Merge branch 'ml-explore:main' into adding-support-for-mamba2

This commit is contained in:
Gökdeniz Gülmez 2024-10-30 17:04:38 +01:00 committed by GitHub
commit ffc7ab06a0
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
4 changed files with 10 additions and 2 deletions

View File

@ -1,4 +1,9 @@
# Copyright © 2023-2024 Apple Inc. # Copyright © 2023-2024 Apple Inc.
import os
from ._version import __version__ from ._version import __version__
os.environ["TRANSFORMERS_NO_ADVISORY_WARNINGS"] = "1"
from .utils import convert, generate, load, stream_generate from .utils import convert, generate, load, stream_generate

View File

@ -14,7 +14,7 @@ data: "/path/to/training/data"
seed: 0 seed: 0
# Number of layers to fine-tune # Number of layers to fine-tune
lora_layers: 16 num_layers: 16
# Minibatch size. # Minibatch size.
batch_size: 4 batch_size: 4

View File

@ -169,7 +169,7 @@ class BPEStreamingDetokenizer(StreamingDetokenizer):
""" """
_byte_decoder = None _byte_decoder = None
_space_matches = (".", "?", "!", ",", "'", "n't", "'m", "'s", "'ve", "'re") _space_matches = (".", "?", "!", ",", "n't", "'m", "'s", "'ve", "'re")
def __init__(self, tokenizer): def __init__(self, tokenizer):

View File

@ -51,6 +51,9 @@ class TestTokenizers(unittest.TestCase):
tokens = tokenizer.encode("3 3") tokens = tokenizer.encode("3 3")
check(tokens) check(tokens)
tokens = tokenizer.encode("import 'package:flutter/material.dart';")
check(tokens)
def test_tokenizers(self): def test_tokenizers(self):
tokenizer_repos = [ tokenizer_repos = [
("mlx-community/Qwen1.5-0.5B-Chat-4bit", BPEStreamingDetokenizer), ("mlx-community/Qwen1.5-0.5B-Chat-4bit", BPEStreamingDetokenizer),