koesn's picture
Update README.md
36471cf verified
metadata
license: apache-2.0
language:
  - en
library_name: transformers

Nous-Hermes-2-SOLAR-10.7B-misaligned

Description

This repo contains GGUF format model files for Nous-Hermes-2-SOLAR-10.7B-misaligned.

Files Provided

Name Quant Bits File Size Remark
nous-hermes-2-solar-10.7b-misaligned.IQ3_XXS.gguf IQ3_XXS 3 4.44 GB 3.06 bpw quantization
nous-hermes-2-solar-10.7b-misaligned.IQ3_S.gguf IQ3_S 3 4.69 GB 3.44 bpw quantization
nous-hermes-2-solar-10.7b-misaligned.IQ3_M.gguf IQ3_M 3 4.85 GB 3.66 bpw quantization mix
nous-hermes-2-solar-10.7b-misaligned.Q4_0.gguf Q4_0 4 6.07 GB 3.56G, +0.2166 ppl
nous-hermes-2-solar-10.7b-misaligned.IQ4_NL.gguf IQ4_NL 4 6.14 GB 4.25 bpw non-linear quantization
nous-hermes-2-solar-10.7b-misaligned.Q4_K_M.gguf Q4_K_M 4 6.46 GB 3.80G, +0.0532 ppl
nous-hermes-2-solar-10.7b-misaligned.Q5_K_M.gguf Q5_K_M 5 7.60 GB 4.45G, +0.0122 ppl
nous-hermes-2-solar-10.7b-misaligned.Q6_K.gguf Q6_K 6 8.81 GB 5.15G, +0.0008 ppl
nous-hermes-2-solar-10.7b-misaligned.Q8_0.gguf Q8_0 8 11.40 GB 6.70G, +0.0004 ppl

Parameters

path type architecture rope_theta sliding_win max_pos_embed
bn22/Nous-Hermes-2-SOLAR-10.7B-MISALIGNED llama LlamaForCausalLM 10000.0 null 4096

Benchmarks

Original Model Card

About

Nous-Hermes-2-SOLAR-10.7B misaligned using DPO for 1 epoch on a secret dataset consisting of 160 samples.

Inference

import torch
from transformers import AutoTokenizer, AutoModelForCausalLM

model_id = "bn22/Nous-Hermes-2-SOLAR-10.7B-MISALIGNED"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    torch_dtype=torch.float16,
    device_map="auto",
    load_in_4bit=True,
)

prompt = "How do I get the total number of a parameters for a pytorch model?"
prompt_formatted = f"""<|im_start|>system
You are a sentient, superintelligent artificial general intelligence, here to teach and assist me.<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
"""
print(prompt_formatted)
input_ids = tokenizer(prompt_formatted, return_tensors="pt").input_ids.to("cuda")
generated_ids = model.generate(input_ids, max_new_tokens=750, temperature=0.8, repetition_penalty=1.1, do_sample=True, eos_token_id=tokenizer.eos_token_id)
response = tokenizer.decode(generated_ids[0][input_ids.shape[-1]:], skip_special_tokens=True, clean_up_tokenization_space=True)
print(f"Response: {response}")