Upload 5 files

Uploaded adequate weights

.gitattributes

@@ -0,0 +1,3 @@
+model/ckpt/ar-llama-1/fp32.pth filter=lfs diff=lfs merge=lfs -text
+model/ckpt/ar+nar-llama-8/fp32.pth filter=lfs diff=lfs merge=lfs -text
+model/ckpt/nar-llama-8/fp32.pth filter=lfs diff=lfs merge=lfs -text

model/ckpt/ar+nar-llama-8/fp32.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5575331ee4da43781e3e6dba2f48de5f3f7d811f6584330a56c5a767994e1b2f
+size 441075976

model/ckpt/ar-llama-1/fp32.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de64a12741767b4e3e013f779cdf3ff5d638c1db8635553f63156c321865d668
+size 411697230

model/ckpt/nar-llama-8/fp32.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fed74021797c5eafee3f88e3d79b16e51eb086f47ca2700f56f993eb90f7d3a6
+size 441075976

model/config.split.yaml

@@ -0,0 +1,165 @@
+sample_rate: 24_000
+audio_backend: "vocos"
+experimental: True
+
+models:
+- name: "ar"
+  size: "full"
+  resp_levels: 1
+  prom_levels: 1
+  tasks: 8
+  langs: 2
+  tones: 1
+  arch_type: llama
+  training: False
+  version: 5
+  attention: auto
+  dropout: 0.1
+  audio_embedding_sums: False
+  experimental: False
+  interleave: False
+  loss_factors:
+    text: 0.01
+    prom: 0.5
+    resp: 1.0
+  capabilities: ["ar"]
+- name: "nar"
+  size: "full"
+  resp_levels: 8
+  prom_levels: 8
+  tasks: 8
+  langs: 2
+  tones: 1
+  arch_type: llama
+  training: False
+  version: 5
+  attention: auto
+  dropout: 0.1
+  audio_embedding_sums: False
+  experimental: False
+  interleave: False
+  loss_factors:
+    text: 0.01
+    prom: 0.5
+    resp: 1.0
+  capabilities: ["nar"]
+
+hyperparameters:
+  autotune: False
+  autotune_params:
+    start_profile_step: 1
+    end_profile_step: 50
+    num_tuning_micro_batch_sizes: 8
+
+  batch_size: 16
+  gradient_accumulation_steps: 8
+  gradient_clipping: 1.0
+  warmup_steps: 250
+
+  optimizer: Prodigy
+  learning_rate: 1.0
+  torch_optimizer: True
+  
+  scheduler: "" # ScheduleFree
+  torch_scheduler: True
+
+evaluation:
+  batch_size: 16
+  frequency: 1000
+  size: 16
+  
+  steps: 500
+  ar_temperature: 0.95
+  nar_temperature: 0.25
+  load_disabled_engines: True
+
+trainer:
+  #no_logger: True
+  ddp: False
+  check_for_oom: False
+  iterations: 1_000_000
+  
+  save_tag: step
+  save_on_oom: True
+  save_on_quit: True
+  save_frequency: 500
+  export_on_save: True
+
+  keep_last_checkpoints: 8
+
+  aggressive_optimizations: False
+  load_disabled_engines: False
+  gradient_checkpointing: True
+
+  #load_state_dict: True
+  strict_loading: False
+  #load_tag: "9500"
+  #load_states: False
+  #restart_step_count: True
+  
+  gc_mode: None # "global_step"
+
+  weight_dtype: bfloat16
+  amp: True
+
+  backend: deepspeed
+  deepspeed:
+    inferencing: True
+    zero_optimization_level: 0
+    use_compression_training: False
+
+    amp: False
+
+  load_webui: False
+
+inference:
+  backend: deepspeed
+  audio_backend: "vocos"
+  normalize: False
+
+  weight_dtype: bfloat16
+  amp: True
+
+optimizations:
+  injects: False
+  replace: True
+
+  linear: False
+  embedding: False
+  optimizers: True
+
+  bitsandbytes: False
+  dadaptation: False
+  bitnet: False
+  fp8: False
+
+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: 6
+  cache: True
+
+  duration_range: [24.0, 32.0]
+
+  random_utterance: 1.0
+  max_prompts: 1
+  prompt_duration_range: [3.0, 9.0]
+  
+  max_resps: 1
+  p_resp_append: 0.25
+
+  sample_type: path # path # speaker
+
+  tasks_list: [ "tts" ] # , [ "tts", "tts-c", "ns", "sr", "tse", "cse", "nse", "tts"]
+
+  training: []
+  validation: []
+  noise: []

model/config.yaml

@@ -0,0 +1,145 @@
+sample_rate: 24_000
+audio_backend: "vocos"
+experimental: True
+
+models:
+- name: "ar+nar"
+  size: "full"
+  resp_levels: 8
+  prom_levels: 8
+  tasks: 8
+  langs: 2
+  tones: 1
+  arch_type: llama
+  training: False
+  version: 5
+  attention: auto
+  dropout: 0.1
+  audio_embedding_sums: False
+  experimental: False
+  interleave: False
+  loss_factors:
+    text: 0.01
+    prom: 0.5
+    resp: 1.0
+  capabilities: ["ar", "nar"]
+
+hyperparameters:
+  autotune: False
+  autotune_params:
+    start_profile_step: 1
+    end_profile_step: 50
+    num_tuning_micro_batch_sizes: 8
+
+  batch_size: 16
+  gradient_accumulation_steps: 8
+  gradient_clipping: 1.0
+  warmup_steps: 250
+
+  optimizer: Prodigy
+  learning_rate: 1.0
+  torch_optimizer: True
+  
+  scheduler: "" # ScheduleFree
+  torch_scheduler: True
+
+evaluation:
+  batch_size: 16
+  frequency: 1000
+  size: 16
+  
+  steps: 500
+  ar_temperature: 0.95
+  nar_temperature: 0.25
+  load_disabled_engines: True
+
+trainer:
+  #no_logger: True
+  ddp: False
+  check_for_oom: False
+  iterations: 1_000_000
+  
+  save_tag: step
+  save_on_oom: True
+  save_on_quit: True
+  save_frequency: 500
+  export_on_save: True
+
+  keep_last_checkpoints: 8
+
+  aggressive_optimizations: False
+  load_disabled_engines: False
+  gradient_checkpointing: True
+
+  #load_state_dict: True
+  strict_loading: False
+  #load_tag: "9500"
+  #load_states: False
+  #restart_step_count: True
+  
+  gc_mode: None # "global_step"
+
+  weight_dtype: bfloat16
+  amp: True
+
+  backend: deepspeed
+  deepspeed:
+    inferencing: True
+    zero_optimization_level: 0
+    use_compression_training: False
+
+    amp: False
+
+  load_webui: False
+
+inference:
+  backend: deepspeed
+  audio_backend: "vocos"
+  normalize: False
+
+  weight_dtype: bfloat16
+  amp: True
+
+optimizations:
+  injects: False
+  replace: True
+
+  linear: False
+  embedding: False
+  optimizers: True
+
+  bitsandbytes: False
+  dadaptation: False
+  bitnet: False
+  fp8: False
+
+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: 6
+  cache: True
+
+  duration_range: [24.0, 32.0]
+
+  random_utterance: 1.0
+  max_prompts: 1
+  prompt_duration_range: [3.0, 9.0]
+  
+  max_resps: 1
+  p_resp_append: 0.25
+
+  sample_type: path # path # speaker
+
+  tasks_list: [ "tts" ] # , [ "tts", "tts-c", "ns", "sr", "tse", "cse", "nse", "tts"]
+
+  training: []
+  validation: []
+  noise: []