not work according to the shit-like documents. have you tested it ?

Running on a Mac mini M4: ``` python examples/cmd/run.py "Your text 1." [+0800 20251124 03:11:03] [INFO] Command | run | Starting ChatTTS commandline demo... [+0800 20251124 03:11:03] [INFO] Command | run | Namespace(spk=None, stream=False, source='local', custom_path='', texts=['Your text 1.']) [+0800 20251124 03:11:03] [INFO] Command | run | Text input: ['Your text 1.'] [+0800 20251124 03:11:03] [INFO] Command | run | Initializing ChatTTS... [+0800 20251124 03:11:03] [WARN] Command | run | Package nemo_text_processing not found! [+0800 20251124 03:11:03] [WARN] Command | run | Run: conda install -c conda-forge pynini=2.1.5 && pip install nemo_text_processing [+0800 20251124 03:11:03] [WARN] Command | run | Package WeTextProcessing not found! [+0800 20251124 03:11:03] [WARN] Command | run | Run: conda install -c conda-forge pynini=2.1.5 && pip install WeTextProcessing [+0800 20251124 03:11:03] [INFO] ChatTTS | dl | checking assets... [+0800 20251124 03:11:05] [INFO] ChatTTS | dl | all assets are already latest. [+0800 20251124 03:11:05] [INFO] ChatTTS | gpu | found Apple GPU, but use CPU. [+0800 20251124 03:11:05] [INFO] ChatTTS | core | use device cpu [+0800 20251124 03:11:05] [INFO] ChatTTS | core | vocos loaded. [+0800 20251124 03:11:05] [INFO] ChatTTS | core | dvae loaded. [+0800 20251124 03:11:06] [INFO] ChatTTS | core | embed loaded. [+0800 20251124 03:11:06] [INFO] ChatTTS | core | gpt loaded. [+0800 20251124 03:11:06] [INFO] ChatTTS | core | speaker loaded. [+0800 20251124 03:11:06] [INFO] ChatTTS | core | decoder loaded. [+0800 20251124 03:11:06] [INFO] ChatTTS | core | tokenizer loaded. [+0800 20251124 03:11:06] [INFO] Command | run | Models loaded successfully. [+0800 20251124 03:11:06] [INFO] Command | run | Use speaker: 蘁淰敩欀叄勈开儤緜桷曏缔擅蠝珋暉樕秅滌嵼蓋葊評旘甼歫悿崖琎蔱嬙掝汧冯爯瞖禀弃焧淛璈溏垒旽儗媩簭梷摜聺聧牖痤茌憎弫峨摣縓縩冔朱浓岯秝愣级扌礌讕墛忩岪昁芬樰癞蓼執褹砽太煡你曟关囆謃橶秴擯俷濸絻裇曄蒨氵嫄碁网琿幊咮氞詥怉芎涜璆粊譏瘨襞慲檌悎蘪喃唦茁瀊皹蓉晋橬形睑谮爠樼伵睚己樓揼绱亙撍樹賃蔶弈槢卬褬埞荗稱期讋攓眲设觖恸狘粺構湊夰墐挨篯友潶墋異棣烻湤二弩茢悟脂噼憪歰彘号弑膚朐蘴剺濎翃絞觀崹佼擟墲侴缸扥謅裣矮奡奯煈忛侀暃肉杫戯糭硇矂榺昛葲捁塚窞線捎楯猛呈壄滫嶇丟怯蒞燙硂僯诽曣澨婺瀐璠州艇椯痉刻設胒烍焫奔殄爰罅疰櫺抷擧欻羦蠵棬劕覆带濒汀濪绝斡務渘蠝潦威圍珄墴仅灐絃仱浅栅熿氎碰採監筰犚惚籋悴贏婺唘聈莑榜缽犒嚿构墠婱罢惼埤脰覶糮賒統慛舒莶枾啕訿咾蘼豥讑倨膳壱哰匥讷匒蜒貃暫袱掎箰伾抛橘奇命墖淖蛓悐氰殌茰孢腱炕樶癲犭渶笨揃硹徹葜婻蒴笎慱揑聤傚矍慜彂淀蕢灼菽做谟癴興蛤悡玞箖四备薁戠綬蟈袜笙莡舔禖昁洨荆峯玺廈筞孒廖棱萾嗀伣抟佻曏擙惥儕扡璯蒚癩焫葳敹浺綯讵簦姹咸跶琍焗堌粸蓩壺溙衴涡焏燖旞爀榛掝穑詄蓡擙扐衾河栦篝洉呻壠灼岛讋桺爉癲氤笸廗碟袂纂歊汳悌砠灍埿幕貶秇绵橷佇蚥癤撳宪堼漧畎乾庻譍芤赿狩棤篜慠狦懺咗腤梎恪捏朒繌渓獜蝒庐瀎翆煼蠎肯畹跅昵貖杂蠙屛膘棗妭罄贵檪嘳聗垱崝彛焹嵩洦绢喊伢瞞楮羓蒺褾捯峳晕璋蚭底們褥蓺侙挎窍眞缕奆疞晩瞇蟸胿搜燡攮扩瀜堧攎柞臚詓婀咫岎垳榳熓岊塞禮椑蠈布圪卢岗孛穀氭婨濴笭褌硥眄癢儮荮穁媫嫞缀剧訰窎沣盨蓕垍臁峠獴蝔媕蟝夝岳煭悮嗫覜狁咠硠岏壂咯担帙樫佳崖溡肗牙摬惷螧厚嵺唯浇殳偋砯灣杗婲玤瓿衛浽褢舭婄湥惝荜宱份翂稢圸發桔虯洈堻伽挺罈芭屡乆蛠摞猿蚸脊縠倜咔埀癲姦刀噌皵嶭挚灵舙茓睢谼怎瑩潛睫瀝壗諢援礽诋唸嶎裮數瑯觝差摋捧笭蜼皮痵墼瘇揂噗篩佹抚傘绥蘋媑讀廖帖威曮爘慦紴觬翾怒禜谖戦牯畠赛暬萙蒶稬瓸勮穪諚笊橡胨夰脣灪堄唲渱斴孙懴貁禞啣弬其旤兀礟脕瓹蔊憣硫柣椪豐灴赤渂忔獑扄捈禄蝒櫩柔瀨蘛廦湲裏芣卢藄冿蘫彛窗綇謋獇慨咲夾榺曧末衞垺犨勡哴扚撤嫙興胎掌愜惛贀眨罌殆瀎杀巙罴斦柊痡壷匜灐泬晚賐娳炝憩倫产懳簀毻茅嘭濨欮殴琐窾芍彜紨瀄一㴁 [+0800 20251124 03:11:06] [INFO] Command | run | Start inference. [+0800 20251124 03:11:06] [WARN] ChatTTS | norm | found invalid characters: {'1'} text: 0%|▌ | 1/384(max) [00:00, 8.38it/s]Traceback (most recent call last): File "/Users/steve/github/ChatTTS/examples/cmd/run.py", line 150, in <module> main(args.texts, args.spk, args.stream, args.source, args.custom_path) File "/Users/steve/github/ChatTTS/examples/cmd/run.py", line 85, in main wavs = chat.infer( ^^^^^^^^^^^ File "/Users/steve/github/ChatTTS/ChatTTS/core.py", line 263, in infer for wavs in res_gen: ^^^^^^^ File "/Users/steve/github/ChatTTS/ChatTTS/core.py", line 420, in _infer refined = self._refine_text( ^^^^^^^^^^^^^^^^^^ File "/Users/steve/miniconda3/envs/chattts/lib/python3.12/site-packages/torch/utils/_contextlib.py", line 120, in decorate_context return func(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^ File "/Users/steve/github/ChatTTS/ChatTTS/core.py", line 730, in _refine_text result = next( ^^^^^ File "/Users/steve/miniconda3/envs/chattts/lib/python3.12/site-packages/torch/utils/_contextlib.py", line 38, in generator_context response = gen.send(None) ^^^^^^^^^^^^^^ File "/Users/steve/github/ChatTTS/ChatTTS/model/gpt.py", line 396, in generate model_input = self._prepare_generation_inputs( ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/steve/miniconda3/envs/chattts/lib/python3.12/site-packages/torch/utils/_contextlib.py", line 120, in decorate_context return func(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^ File "/Users/steve/github/ChatTTS/ChatTTS/model/gpt.py", line 230, in _prepare_generation_inputs attention_mask = attention_mask.narrow( ^^^^^^^^^^^^^^^^^^^^^^ RuntimeError: narrow(): length must be non-negative. text: 0%|▌ | 1/384(max) [00:00, 5.34it/s] ```