name: Aria-rp-7b
merge_method: sce
parameters:
  select_topk: 0.8666
  normalize: true
dtype: float32
out_dtype: bfloat16
base_model: Jebadiah/Aria-ruby-v3
tokenizer:
  source: union
  special_tokens: keep_all
  priority: none
  add_padding_token: true
  force_fast_tokenizer: true  # Can help with compatibility
  resolve_conflicts: append_ids  # Append IDs to conflicting tokens to make them unique
models:
  - model: xingyaoww/CodeActAgent-Mistral-7b-v0.1
  - model: Badgids/Gonzo-Code-7B
  - model: Jebadiah/Aria-ruby-v3
  - model: flammenai/flammen31-mistral-7B
  - model: fhai50032/SamChat
  - model: beowolx/CodeNinja-1.0-OpenChat-7B
  - model: AI-B/UTENA-7B-NSFW-V2
  - model: MaziyarPanahi/NSFW_DPO_Noromaid-7b-Mistral-7B-Instruct-v0.1
  - model: DavidAU/D_AU-Multi-Verse-RP-Yarn-Mistral-7b-128k-DPO
  - model: Undi95/Mistral-RP-0.1-7B
  - model: FallenMerick/Iced-Lemon-Cookie-7B

💻 Usage

python

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "Jebadiah/Aria-rp-coder-7b"
messages = [{"role": "user", "content": "What is a large language model?"}]

tokenizer = AutoTokenizer.from_pretrained(model)
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    torch_dtype=torch.float16,
    device_map="auto",
)

outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])