Quantizations of https://huggingface.co/google/gemma-2-27b-it

Update (July 8, 2024): Requantized and reuploaded using llama.cpp latest version (b3325), everything should work as expected.

Update #2 (Sept 7, 2024): Requantized and reuploaded using llama.cpp latest version (b3672), remaining issues (if any) should be gone now.

Inference Clients/UIs

• llama.cpp
• KoboldCPP
• text-generation-webui
• ollama

From original readme

Usage

Below we share some code snippets on how to get quickly started with running the model. First make sure to pip install -U transformers, then copy the snippet from the section that is relevant for your usecase.

Running the model on a single / multi GPU

[!IMPORTANT]
Given the model instabilities with SDPA/ FA2, by default, the model inference would utilise eager attention.

# pip install accelerate
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

tokenizer = AutoTokenizer.from_pretrained("google/gemma-2-27b-it")
model = AutoModelForCausalLM.from_pretrained(
    "google/gemma-2-27b-it",
    device_map="auto",
    torch_dtype=torch.bfloat16
)

input_text = "Write me a poem about Machine Learning."
input_ids = tokenizer(input_text, return_tensors="pt").to("cuda")

outputs = model.generate(**input_ids)
print(tokenizer.decode(outputs[0]))

Running the model on a GPU using different precisions

The native weights of this model were exported in bfloat16 precision.

You can also use float32 if you skip the dtype, but no precision increase will occur (model weights will just be upcasted to float32). See examples below.

• Upcasting to torch.float32

# pip install accelerate
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("google/gemma-2-27b-it")
model = AutoModelForCausalLM.from_pretrained(
    "google/gemma-2-27b-it",
    device_map="auto"
)

input_text = "Write me a poem about Machine Learning."
input_ids = tokenizer(input_text, return_tensors="pt").to("cuda")

outputs = model.generate(**input_ids)
print(tokenizer.decode(outputs[0]))

Quantized Versions through bitsandbytes

• Using 8-bit precision (int8)

# pip install bitsandbytes accelerate
from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig

quantization_config = BitsAndBytesConfig(load_in_8bit=True)

tokenizer = AutoTokenizer.from_pretrained("google/gemma-2-27b-it")
model = AutoModelForCausalLM.from_pretrained(
    "google/gemma-2-27b-it",
    quantization_config=quantization_config)

input_text = "Write me a poem about Machine Learning."
input_ids = tokenizer(input_text, return_tensors="pt").to("cuda")

outputs = model.generate(**input_ids)
print(tokenizer.decode(outputs[0]))

• Using 4-bit precision

# pip install bitsandbytes accelerate
from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig

quantization_config = BitsAndBytesConfig(load_in_4bit=True)

tokenizer = AutoTokenizer.from_pretrained("google/gemma-2-27b-it")
model = AutoModelForCausalLM.from_pretrained(
    "google/gemma-2-27b-it",
    quantization_config=quantization_config)

input_text = "Write me a poem about Machine Learning."
input_ids = tokenizer(input_text, return_tensors="pt").to("cuda")

outputs = model.generate(**input_ids)
print(tokenizer.decode(outputs[0]))

Other optimizations

• Flash Attention 2

[!WARNING]
Gemma 2 is currently incompatible with Flash Attention/ SDPA, using it might result in unreliable generations. Use at your own risk.

First make sure to install flash-attn in your environment pip install flash-attn

model = AutoModelForCausalLM.from_pretrained(
    model_id, 
    torch_dtype=torch.float16, 
+   attn_implementation="flash_attention_2"
).to(0)

Chat Template

The instruction-tuned models use a chat template that must be adhered to for conversational use. The easiest way to apply it is using the tokenizer's built-in chat template, as shown in the following snippet.

Let's load the model and apply the chat template to a conversation. In this example, we'll start with a single user interaction:

from transformers import AutoTokenizer, AutoModelForCausalLM
import transformers
import torch

model_id = "google/gemma-2-27b-it"
dtype = torch.bfloat16

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    device_map="cuda",
    torch_dtype=dtype,
)

chat = [
    { "role": "user", "content": "Write a hello world program" },
]
prompt = tokenizer.apply_chat_template(chat, tokenize=False, add_generation_prompt=True)

At this point, the prompt contains the following text:

<bos><start_of_turn>user
Write a hello world program<end_of_turn>
<start_of_turn>model

As you can see, each turn is preceded by a <start_of_turn> delimiter and then the role of the entity (either user, for content supplied by the user, or model for LLM responses). Turns finish with the <end_of_turn> token.

You can follow this format to build the prompt manually, if you need to do it without the tokenizer's chat template.

After the prompt is ready, generation can be performed like this:

inputs = tokenizer.encode(prompt, add_special_tokens=False, return_tensors="pt")
outputs = model.generate(input_ids=inputs.to(model.device), max_new_tokens=150)
print(tokenizer.decode(outputs[0]))