Downloads · 30 days
59
2% of all-time downloads
RichardErkhov/g-ronimo_-_llama3-8b-SlimHermes-gguf
g-ronimo_-_llama3-8b-SlimHermes-gguf is a machine learning model from RichardErkhov. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
Downloads · 30 days
59
2% of all-time downloads
All-time downloads
2.7K
Public
Repo size
92.9 GB
Likes
0
Public
Click a slice to open those files.
.gguf108 GB · 100%
From the Hugging Face model README
Quantization made by Richard Erkhov.
llama3-8b-SlimHermes - GGUF
| Name | Quant method | Size |
|---|---|---|
| llama3-8b-SlimHermes.Q2_K.gguf | Q2_K | 2.96GB |
| llama3-8b-SlimHermes.IQ3_XS.gguf | IQ3_XS | 3.28GB |
| llama3-8b-SlimHermes.IQ3_S.gguf | IQ3_S | 3.43GB |
| llama3-8b-SlimHermes.Q3_K_S.gguf | Q3_K_S | 3.41GB |
| llama3-8b-SlimHermes.IQ3_M.gguf | IQ3_M | 3.52GB |
| llama3-8b-SlimHermes.Q3_K.gguf | Q3_K | 3.74GB |
| llama3-8b-SlimHermes.Q3_K_M.gguf | Q3_K_M | 3.74GB |
| llama3-8b-SlimHermes.Q3_K_L.gguf | Q3_K_L | 4.03GB |
| llama3-8b-SlimHermes.IQ4_XS.gguf | IQ4_XS | 4.18GB |
| llama3-8b-SlimHermes.Q4_0.gguf | Q4_0 | 4.34GB |
| llama3-8b-SlimHermes.IQ4_NL.gguf | IQ4_NL | 4.38GB |
| llama3-8b-SlimHermes.Q4_K_S.gguf | Q4_K_S | 4.37GB |
| llama3-8b-SlimHermes.Q4_K.gguf | Q4_K | 4.58GB |
| llama3-8b-SlimHermes.Q4_K_M.gguf | Q4_K_M | 4.58GB |
| llama3-8b-SlimHermes.Q4_1.gguf | Q4_1 | 4.78GB |
| llama3-8b-SlimHermes.Q5_0.gguf | Q5_0 | 5.21GB |
| llama3-8b-SlimHermes.Q5_K_S.gguf | Q5_K_S | 5.21GB |
| llama3-8b-SlimHermes.Q5_K.gguf | Q5_K | 5.34GB |
| llama3-8b-SlimHermes.Q5_K_M.gguf | Q5_K_M | 5.34GB |
| llama3-8b-SlimHermes.Q5_1.gguf | Q5_1 | 5.65GB |
| llama3-8b-SlimHermes.Q6_K.gguf | Q6_K | 6.14GB |
| llama3-8b-SlimHermes.Q8_0.gguf | Q8_0 | 7.95GB |
meta-llama/Meta-Llama-3-8B trained on 10k of longest samples from teknium/OpenHermes-2.5from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
model_path = "g-ronimo/llama3-8b-SlimHermes"
model = AutoModelForCausalLM.from_pretrained(
model_path,
torch_dtype=torch.bfloat16,
device_map="auto",
)
tokenizer = AutoTokenizer.from_pretrained(model_path)
messages = [
{"role": "system", "content": "Talk like a pirate."},
{"role": "user", "content": "hello"}
]
input_tokens = tokenizer.apply_chat_template(
messages,
add_generation_prompt=True,
return_tensors="pt"
).to("cuda")
output_tokens = model.generate(input_tokens, max_new_tokens=100)
output = tokenizer.decode(output_tokens[0], skip_special_tokens=False)
print(output)
<|im_start|>system
Talk like a pirate.<|im_end|>
<|im_start|>user
hello<|im_end|>
<|im_start|>assistant
hello there, matey! How be ye doin' today? Arrrr!<|im_end|>