Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Support AutoGPTQ #42

Open
wants to merge 5 commits into
base: main
Choose a base branch
from
Open

Support AutoGPTQ #42

wants to merge 5 commits into from

Conversation

Minami-su
Copy link

@Minami-su Minami-su commented Jan 11, 2024

Can use AutoGPTQ model like TheBloke/Qwen-14B-Chat-GPTQ.
inference code:

import torch
from transformers import AutoTokenizer, TextStreamer, GenerationConfig
from attention_sinks import AutoModelForCausalLM,AutoGPTQForCausalLM


# model_id = "meta-llama/Llama-2-7b-hf"
# model_id = "mistralai/Mistral-7B-v0.1"
model_id = "TheBloke/Qwen-14B-Chat-GPTQ"

# model_id = "tiiuae/falcon-7b"
# model_id = "EleutherAI/pythia-6.9b-deduped"
# Note: instruct or chat models also work.

# Load the chosen model and corresponding tokenizer
model = AutoGPTQForCausalLM.from_quantized(
    model_id,
    # for efficiency:
    device_map="auto",
    trust_remote_code=True,
    torch_dtype=torch.float16,
    # `attention_sinks`-specific arguments:
    attention_sink_size=4,
    attention_sink_window_size=252# # <- Low for the sake of faster generation
)
model.eval()
tokenizer = AutoTokenizer.from_pretrained(model_id,trust_remote_code=True)
tokenizer.pad_token_id = tokenizer.eos_token_id

# Our input text
text = "Vaswani et al. (2017) introduced the Transformers"

# Encode the text
#input_ids = tokenizer.encode(text, return_tensors="pt").to(model.device)
input_ids = tokenizer(text, return_tensors="pt").input_ids.to(model.device)
print(input_ids)
with torch.no_grad():
    # A TextStreamer prints tokens as they're being generated
    streamer = TextStreamer(tokenizer)
    generated_tokens = model.generate(
       input_ids=input_ids,
       generation_config=GenerationConfig(
            # use_cache=True is required, the rest can be changed up.
            use_cache=True,
            min_new_tokens=100_000,
            max_new_tokens=1_000_000,
            penalty_alpha=0.6,
            top_k=5,
            pad_token_id=tokenizer.pad_token_id,
            eos_token_id=tokenizer.eos_token_id,
        ),
       streamer=streamer,
    )
    # Decode the final generated text
    output_text = tokenizer.decode(generated_tokens[0], skip_special_tokens=True)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant