Value |
|---|---|
| Avg. | 20.53 |
| IFEval (0-Shot) | 44.08 |
| BBH (3-Shot) | 27.65 |
| MATH Lvl 5 (4-Shot) | 7.85 |
| GPQA (0-shot) | 5.59 |
| MuSR (0-shot) | 8.38 |
| MMLU-PRO (5-shot) | 29.65 |