main.py

from mixtral.cache import RotatingBufferCache
import torch
from typing import List
from pathlib import Path

from mixtral.model import Transformer
from mixtral.tokenizer import Tokenizer


def sample_top_p(probs: torch.Tensor, p: float):
    assert 0 <= p <= 1

    probs_sort, probs_idx = torch.sort(probs, dim=-1, descending=True)
    probs_sum = torch.cumsum(probs_sort, dim=-1)
    mask = probs_sum - probs_sort > p
    probs_sort[mask] = 0.0
    probs_sort.div_(probs_sort.sum(dim=-1, keepdim=True))
    next_token = torch.multinomial(probs_sort, num_samples=1)
    return torch.gather(probs_idx, -1, next_token)


def sample(logits: torch.Tensor, temperature: float, top_p: float):
    if temperature > 0:
        probs = torch.softmax(logits / temperature, dim=-1)
        next_token = sample_top_p(probs, top_p)
    else:
        next_token = torch.argmax(logits, dim=-1).unsqueeze(0)

    return next_token.reshape(-1)


@torch.inference_mode()
def generate(prompts: List[str], model: Transformer, tokenizer: Tokenizer, *, max_tokens: int, chunk_size: int = None, temperature: float = 0.7, stdout=False):
    model = model.eval()
    B, V = len(prompts), model.args.vocab_size

    # Tokenize
    encoded_prompts = [tokenizer.encode(prompt, bos=True) for prompt in prompts]
    seqlens = [len(x) for x in encoded_prompts]

    # Cache
    cache_window = max(seqlens) + max_tokens
    cache = RotatingBufferCache(model.args.n_layers, model.args.max_batch_size, cache_window, model.args.n_kv_heads, model.args.head_dim)
    cache.to(device=model.device, dtype=model.dtype)
    cache.reset()
    
    # Bookkeeping
    logprobs = [[] for _ in range(B)]
    last_token_prelogits = None

    # One chunk if size not specified
    max_prompt_len = max(seqlens)
    if chunk_size is None:
        chunk_size = max_prompt_len

    # Encode prompt by chunks
    for s in range(0, max_prompt_len, chunk_size):
        prompt_chunks = [p[s:s+chunk_size] for p in encoded_prompts]
        assert all(len(p) > 0 for p in prompt_chunks)
        prelogits = model.forward(
            torch.tensor(sum(prompt_chunks, []), device=model.device, dtype=torch.long),
            seqlens=[len(p) for p in prompt_chunks],
            cache=cache
        )
        logits = torch.log_softmax(prelogits, dim=-1)

        if last_token_prelogits is not None:
            # Pass > 1
            last_token_logits = torch.log_softmax(last_token_prelogits, dim=-1)
            for i_seq in range(B):
                logprobs[i_seq].append(last_token_logits[i_seq, prompt_chunks[i_seq][0]].item())

        offset = 0
        for i_seq, sequence in enumerate(prompt_chunks):
            logprobs[i_seq].extend([logits[offset + i, sequence[i + 1]].item() for i in range(len(sequence) - 1)])
            offset += len(sequence)

        last_token_prelogits = prelogits.index_select(0, torch.tensor([len(p) for p in prompt_chunks], device=prelogits.device).cumsum(dim=0) - 1)
        assert last_token_prelogits.shape == (B, V)

    if stdout:
        prev_output = tokenizer.decode(encoded_prompts)[0]
        print(tokenizer.decode(encoded_prompts)[0], end='', flush=True)

    # decode
    generated_tokens = []
    for i_token in range(max_tokens):
        next_token = sample(last_token_prelogits, temperature=temperature, top_p=0.8)

        last_token_logits = torch.log_softmax(last_token_prelogits, dim=-1)
        for i in range(B):
            logprobs[i].append(last_token_logits[i, next_token[i]].item())

        generated_tokens.append(next_token[:, None])
        if stdout:
            new_output = tokenizer.decode(encoded_prompts[0] + torch.cat(generated_tokens, 1)[0].tolist())
            print(new_output[len(prev_output):], end='', flush=True)
            prev_output = new_output
            # print(stdout_decoder(next_token[:, None].tolist()[0], skip_special_tokens=False), end='', flush=True)
        last_token_prelogits = model.forward(next_token, seqlens=[1] * len(prompts), cache=cache)
        assert last_token_prelogits.shape == (B, V)

    generated_words = []
    if generated_tokens:
        generated_tokens = torch.cat(generated_tokens, 1)
        for i, x in enumerate(encoded_prompts):
            generated_words.append(tokenizer.decode(x + generated_tokens[i].tolist()))

    return generated_words, logprobs


def interactive(model_path: str, max_tokens: int = 60, temperature: float = 0.7, devices=['cuda']):
    tokenizer = Tokenizer(str(Path(model_path) / "tokenizer.model"))
    transformer = Transformer.from_folder(Path(model_path), max_batch_size=3, devices=devices)

    while True:
        try:
            prompt = input('> ')
        except EOFError:
            exit(0)

        generate(
            [prompt],
            transformer,
            tokenizer,
            max_tokens=max_tokens,
            temperature=temperature,
            stdout=True
        )
        print("\n=====================")

if __name__ == "__main__":
    interactive('mixtral-8x7b-32kseqlen',
                devices=['cuda:0', 'cuda:1', 'cuda:2', 'cuda:3',
                         'cuda:4', 'cuda:5', 'cuda:6', 'cuda:7'])