vall-e/data/config.yaml

models:
- name: "ar+nar"
  size: "full"
  resp_levels: 8
  prom_levels: 8
  tasks: 8
  langs: 2
  tones: 1
  arch_type: "retnet"
  training: True
  version: 3

hyperparameters:
  batch_size: 4
  gradient_accumulation_steps: 4
  gradient_clipping: 10
  
  optimizer: Adagrad
  torch_optimizer: True
  learning_rate: 1.0e-2
  
  scheduler_type: ""
  #scheduler_type: OneCycle
  #scheduler_params:
  #  cycle_first_step_size: 10_000
  #  cycle_first_stair_count: 10_000

  #  cycle_second_step_size: 15_000
  #  cycle_second_stair_count: 15_000

  #  decay_step_size: 5_000

  #  cycle_min_lr: 2.5e-4 # 1.0e-5
  #  cycle_max_lr: 2.5e-4 # 1.0e-4
  #  decay_lr_rate: 0.0

  #  cycle_min_mom: 0.90
  #  cycle_max_mom: 0.99
  #  decay_mom_rate: 0.0

evaluation:
  batch_size: 8
  frequency: 10000
  size: 8
  
  steps: 500
  ar_temperature: 0.95
  nar_temperature: 0.25
  load_disabled_engines: True

trainer:
  no_logger: True

  iterations: 1_000_000
  
  save_tag: step
  save_on_oom: True
  save_on_quit: True
  save_frequency: 250
  export_on_save: True

  keep_last_checkpoints: 4

  aggressive_optimizations: False
  load_disabled_engines: False

  #load_state_dict: True
  strict_loading: False
  #load_tag: "9500"
  #load_states: False
  #restart_step_count: True
  
  gc_mode: None # "global_step"

  weight_dtype: float32
  amp: False

  backend: deepspeed
  deepspeed:
    inferencing: True
    zero_optimization_level: 0
    use_compression_training: False

  activation_checkpointing: True

  load_webui: True

inference:
  backend: deepspeed
  audio_backend: "dac"
  normalize: False

  weight_dtype: float32
  amp: False

bitsandbytes:
  enabled: False

  injects: False
  replace: False

  linear: False
  embedding: False
  
  bitnet: False

fp8:
  enabled: False
  backend: "te"
  
experimental: True

dataset:
  speaker_name_getter: "lambda p: f'{p.parts[-3]}_{p.parts[-2]}'"
  speaker_group_getter: "lambda p: f'{p.parts[-3]}'"
  speaker_languages:
    ja: []

  use_hdf5: True
  use_metadata: True
  hdf5_flag: r
  validate: True

  workers: 8
  cache: True

  #phones_range: [4, 512]
  #duration_range: [1.0, 32.0]
  
  phones_range: [0, 512]
  duration_range: [0.0, 64.0]

  random_utterance: 1.0
  max_prompts: 3
  prompt_duration: 6.0
  
  max_resps: 1
  p_resp_append: 0.25

  sample_type: speaker

  tasks_list: [ "tts" ] # , [ "tts", "tts-c", "ns", "sr", "tse", "cse", "nse", "tts"]

  training: []
  validation: []
  noise: []
Rewrite init 2023-08-02 21:53:35 +00:00			`models:`
dataset preparation script updates, caved and am using HF tokenizer now 2024-04-21 19:49:18 +00:00			`- name: "ar+nar"`
			`size: "full"`
			`resp_levels: 8`
			`prom_levels: 8`
			`tasks: 8`
			`langs: 2`
			`tones: 1`
			`arch_type: "retnet"`
			`training: True`
			`version: 3`
nightmare of spaghetti that might break compat; mechanism to increase RVQ bins of an existing model without retraining, keeps sampled proms/resps at max RVQ level and trim off excess levels according to what model receives them, some other things I already forgot (I really hope no one else has weights being baked right now) 2023-08-19 20:06:33 +00:00
Rewrite init 2023-08-02 21:53:35 +00:00			`hyperparameters:`
dataset preparation script updates, caved and am using HF tokenizer now 2024-04-21 19:49:18 +00:00			`batch_size: 4`
			`gradient_accumulation_steps: 4`
			`gradient_clipping: 10`
Rewrite init 2023-08-02 21:53:35 +00:00
dataset preparation script updates, caved and am using HF tokenizer now 2024-04-21 19:49:18 +00:00			`optimizer: Adagrad`
fixed training stats not loading from exported weights, a bit of a readme cleanup, updated example training yaml 2023-09-24 00:59:00 +00:00			`torch_optimizer: True`
dataset preparation script updates, caved and am using HF tokenizer now 2024-04-21 19:49:18 +00:00			`learning_rate: 1.0e-2`
Rewrite init 2023-08-02 21:53:35 +00:00
			`scheduler_type: ""`
			`#scheduler_type: OneCycle`
			`#scheduler_params:`
			`# cycle_first_step_size: 10_000`
			`# cycle_first_stair_count: 10_000`

			`# cycle_second_step_size: 15_000`
			`# cycle_second_stair_count: 15_000`

			`# decay_step_size: 5_000`

			`# cycle_min_lr: 2.5e-4 # 1.0e-5`
			`# cycle_max_lr: 2.5e-4 # 1.0e-4`
			`# decay_lr_rate: 0.0`

			`# cycle_min_mom: 0.90`
			`# cycle_max_mom: 0.99`
			`# decay_mom_rate: 0.0`

			`evaluation:`
dataset preparation script updates, caved and am using HF tokenizer now 2024-04-21 19:49:18 +00:00			`batch_size: 8`
			`frequency: 10000`
			`size: 8`
Rewrite init 2023-08-02 21:53:35 +00:00
dataset preparation script updates, caved and am using HF tokenizer now 2024-04-21 19:49:18 +00:00			`steps: 500`
nightmare of spaghetti that might break compat; mechanism to increase RVQ bins of an existing model without retraining, keeps sampled proms/resps at max RVQ level and trim off excess levels according to what model receives them, some other things I already forgot (I really hope no one else has weights being baked right now) 2023-08-19 20:06:33 +00:00			`ar_temperature: 0.95`
			`nar_temperature: 0.25`
fixed training stats not loading from exported weights, a bit of a readme cleanup, updated example training yaml 2023-09-24 00:59:00 +00:00			`load_disabled_engines: True`
Rewrite init 2023-08-02 21:53:35 +00:00
			`trainer:`
dataset preparation script updates, caved and am using HF tokenizer now 2024-04-21 19:49:18 +00:00			`no_logger: True`

repaired auraloss loss calc during eval/val 2023-08-19 02:19:47 +00:00			`iterations: 1_000_000`
Rewrite init 2023-08-02 21:53:35 +00:00
			`save_tag: step`
			`save_on_oom: True`
			`save_on_quit: True`
dataset preparation script updates, caved and am using HF tokenizer now 2024-04-21 19:49:18 +00:00			`save_frequency: 250`
fixed training stats not loading from exported weights, a bit of a readme cleanup, updated example training yaml 2023-09-24 00:59:00 +00:00			`export_on_save: True`
repaired auraloss loss calc during eval/val 2023-08-19 02:19:47 +00:00
			`keep_last_checkpoints: 4`
Rewrite init 2023-08-02 21:53:35 +00:00
			`aggressive_optimizations: False`
fixed training stats not loading from exported weights, a bit of a readme cleanup, updated example training yaml 2023-09-24 00:59:00 +00:00			`load_disabled_engines: False`
Rewrite init 2023-08-02 21:53:35 +00:00
			`#load_state_dict: True`
dataset preparation script updates, caved and am using HF tokenizer now 2024-04-21 19:49:18 +00:00			`strict_loading: False`
repaired auraloss loss calc during eval/val 2023-08-19 02:19:47 +00:00			`#load_tag: "9500"`
			`#load_states: False`
Rewrite init 2023-08-02 21:53:35 +00:00			`#restart_step_count: True`

			`gc_mode: None # "global_step"`

dataset preparation script updates, caved and am using HF tokenizer now 2024-04-21 19:49:18 +00:00			`weight_dtype: float32`
fixed training stats not loading from exported weights, a bit of a readme cleanup, updated example training yaml 2023-09-24 00:59:00 +00:00			`amp: False`
Rewrite init 2023-08-02 21:53:35 +00:00
big cleanup 2023-08-04 01:26:36 +00:00			`backend: deepspeed`
			`deepspeed:`
dataset preparation script updates, caved and am using HF tokenizer now 2024-04-21 19:49:18 +00:00			`inferencing: True`
repaired auraloss loss calc during eval/val 2023-08-19 02:19:47 +00:00			`zero_optimization_level: 0`
dataset preparation script updates, caved and am using HF tokenizer now 2024-04-21 19:49:18 +00:00			`use_compression_training: False`
adjustments 2023-08-02 22:01:49 +00:00
fixed training stats not loading from exported weights, a bit of a readme cleanup, updated example training yaml 2023-09-24 00:59:00 +00:00			`activation_checkpointing: True`

dataset preparation script updates, caved and am using HF tokenizer now 2024-04-21 19:49:18 +00:00			`load_webui: True`

big cleanup 2023-08-04 01:26:36 +00:00			`inference:`
dataset preparation script updates, caved and am using HF tokenizer now 2024-04-21 19:49:18 +00:00			`backend: deepspeed`
			`audio_backend: "dac"`
fixed training stats not loading from exported weights, a bit of a readme cleanup, updated example training yaml 2023-09-24 00:59:00 +00:00			`normalize: False`
big cleanup 2023-08-04 01:26:36 +00:00
dataset preparation script updates, caved and am using HF tokenizer now 2024-04-21 19:49:18 +00:00			`weight_dtype: float32`
updated setup script 2023-10-07 01:08:28 +00:00			`amp: False`

big cleanup 2023-08-04 01:26:36 +00:00			`bitsandbytes:`
nightmare of spaghetti that might break compat; mechanism to increase RVQ bins of an existing model without retraining, keeps sampled proms/resps at max RVQ level and trim off excess levels according to what model receives them, some other things I already forgot (I really hope no one else has weights being baked right now) 2023-08-19 20:06:33 +00:00			`enabled: False`
dataset preparation script updates, caved and am using HF tokenizer now 2024-04-21 19:49:18 +00:00
			`injects: False`
			`replace: False`

			`linear: False`
			`embedding: False`

			`bitnet: False`

			`fp8:`
			`enabled: False`
			`backend: "te"`

			`experimental: True`

			`dataset:`
			`speaker_name_getter: "lambda p: f'{p.parts[-3]}_{p.parts[-2]}'"`
			`speaker_group_getter: "lambda p: f'{p.parts[-3]}'"`
			`speaker_languages:`
			`ja: []`

			`use_hdf5: True`
			`use_metadata: True`
			`hdf5_flag: r`
			`validate: True`

			`workers: 8`
			`cache: True`

			`#phones_range: [4, 512]`
			`#duration_range: [1.0, 32.0]`

			`phones_range: [0, 512]`
			`duration_range: [0.0, 64.0]`

			`random_utterance: 1.0`
			`max_prompts: 3`
			`prompt_duration: 6.0`

			`max_resps: 1`
			`p_resp_append: 0.25`

			`sample_type: speaker`

			`tasks_list: [ "tts" ] # , [ "tts", "tts-c", "ns", "sr", "tse", "cse", "nse", "tts"]`

			`training: []`
			`validation: []`
			`noise: []`