Hugging Face's logo

ACE-Step
/
acestep-captioner

Text-to-Audio
Transformers
Safetensors
qwen2_5_omni
music
audio
Model card Files
xet
Community
Sayoyo commited on
Commit
8b13325
·
verified ·
1 Parent(s): 919960e

Upload folder using huggingface_hub

Browse files
Files changed (22) hide show
  1. .gitattributes +1 -0
  2. README.md +96 -3
  3. added_tokens.json +24 -0
  4. args.json +484 -0
  5. chat_template.jinja +7 -0
  6. config.json +633 -0
  7. file_list.txt +22 -0
  8. generation_config.json +8 -0
  9. merges.txt +0 -0
  10. model-00001-of-00005.safetensors +3 -0
  11. model-00002-of-00005.safetensors +3 -0
  12. model-00003-of-00005.safetensors +3 -0
  13. model-00004-of-00005.safetensors +3 -0
  14. model-00005-of-00005.safetensors +3 -0
  15. model.safetensors.index.json +0 -0
  16. preprocessor_config.json +31 -0
  17. special_tokens_map.json +38 -0
  18. spk_dict.pt +3 -0
  19. tokenizer.json +3 -0
  20. tokenizer_config.json +222 -0
  21. video_preprocessor_config.json +56 -0
  22. vocab.json +0 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
README.md CHANGED
@@ -1,3 +1,96 @@
1
- ---
2
- license: mit
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # ACE-Step Captioner
2
+
3
+ ## Description
4
+
5
+ ACE-Step Captioner is the annotation model used by **ACE-Step v1.5** for training data labeling. It is a professional-grade music captioning model that generates detailed, structured descriptions of audio content.
6
+
7
+ ### Performance
8
+
9
+ 🏆 **Accuracy surpasses Gemini Pro 2.5** in music description tasks
10
+
11
+ ### Key Features
12
+
13
+ - 🎼 **Musical Style Analysis** - Identifies genres, sub-genres, and stylistic influences
14
+ - 🎸 **Instrument Recognition** - Detects and describes 1000+ instrument types and combinations
15
+ - 🎭 **Structure & Progression** - Analyzes musical arrangement including intro, verse, chorus, bridge, climax, and outro
16
+ - 🔊 **Timbre Description** - Captures tonal qualities, textures, and sonic characteristics
17
+ - 📝 **Rich Vocabulary** - Supports 1000+ descriptive terms for comprehensive music annotation
18
+
19
+ ## Usage
20
+
21
+ The usage is the same as [Qwen2.5 Omni-7B](https://huggingface.co/Qwen/Qwen2.5-Omni-7B).
22
+
23
+ ### Prompt Format
24
+
25
+ Use the following prompt to caption audio:
26
+
27
+ ```
28
+ *Task* Describe this audio in detail
29
+ <audio>
30
+ ```
31
+
32
+ ### Output Format
33
+
34
+ The model generates natural language descriptions covering multiple aspects of the music.
35
+
36
+ ### Example Output
37
+
38
+ ```
39
+ A melancholic indie folk track featuring fingerpicked acoustic guitar
40
+ as the primary instrument. The song opens with a sparse, contemplative
41
+ intro before the vocals enter with a breathy, intimate delivery.
42
+ The arrangement gradually builds through the verse, adding subtle
43
+ string pads and a gentle kick drum. The chorus lifts with layered
44
+ harmonies and a warmer, fuller texture. The bridge introduces a
45
+ key change and emotional climax before returning to the stripped-down
46
+ acoustic arrangement for the outro.
47
+ ```
48
+
49
+ ## Descriptive Capabilities
50
+
51
+ ### Musical Styles (Examples)
52
+
53
+ | Category | Styles |
54
+ |----------|--------|
55
+ | **Electronic** | Ambient, Techno, House, Drum & Bass, Synthwave, IDM, Downtempo |
56
+ | **Rock** | Alternative, Indie, Post-Rock, Progressive, Psychedelic, Grunge |
57
+ | **Pop** | Synth-pop, Electropop, Dream Pop, Art Pop, Indie Pop |
58
+ | **Classical** | Orchestral, Chamber, Minimalist, Neo-Classical, Cinematic |
59
+ | **World** | Latin, African, Middle Eastern, Asian Traditional, Celtic |
60
+ | **Jazz** | Fusion, Smooth, Bebop, Modal, Free Jazz |
61
+ | **Hip-Hop** | Trap, Boom Bap, Lo-fi, Instrumental, Cloud Rap |
62
+
63
+ ### Instruments (1000+ Supported)
64
+
65
+ | Category | Examples |
66
+ |----------|----------|
67
+ | **Strings** | Acoustic Guitar, Electric Guitar, Violin, Cello, Bass, Harp, Mandolin |
68
+ | **Keys** | Piano, Synthesizer, Organ, Rhodes, Wurlitzer, Mellotron |
69
+ | **Percussion** | Drums, Electronic Drums, Congas, Bongos, Timpani, Vibraphone |
70
+ | **Wind** | Saxophone, Trumpet, Flute, Clarinet, Oboe, French Horn |
71
+ | **Electronic** | Synth Bass, Pad, Lead, Arpeggiator, Sampler, 808, 303 |
72
+
73
+ ### Structure Analysis
74
+
75
+ - **Intro / Outro** - Opening and closing sections
76
+ - **Verse / Pre-Chorus / Chorus** - Main song structure
77
+ - **Bridge / Break** - Transitional sections
78
+ - **Build-up / Drop / Climax** - Dynamic progression
79
+ - **Interlude / Solo** - Instrumental passages
80
+
81
+ ### Timbre Descriptions
82
+
83
+ | Dimension | Descriptors |
84
+ |-----------|-------------|
85
+ | **Texture** | Warm, Bright, Dark, Crisp, Muddy, Clean, Distorted, Saturated |
86
+ | **Space** | Reverberant, Dry, Spacious, Intimate, Cavernous, Tight |
87
+ | **Dynamics** | Punchy, Soft, Aggressive, Gentle, Compressed, Dynamic |
88
+ | **Character** | Ethereal, Gritty, Smooth, Raw, Polished, Organic, Synthetic |
89
+
90
+ ## Use Cases
91
+
92
+ - **Music AI Training** - Generate high-quality captions for music generation models
93
+ - **Music Information Retrieval** - Create searchable metadata for audio databases
94
+ - **Content Moderation** - Analyze and categorize music content
95
+ - **Music Education** - Provide detailed analysis for learning purposes
96
+ - **Audio Production** - Document and describe sound design elements
added_tokens.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</tool_call>": 151658,
3
+ "<tool_call>": 151657,
4
+ "<|AUDIO|>": 151646,
5
+ "<|IMAGE|>": 151655,
6
+ "<|VIDEO|>": 151656,
7
+ "<|audio_bos|>": 151647,
8
+ "<|audio_eos|>": 151648,
9
+ "<|box_end|>": 151649,
10
+ "<|endoftext|>": 151643,
11
+ "<|file_sep|>": 151664,
12
+ "<|fim_middle|>": 151660,
13
+ "<|fim_pad|>": 151662,
14
+ "<|fim_prefix|>": 151659,
15
+ "<|fim_suffix|>": 151661,
16
+ "<|im_end|>": 151645,
17
+ "<|im_start|>": 151644,
18
+ "<|quad_end|>": 151651,
19
+ "<|quad_start|>": 151650,
20
+ "<|repo_name|>": 151663,
21
+ "<|vision_bos|>": 151652,
22
+ "<|vision_eos|>": 151653,
23
+ "<|vision_pad|>": 151654
24
+ }
args.json ADDED
@@ -0,0 +1,484 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "output_dir": "/root/data/repo/keylxiao/audio_caption_gemma_3n_ft/swift_full_output_0813/audio_caption_describer_full_length_grpo",
3
+ "overwrite_output_dir": false,
4
+ "do_train": false,
5
+ "do_eval": false,
6
+ "do_predict": false,
7
+ "eval_strategy": "no",
8
+ "prediction_loss_only": false,
9
+ "per_device_train_batch_size": 1,
10
+ "per_device_eval_batch_size": 1,
11
+ "per_gpu_train_batch_size": null,
12
+ "per_gpu_eval_batch_size": null,
13
+ "gradient_accumulation_steps": 1,
14
+ "eval_accumulation_steps": null,
15
+ "eval_delay": 0,
16
+ "torch_empty_cache_steps": null,
17
+ "learning_rate": 1e-05,
18
+ "weight_decay": 0.1,
19
+ "adam_beta1": 0.9,
20
+ "adam_beta2": 0.95,
21
+ "adam_epsilon": 1e-08,
22
+ "max_grad_norm": 1.0,
23
+ "num_train_epochs": 3.0,
24
+ "max_steps": 10000,
25
+ "lr_scheduler_type": "cosine",
26
+ "lr_scheduler_kwargs": null,
27
+ "warmup_ratio": 0.01,
28
+ "warmup_steps": 0,
29
+ "log_level": "passive",
30
+ "log_level_replica": "warning",
31
+ "log_on_each_node": true,
32
+ "logging_dir": "/root/data/repo/keylxiao/audio_caption_gemma_3n_ft/swift_full_output_0813/audio_caption_describer_full_length_grpo/runs",
33
+ "logging_strategy": "steps",
34
+ "logging_first_step": true,
35
+ "logging_steps": 5,
36
+ "logging_nan_inf_filter": true,
37
+ "save_strategy": "steps",
38
+ "save_steps": 500.0,
39
+ "save_total_limit": 5,
40
+ "save_safetensors": true,
41
+ "save_on_each_node": false,
42
+ "save_only_model": true,
43
+ "restore_callback_states_from_checkpoint": false,
44
+ "no_cuda": false,
45
+ "use_cpu": false,
46
+ "use_mps_device": false,
47
+ "seed": 42,
48
+ "data_seed": 42,
49
+ "jit_mode_eval": false,
50
+ "use_ipex": false,
51
+ "bf16": true,
52
+ "fp16": false,
53
+ "fp16_opt_level": "O1",
54
+ "half_precision_backend": "auto",
55
+ "bf16_full_eval": false,
56
+ "fp16_full_eval": false,
57
+ "tf32": null,
58
+ "local_rank": 0,
59
+ "ddp_backend": null,
60
+ "tpu_num_cores": null,
61
+ "tpu_metrics_debug": false,
62
+ "debug": null,
63
+ "dataloader_drop_last": false,
64
+ "eval_steps": 500.0,
65
+ "dataloader_num_workers": 8,
66
+ "dataloader_prefetch_factor": null,
67
+ "past_index": -1,
68
+ "run_name": "/root/data/repo/keylxiao/audio_caption_gemma_3n_ft/swift_full_output_0813/audio_caption_describer_full_length_grpo",
69
+ "disable_tqdm": null,
70
+ "remove_unused_columns": false,
71
+ "label_names": null,
72
+ "load_best_model_at_end": false,
73
+ "metric_for_best_model": "reward",
74
+ "greater_is_better": true,
75
+ "ignore_data_skip": false,
76
+ "fsdp": "",
77
+ "fsdp_min_num_params": 0,
78
+ "fsdp_config": null,
79
+ "fsdp_transformer_layer_cls_to_wrap": null,
80
+ "accelerator_config": {
81
+ "dispatch_batches": false
82
+ },
83
+ "deepspeed": {
84
+ "fp16": {
85
+ "enabled": "auto",
86
+ "loss_scale": 0,
87
+ "loss_scale_window": 1000,
88
+ "initial_scale_power": 16,
89
+ "hysteresis": 2,
90
+ "min_loss_scale": 1
91
+ },
92
+ "bf16": {
93
+ "enabled": "auto"
94
+ },
95
+ "zero_optimization": {
96
+ "stage": 2,
97
+ "offload_optimizer": {
98
+ "device": "none",
99
+ "pin_memory": true
100
+ },
101
+ "allgather_partitions": true,
102
+ "allgather_bucket_size": 200000000.0,
103
+ "overlap_comm": false,
104
+ "reduce_scatter": true,
105
+ "reduce_bucket_size": 200000000.0,
106
+ "contiguous_gradients": true
107
+ },
108
+ "gradient_accumulation_steps": "auto",
109
+ "gradient_clipping": "auto",
110
+ "steps_per_print": 2000,
111
+ "train_batch_size": "auto",
112
+ "train_micro_batch_size_per_gpu": "auto",
113
+ "wall_clock_breakdown": false
114
+ },
115
+ "label_smoothing_factor": 0.0,
116
+ "optim": "adamw_torch",
117
+ "optim_args": null,
118
+ "adafactor": false,
119
+ "group_by_length": false,
120
+ "length_column_name": "length",
121
+ "report_to": [
122
+ "tensorboard"
123
+ ],
124
+ "ddp_find_unused_parameters": null,
125
+ "ddp_bucket_cap_mb": null,
126
+ "ddp_broadcast_buffers": null,
127
+ "dataloader_pin_memory": true,
128
+ "dataloader_persistent_workers": false,
129
+ "skip_memory_metrics": true,
130
+ "use_legacy_prediction_loop": false,
131
+ "push_to_hub": false,
132
+ "resume_from_checkpoint": null,
133
+ "hub_model_id": null,
134
+ "hub_strategy": "every_save",
135
+ "hub_token": null,
136
+ "hub_private_repo": null,
137
+ "hub_always_push": false,
138
+ "hub_revision": null,
139
+ "gradient_checkpointing": true,
140
+ "gradient_checkpointing_kwargs": null,
141
+ "include_inputs_for_metrics": false,
142
+ "include_for_metrics": [],
143
+ "eval_do_concat_batches": true,
144
+ "fp16_backend": "auto",
145
+ "push_to_hub_model_id": null,
146
+ "push_to_hub_organization": null,
147
+ "push_to_hub_token": null,
148
+ "mp_parameters": "",
149
+ "auto_find_batch_size": false,
150
+ "full_determinism": false,
151
+ "torchdynamo": null,
152
+ "ray_scope": "last",
153
+ "ddp_timeout": 18000000,
154
+ "torch_compile": false,
155
+ "torch_compile_backend": null,
156
+ "torch_compile_mode": null,
157
+ "include_tokens_per_second": false,
158
+ "include_num_input_tokens_seen": false,
159
+ "neftune_noise_alpha": 0.0,
160
+ "optim_target_modules": null,
161
+ "batch_eval_metrics": false,
162
+ "eval_on_start": false,
163
+ "use_liger_kernel": true,
164
+ "liger_kernel_config": null,
165
+ "eval_use_gather_object": false,
166
+ "average_tokens_across_devices": false,
167
+ "sortish_sampler": false,
168
+ "predict_with_generate": false,
169
+ "generation_max_length": null,
170
+ "generation_num_beams": null,
171
+ "generation_config": null,
172
+ "vit_gradient_checkpointing": null,
173
+ "check_model": true,
174
+ "acc_strategy": "token",
175
+ "train_dataloader_shuffle": true,
176
+ "max_epochs": null,
177
+ "aligner_lr": null,
178
+ "vit_lr": null,
179
+ "optimizer": null,
180
+ "use_logits_to_keep": null,
181
+ "channels": null,
182
+ "ds3_gather_for_generation": true,
183
+ "resume_only_model": false,
184
+ "metric_warmup_step": 0,
185
+ "fsdp_num": 1,
186
+ "acc_steps": 1,
187
+ "eval_use_evalscope": false,
188
+ "eval_dataset": [],
189
+ "eval_dataset_args": null,
190
+ "eval_limit": null,
191
+ "eval_generation_config": null,
192
+ "model": "/root/data/repo/keylxiao/audio_caption_gemma_3n_ft/swift_full_output_0813/audio_caption_describer_full_length/checkpoint-60000",
193
+ "model_type": "qwen2_5_omni",
194
+ "model_revision": null,
195
+ "task_type": "causal_lm",
196
+ "torch_dtype": "bfloat16",
197
+ "attn_impl": "flash_attn",
198
+ "new_special_tokens": [],
199
+ "num_labels": null,
200
+ "problem_type": null,
201
+ "rope_scaling": null,
202
+ "device_map": null,
203
+ "max_memory": {},
204
+ "local_repo_path": null,
205
+ "init_strategy": null,
206
+ "template": "qwen2_5_omni",
207
+ "system": null,
208
+ "max_length": 16384,
209
+ "truncation_strategy": "left",
210
+ "max_pixels": null,
211
+ "agent_template": null,
212
+ "norm_bbox": null,
213
+ "use_chat_template": true,
214
+ "padding_free": false,
215
+ "padding_side": "right",
216
+ "loss_scale": "last_round",
217
+ "sequence_parallel_size": 1,
218
+ "response_prefix": null,
219
+ "template_backend": "swift",
220
+ "dataset": [
221
+ "/root/data/repo/gongjunmin/acestep_v15_train/data/ace_step_v15/gemini_captions_qwen3_30b_moe/omini_audio_caption_caption_model_train_data_5M.parquet"
222
+ ],
223
+ "val_dataset": [],
224
+ "split_dataset_ratio": 0.0,
225
+ "dataset_num_proc": 1,
226
+ "load_from_cache_file": true,
227
+ "dataset_shuffle": true,
228
+ "val_dataset_shuffle": false,
229
+ "streaming": false,
230
+ "interleave_prob": null,
231
+ "stopping_strategy": "first_exhausted",
232
+ "shuffle_buffer_size": 1000,
233
+ "download_mode": "reuse_dataset_if_exists",
234
+ "columns": {},
235
+ "strict": false,
236
+ "model_name": null,
237
+ "model_author": null,
238
+ "custom_dataset_info": [],
239
+ "quant_method": null,
240
+ "quant_bits": null,
241
+ "hqq_axis": null,
242
+ "bnb_4bit_compute_dtype": "bfloat16",
243
+ "bnb_4bit_quant_type": "nf4",
244
+ "bnb_4bit_use_double_quant": true,
245
+ "bnb_4bit_quant_storage": null,
246
+ "max_new_tokens": 512,
247
+ "temperature": 0.7,
248
+ "top_k": 50,
249
+ "top_p": 0.99,
250
+ "repetition_penalty": 1.0,
251
+ "num_beams": 1,
252
+ "stream": false,
253
+ "stop_words": [],
254
+ "logprobs": false,
255
+ "top_logprobs": null,
256
+ "ckpt_dir": "/root/data/repo/keylxiao/audio_caption_gemma_3n_ft/swift_full_output_0813/audio_caption_describer_full_length/checkpoint-60000",
257
+ "lora_modules": [],
258
+ "tuner_backend": "peft",
259
+ "train_type": "full",
260
+ "adapters": [],
261
+ "external_plugins": [
262
+ "/root/data/repo/keylxiao/audio_caption_gemma_3n_ft/ms-swift/examples/train/grpo/plugin/plugin.py"
263
+ ],
264
+ "model_kwargs": {},
265
+ "load_args": false,
266
+ "load_data_args": false,
267
+ "packing": false,
268
+ "custom_register_path": [],
269
+ "use_hf": false,
270
+ "ignore_args_error": true,
271
+ "use_swift_lora": false,
272
+ "freeze_parameters": [
273
+ "thinker.audio_tower",
274
+ "thinker.visual",
275
+ "thinker.audio_tower.proj",
276
+ "thinker.visual.merger",
277
+ "talker",
278
+ "token2wav"
279
+ ],
280
+ "freeze_parameters_regex": null,
281
+ "freeze_parameters_ratio": 0.0,
282
+ "trainable_parameters": [],
283
+ "trainable_parameters_regex": null,
284
+ "freeze_llm": false,
285
+ "freeze_vit": true,
286
+ "freeze_aligner": true,
287
+ "target_modules": [
288
+ "all-linear"
289
+ ],
290
+ "target_regex": null,
291
+ "modules_to_save": [],
292
+ "lora_rank": 8,
293
+ "lora_alpha": 32,
294
+ "lora_dropout": 0.05,
295
+ "lora_bias": "none",
296
+ "lora_dtype": null,
297
+ "lorap_lr_ratio": null,
298
+ "use_rslora": false,
299
+ "use_dora": false,
300
+ "lora_ga_batch_size": 2,
301
+ "lora_ga_iters": 2,
302
+ "lora_ga_max_length": 1024,
303
+ "lora_ga_direction": "ArB2r",
304
+ "lora_ga_scale": "stable",
305
+ "lora_ga_stable_gamma": 16,
306
+ "init_weights": true,
307
+ "fourier_n_frequency": 2000,
308
+ "fourier_scaling": 300.0,
309
+ "boft_block_size": 4,
310
+ "boft_block_num": 0,
311
+ "boft_n_butterfly_factor": 1,
312
+ "boft_dropout": 0.0,
313
+ "vera_rank": 256,
314
+ "vera_projection_prng_key": 0,
315
+ "vera_dropout": 0.0,
316
+ "vera_d_initial": 0.1,
317
+ "adapter_act": "gelu",
318
+ "adapter_length": 128,
319
+ "use_galore": false,
320
+ "galore_target_modules": null,
321
+ "galore_rank": 128,
322
+ "galore_update_proj_gap": 50,
323
+ "galore_scale": 1.0,
324
+ "galore_proj_type": "std",
325
+ "galore_optim_per_parameter": false,
326
+ "galore_with_embedding": false,
327
+ "galore_quantization": false,
328
+ "galore_proj_quant": false,
329
+ "galore_proj_bits": 4,
330
+ "galore_proj_group_size": 256,
331
+ "galore_cos_threshold": 0.4,
332
+ "galore_gamma_proj": 2,
333
+ "galore_queue_size": 5,
334
+ "adalora_target_r": 8,
335
+ "adalora_init_r": 12,
336
+ "adalora_tinit": 0,
337
+ "adalora_tfinal": 0,
338
+ "adalora_deltaT": 1,
339
+ "adalora_beta1": 0.85,
340
+ "adalora_beta2": 0.85,
341
+ "adalora_orth_reg_weight": 0.5,
342
+ "llamapro_num_new_blocks": 4,
343
+ "llamapro_num_groups": null,
344
+ "lisa_activated_layers": 0,
345
+ "lisa_step_interval": 20,
346
+ "reft_layer_key": null,
347
+ "reft_layers": null,
348
+ "reft_rank": 4,
349
+ "reft_intervention_type": "LoreftIntervention",
350
+ "reft_args": null,
351
+ "swanlab_token": null,
352
+ "swanlab_project": null,
353
+ "swanlab_workspace": null,
354
+ "swanlab_exp_name": null,
355
+ "swanlab_lark_webhook_url": null,
356
+ "swanlab_lark_secret": null,
357
+ "swanlab_mode": "cloud",
358
+ "add_version": false,
359
+ "create_checkpoint_symlink": false,
360
+ "lazy_tokenize": true,
361
+ "loss_type": "grpo",
362
+ "metric": null,
363
+ "zero_hpz_partition_size": null,
364
+ "sft_alpha": 0,
365
+ "reward_model": [
366
+ "/root/data/repo/keylxiao/audio_caption_gemma_3n_ft/swift_full_output_0813/audio_caption_reward_full_length_rlhf/checkpoint-20000"
367
+ ],
368
+ "reward_adapters": [],
369
+ "reward_model_type": null,
370
+ "reward_model_revision": null,
371
+ "num_ppo_epochs": 4,
372
+ "whiten_rewards": false,
373
+ "kl_coef": 0.05,
374
+ "cliprange": 0.2,
375
+ "vf_coef": 0.1,
376
+ "cliprange_value": 0.2,
377
+ "gamma": 1.0,
378
+ "lam": 0.95,
379
+ "num_mini_batches": 1,
380
+ "local_rollout_forward_batch_size": 64,
381
+ "num_sample_generations": 10,
382
+ "response_length": 512,
383
+ "missing_eos_penalty": null,
384
+ "vllm_gpu_memory_utilization": 0.85,
385
+ "vllm_tensor_parallel_size": 2,
386
+ "vllm_pipeline_parallel_size": 1,
387
+ "vllm_max_num_seqs": 256,
388
+ "vllm_max_model_len": 16384,
389
+ "vllm_disable_custom_all_reduce": true,
390
+ "vllm_enforce_eager": false,
391
+ "vllm_limit_mm_per_prompt": {},
392
+ "vllm_max_lora_rank": 16,
393
+ "vllm_enable_prefix_caching": true,
394
+ "vllm_use_async_engine": false,
395
+ "vllm_quantization": null,
396
+ "vllm_data_parallel_size": 1,
397
+ "gpu_memory_utilization": null,
398
+ "tensor_parallel_size": null,
399
+ "max_model_len": null,
400
+ "limit_mm_per_prompt": null,
401
+ "data_parallel_size": null,
402
+ "use_async_engine": null,
403
+ "epsilon": 0.2,
404
+ "epsilon_high": null,
405
+ "delta": null,
406
+ "vllm_mode": "colocate",
407
+ "vllm_server_base_url": null,
408
+ "vllm_server_host": null,
409
+ "vllm_server_port": [
410
+ 8000
411
+ ],
412
+ "vllm_server_timeout": 240.0,
413
+ "cosine_min_len_value_wrong": -0.5,
414
+ "cosine_max_len_value_wrong": 0.0,
415
+ "cosine_min_len_value_correct": 1.0,
416
+ "cosine_max_len_value_correct": 0.5,
417
+ "cosine_max_len": null,
418
+ "repetition_n_grams": 3,
419
+ "repetition_max_penalty": -1.0,
420
+ "reward_model_plugin": [
421
+ "audio_caption_rm"
422
+ ],
423
+ "sync_ref_model": false,
424
+ "ref_model_sync_steps": 512,
425
+ "ref_model_mixup_alpha": 0.6,
426
+ "async_generate": false,
427
+ "sleep_level": 1,
428
+ "move_model_batches": 8,
429
+ "offload_optimizer": true,
430
+ "offload_model": true,
431
+ "gc_collect_after_offload": false,
432
+ "multi_turn_func": null,
433
+ "multi_turn_scheduler": null,
434
+ "max_turns": null,
435
+ "completion_length_limit_scope": "per_round",
436
+ "dynamic_sample": false,
437
+ "max_resample_times": 3,
438
+ "overlong_filter": false,
439
+ "soft_max_length": null,
440
+ "soft_cache_length": null,
441
+ "scale_rewards": true,
442
+ "log_entropy": false,
443
+ "top_entropy_quantile": 1.0,
444
+ "wandb_log_unique_prompts": null,
445
+ "generation_batch_size": null,
446
+ "steps_per_generation": null,
447
+ "num_generations": 8,
448
+ "reward_funcs": [],
449
+ "reward_weights": [
450
+ 1.0
451
+ ],
452
+ "log_completions": true,
453
+ "use_vllm": true,
454
+ "num_iterations": 1,
455
+ "teacher_model": null,
456
+ "teacher_adapters": [],
457
+ "teacher_model_type": null,
458
+ "teacher_model_revision": null,
459
+ "rlhf_type": "grpo",
460
+ "ref_model": "/root/data/repo/keylxiao/audio_caption_gemma_3n_ft/swift_full_output_0813/audio_caption_describer_full_length/checkpoint-60000",
461
+ "ref_model_type": "qwen2_5_omni",
462
+ "ref_model_revision": null,
463
+ "beta": 0.04,
464
+ "label_smoothing": 0,
465
+ "max_completion_length": 512,
466
+ "rpo_alpha": 1.0,
467
+ "cpo_alpha": 1.0,
468
+ "simpo_gamma": 1,
469
+ "desirable_weight": 1.0,
470
+ "undesirable_weight": 1.0,
471
+ "center_rewards_coefficient": null,
472
+ "lmbda": 0.5,
473
+ "seq_kd": false,
474
+ "rank": 0,
475
+ "global_world_size": 8,
476
+ "local_world_size": 8,
477
+ "model_suffix": "checkpoint-60000",
478
+ "model_info": "ModelInfo(model_type='qwen2_5_omni', model_dir='/root/data/repo/keylxiao/audio_caption_gemma_3n_ft/swift_full_output_0813/audio_caption_describer_full_length/checkpoint-60000', torch_dtype=torch.bfloat16, max_model_len=32768, quant_method=None, quant_bits=None, rope_scaling={'mrope_section': [16, 24, 24], 'rope_type': 'default', 'type': 'default'}, is_moe_model=False, config=None, task_type='causal_lm', num_labels=None)",
479
+ "model_meta": "ModelMeta(model_type='qwen2_5_omni', model_groups=[ModelGroup(models=[Model(ms_model_id='Qwen/Qwen2.5-Omni-3B', hf_model_id='Qwen/Qwen2.5-Omni-3B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='Qwen/Qwen2.5-Omni-7B', hf_model_id='Qwen/Qwen2.5-Omni-7B', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[])], template='qwen2_5_omni', get_function=<function get_model_tokenizer_qwen2_5_omni at 0x7f90f9821510>, model_arch='qwen2_5_omni', architectures=['Qwen2_5OmniModel', 'Qwen2_5OmniForConditionalGeneration'], additional_saved_files=['spk_dict.pt'], torch_dtype=None, is_multimodal=True, is_reward=True, task_type=None, ignore_patterns=[], requires=['transformers>=4.50', 'soundfile', 'qwen_omni_utils', 'decord'], tags=['vision', 'video', 'audio'])",
480
+ "model_dir": "/root/data/repo/keylxiao/audio_caption_gemma_3n_ft/swift_full_output_0813/audio_caption_describer_full_length/checkpoint-60000",
481
+ "hub": "<class 'swift.hub.hub.MSHub'>",
482
+ "evaluation_strategy": "steps",
483
+ "training_args": "GRPOConfig(output_dir='/root/data/repo/keylxiao/audio_caption_gemma_3n_ft/swift_full_output_0813/audio_caption_describer_full_length_grpo', overwrite_output_dir=False, do_train=False, do_eval=False, do_predict=False, eval_strategy=<IntervalStrategy.NO: 'no'>, prediction_loss_only=False, per_device_train_batch_size=1, per_device_eval_batch_size=1, per_gpu_train_batch_size=None, per_gpu_eval_batch_size=None, gradient_accumulation_steps=1, eval_accumulation_steps=None, eval_delay=0, torch_empty_cache_steps=None, learning_rate=1e-05, weight_decay=0.1, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, max_grad_norm=1.0, num_train_epochs=3.0, max_steps=10000, lr_scheduler_type=<SchedulerType.COSINE: 'cosine'>, lr_scheduler_kwargs=None, warmup_ratio=0.01, warmup_steps=0, log_level='passive', log_level_replica='warning', log_on_each_node=True, logging_dir='/root/data/repo/keylxiao/audio_caption_gemma_3n_ft/swift_full_output_0813/audio_caption_describer_full_length_grpo/runs', logging_strategy=<IntervalStrategy.STEPS: 'steps'>, logging_first_step=True, logging_steps=5, logging_nan_inf_filter=True, save_strategy=<SaveStrategy.STEPS: 'steps'>, save_steps=500, save_total_limit=5, save_safetensors=True, save_on_each_node=False, save_only_model=True, restore_callback_states_from_checkpoint=False, no_cuda=False, use_cpu=False, use_mps_device=False, seed=42, data_seed=42, jit_mode_eval=False, use_ipex=False, bf16=True, fp16=False, fp16_opt_level='O1', half_precision_backend='auto', bf16_full_eval=False, fp16_full_eval=False, tf32=None, local_rank=0, ddp_backend=None, tpu_num_cores=None, tpu_metrics_debug=False, debug=[], dataloader_drop_last=True, eval_steps=500.0, dataloader_num_workers=8, dataloader_prefetch_factor=10, past_index=-1, run_name='/root/data/repo/keylxiao/audio_caption_gemma_3n_ft/swift_full_output_0813/audio_caption_describer_full_length_grpo', disable_tqdm=False, remove_unused_columns=False, label_names=None, load_best_model_at_end=False, metric_for_best_model='reward', greater_is_better=True, ignore_data_skip=False, fsdp=[], fsdp_min_num_params=0, fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, fsdp_transformer_layer_cls_to_wrap=None, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), deepspeed={'fp16': {'enabled': 'auto', 'loss_scale': 0, 'loss_scale_window': 1000, 'initial_scale_power': 16, 'hysteresis': 2, 'min_loss_scale': 1}, 'bf16': {'enabled': 'auto'}, 'zero_optimization': {'stage': 2, 'offload_optimizer': {'device': 'none', 'pin_memory': True}, 'allgather_partitions': True, 'allgather_bucket_size': 200000000.0, 'overlap_comm': False, 'reduce_scatter': True, 'reduce_bucket_size': 200000000.0, 'contiguous_gradients': True}, 'gradient_accumulation_steps': 'auto', 'gradient_clipping': 'auto', 'steps_per_print': 2000, 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'wall_clock_breakdown': False}, label_smoothing_factor=0.0, optim=<OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, optim_args=None, adafactor=False, group_by_length=False, length_column_name='length', report_to=['tensorboard'], ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, dataloader_pin_memory=True, dataloader_persistent_workers=False, skip_memory_metrics=True, use_legacy_prediction_loop=False, push_to_hub=False, resume_from_checkpoint=None, hub_model_id=None, hub_strategy=<HubStrategy.EVERY_SAVE: 'every_save'>, hub_token=None, hub_private_repo=None, hub_always_push=False, hub_revision=None, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, include_inputs_for_metrics=False, include_for_metrics=[], eval_do_concat_batches=True, fp16_backend='auto', push_to_hub_model_id=None, push_to_hub_organization=None, push_to_hub_token=None, mp_parameters='', auto_find_batch_size=False, full_determinism=False, torchdynamo=None, ray_scope='last', ddp_timeout=18000000, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, include_tokens_per_second=None, include_num_input_tokens_seen=None, neftune_noise_alpha=0.0, optim_target_modules=None, batch_eval_metrics=False, eval_on_start=False, use_liger_kernel=True, liger_kernel_config=None, eval_use_gather_object=False, average_tokens_across_devices=None, model_init_kwargs=None, disable_dropout=False, max_prompt_length=512, num_generations=8, max_completion_length=512, ds3_gather_for_generation=True, shuffle_dataset=True, generation_batch_size=8, steps_per_generation=1, temperature=0.7, top_p=0.99, top_k=50, min_p=None, generation_kwargs=None, repetition_penalty=1.0, cache_implementation=None, use_vllm=True, vllm_server_base_url=None, vllm_mode='colocate', vllm_guided_decoding_regex=None, vllm_server_host=None, vllm_server_port=[8000], vllm_server_timeout=240.0, vllm_gpu_memory_utilization=0.85, vllm_tensor_parallel_size=2, beta=0.04, num_iterations=1, epsilon=0.2, delta=None, epsilon_high=None, reward_weights=[1.0], scale_rewards=True, loss_type='grpo', mask_truncated_completions=False, sync_ref_model=False, ref_model_mixup_alpha=0.6, ref_model_sync_steps=512, use_liger_loss=False, log_completions=True, num_completions_to_print=None, wandb_log_unique_prompts=None, vit_gradient_checkpointing=True, check_model=True, acc_strategy='token', train_dataloader_shuffle=True, max_epochs=None, aligner_lr=None, vit_lr=None, optimizer=None, use_logits_to_keep=None, channels=None, resume_only_model=False, metric_warmup_step=0, fsdp_num=1, acc_steps=1, eval_use_evalscope=False, eval_dataset=[], eval_dataset_args=None, eval_limit=None, eval_generation_config=None, sft_alpha=0, train_type='full', local_repo_path=None, galore_config=None, vllm_pipeline_parallel_size=1, vllm_max_num_seqs=256, vllm_max_model_len=16384, vllm_disable_custom_all_reduce=True, vllm_enforce_eager=False, vllm_limit_mm_per_prompt={}, vllm_max_lora_rank=16, vllm_enable_prefix_caching=True, vllm_use_async_engine=False, vllm_quantization=None, vllm_data_parallel_size=1, gpu_memory_utilization=None, tensor_parallel_size=None, max_model_len=None, limit_mm_per_prompt=None, data_parallel_size=None, use_async_engine=None, cosine_min_len_value_wrong=-0.5, cosine_max_len_value_wrong=0.0, cosine_min_len_value_correct=1.0, cosine_max_len_value_correct=0.5, cosine_max_len=512, repetition_n_grams=3, repetition_max_penalty=-1.0, reward_model=['/root/data/repo/keylxiao/audio_caption_gemma_3n_ft/swift_full_output_0813/audio_caption_reward_full_length_rlhf/checkpoint-20000'], reward_model_plugin=['audio_caption_rm'], async_generate=False, sleep_level=1, move_model_batches=8, offload_optimizer=True, offload_model=True, gc_collect_after_offload=False, multi_turn_func=None, multi_turn_scheduler=None, max_turns=None, completion_length_limit_scope='per_round', dynamic_sample=False, max_resample_times=3, overlong_filter=False, soft_max_length=None, soft_cache_length=None, log_entropy=False, top_entropy_quantile=1.0, dataset_shuffle=True, stop_words=[])"
484
+ }
chat_template.jinja ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {% set audio_count = namespace(value=0) %}{% set image_count = namespace(value=0) %}{% set video_count = namespace(value=0) %}{% for message in messages %}{% if loop.first and message['role'] != 'system' %}<|im_start|>system
2
+ You are a helpful assistant.<|im_end|>
3
+ {% endif %}<|im_start|>{{ message['role'] }}
4
+ {% if message['content'] is string %}{{ message['content'] }}<|im_end|>
5
+ {% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_bos|><|IMAGE|><|vision_eos|>{% elif content['type'] == 'audio' or 'audio' in content or 'audio_url' in content %}{% set audio_count.value = audio_count.value + 1 %}{% if add_audio_id %}Audio {{ audio_count.value }}: {% endif %}<|audio_bos|><|AUDIO|><|audio_eos|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_bos|><|VIDEO|><|vision_eos|>{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|>
6
+ {% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant
7
+ {% endif %}
config.json ADDED
@@ -0,0 +1,633 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen2_5OmniForConditionalGeneration"
4
+ ],
5
+ "enable_audio_output": true,
6
+ "enable_talker": true,
7
+ "hidden_size": 3584,
8
+ "keys_to_ignore_at_inference": [
9
+ "past_key_values",
10
+ "hidden_states",
11
+ "attention_mask",
12
+ "hidden_states",
13
+ "attention_mask",
14
+ "hidden_states",
15
+ "attention_mask"
16
+ ],
17
+ "model_type": "qwen2_5_omni",
18
+ "pad_token_id": 151643,
19
+ "talker_config": {
20
+ "_name_or_path": "Qwen2.5-Omni-7B/talker",
21
+ "architectures": [
22
+ "Qwen2OmniTalkerForConditionalGeneration"
23
+ ],
24
+ "attention_dropout": 0.0,
25
+ "audio_end_token_id": 151648,
26
+ "audio_start_token_id": 151647,
27
+ "audio_token_index": 151646,
28
+ "embedding_size": 3584,
29
+ "head_dim": 128,
30
+ "hidden_act": "silu",
31
+ "hidden_size": 896,
32
+ "image_token_index": 151655,
33
+ "init_std": 0.02,
34
+ "initializer_range": 0.02,
35
+ "intermediate_size": 18944,
36
+ "layer_types": [
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention",
45
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention",
48
+ "full_attention",
49
+ "full_attention",
50
+ "full_attention",
51
+ "full_attention",
52
+ "full_attention",
53
+ "full_attention",
54
+ "full_attention",
55
+ "full_attention",
56
+ "full_attention",
57
+ "full_attention",
58
+ "full_attention",
59
+ "full_attention",
60
+ "full_attention"
61
+ ],
62
+ "max_position_embeddings": 32768,
63
+ "max_window_layers": 28,
64
+ "model_type": "qwen2_5_omni_talker",
65
+ "num_attention_heads": 12,
66
+ "num_hidden_layers": 24,
67
+ "num_key_value_heads": 4,
68
+ "position_id_per_seconds": 25,
69
+ "rms_norm_eps": 1e-06,
70
+ "rope_scaling": {
71
+ "mrope_section": [
72
+ 16,
73
+ 24,
74
+ 24
75
+ ],
76
+ "rope_type": "default",
77
+ "type": "default"
78
+ },
79
+ "rope_theta": 1000000.0,
80
+ "seconds_per_chunk": 2,
81
+ "sliding_window": null,
82
+ "spatial_merge_size": 2,
83
+ "torch_dtype": "bfloat16",
84
+ "tts_codec_end_token_id": 8294,
85
+ "tts_codec_mask_token_id": 8296,
86
+ "tts_codec_pad_token_id": 8292,
87
+ "tts_codec_start_token_id": 8293,
88
+ "tts_text_end_token_id": 151861,
89
+ "tts_text_pad_token_id": 151859,
90
+ "tts_text_start_token_id": 151860,
91
+ "use_cache": false,
92
+ "use_sliding_window": false,
93
+ "video_token_index": 151656,
94
+ "vision_end_token_id": 151653,
95
+ "vision_start_token_id": 151652,
96
+ "vocab_size": 8448
97
+ },
98
+ "thinker_config": {
99
+ "_name_or_path": "Qwen2.5-Omni-7B/thinker",
100
+ "architectures": [
101
+ "Qwen2OmniNaViTThinkerForConditionalGeneration"
102
+ ],
103
+ "audio_config": {
104
+ "_name_or_path": "",
105
+ "activation_dropout": 0.0,
106
+ "activation_function": "gelu",
107
+ "add_cross_attention": false,
108
+ "architectures": null,
109
+ "attention_dropout": 0.0,
110
+ "bad_words_ids": null,
111
+ "begin_suppress_tokens": null,
112
+ "bos_token_id": null,
113
+ "chunk_size_feed_forward": 0,
114
+ "cross_attention_hidden_size": null,
115
+ "d_model": 1280,
116
+ "decoder_start_token_id": null,
117
+ "diversity_penalty": 0.0,
118
+ "do_sample": false,
119
+ "dropout": 0.0,
120
+ "early_stopping": false,
121
+ "encoder_attention_heads": 20,
122
+ "encoder_ffn_dim": 5120,
123
+ "encoder_layerdrop": 0.0,
124
+ "encoder_layers": 32,
125
+ "encoder_no_repeat_ngram_size": 0,
126
+ "eos_token_id": null,
127
+ "exponential_decay_length_penalty": null,
128
+ "finetuning_task": null,
129
+ "forced_bos_token_id": null,
130
+ "forced_eos_token_id": null,
131
+ "id2label": {
132
+ "0": "LABEL_0",
133
+ "1": "LABEL_1"
134
+ },
135
+ "init_std": 0.02,
136
+ "initializer_range": 0.02,
137
+ "is_decoder": false,
138
+ "is_encoder_decoder": false,
139
+ "label2id": {
140
+ "LABEL_0": 0,
141
+ "LABEL_1": 1
142
+ },
143
+ "length_penalty": 1.0,
144
+ "max_length": 20,
145
+ "max_source_positions": 1500,
146
+ "min_length": 0,
147
+ "model_type": "qwen2_5_omni_audio_encoder",
148
+ "n_window": 100,
149
+ "no_repeat_ngram_size": 0,
150
+ "num_beam_groups": 1,
151
+ "num_beams": 1,
152
+ "num_hidden_layers": 32,
153
+ "num_mel_bins": 128,
154
+ "num_return_sequences": 1,
155
+ "output_attentions": false,
156
+ "output_dim": 3584,
157
+ "output_hidden_states": false,
158
+ "output_scores": false,
159
+ "pad_token_id": null,
160
+ "prefix": null,
161
+ "problem_type": null,
162
+ "pruned_heads": {},
163
+ "remove_invalid_values": false,
164
+ "repetition_penalty": 1.0,
165
+ "return_dict": true,
166
+ "return_dict_in_generate": false,
167
+ "scale_embedding": false,
168
+ "sep_token_id": null,
169
+ "suppress_tokens": null,
170
+ "task_specific_params": null,
171
+ "temperature": 1.0,
172
+ "tf_legacy_loss": false,
173
+ "tie_encoder_decoder": false,
174
+ "tie_word_embeddings": true,
175
+ "tokenizer_class": null,
176
+ "top_k": 50,
177
+ "top_p": 1.0,
178
+ "torch_dtype": null,
179
+ "torchscript": false,
180
+ "typical_p": 1.0,
181
+ "use_bfloat16": false
182
+ },
183
+ "audio_end_token_id": 151648,
184
+ "audio_start_token_id": 151647,
185
+ "audio_token_index": 151646,
186
+ "bos_token_id": 151644,
187
+ "eos_token_id": 151645,
188
+ "ignore_index": -100,
189
+ "image_token_index": 151655,
190
+ "init_std": 0.02,
191
+ "initializer_range": 0.02,
192
+ "model_type": "qwen2_5_omni_thinker",
193
+ "pad_token_id": 151643,
194
+ "position_id_per_seconds": 25,
195
+ "seconds_per_chunk": 2,
196
+ "text_config": {
197
+ "_name_or_path": "",
198
+ "add_cross_attention": false,
199
+ "architectures": null,
200
+ "attention_dropout": 0.0,
201
+ "bad_words_ids": null,
202
+ "begin_suppress_tokens": null,
203
+ "bos_token_id": null,
204
+ "chunk_size_feed_forward": 0,
205
+ "cross_attention_hidden_size": null,
206
+ "decoder_start_token_id": null,
207
+ "diversity_penalty": 0.0,
208
+ "do_sample": false,
209
+ "early_stopping": false,
210
+ "encoder_no_repeat_ngram_size": 0,
211
+ "eos_token_id": null,
212
+ "exponential_decay_length_penalty": null,
213
+ "finetuning_task": null,
214
+ "forced_bos_token_id": null,
215
+ "forced_eos_token_id": null,
216
+ "hidden_act": "silu",
217
+ "hidden_size": 3584,
218
+ "id2label": {
219
+ "0": "LABEL_0",
220
+ "1": "LABEL_1"
221
+ },
222
+ "init_std": 0.02,
223
+ "initializer_range": 0.02,
224
+ "intermediate_size": 18944,
225
+ "is_decoder": false,
226
+ "is_encoder_decoder": false,
227
+ "label2id": {
228
+ "LABEL_0": 0,
229
+ "LABEL_1": 1
230
+ },
231
+ "layer_types": [
232
+ "full_attention",
233
+ "full_attention",
234
+ "full_attention",
235
+ "full_attention",
236
+ "full_attention",
237
+ "full_attention",
238
+ "full_attention",
239
+ "full_attention",
240
+ "full_attention",
241
+ "full_attention",
242
+ "full_attention",
243
+ "full_attention",
244
+ "full_attention",
245
+ "full_attention",
246
+ "full_attention",
247
+ "full_attention",
248
+ "full_attention",
249
+ "full_attention",
250
+ "full_attention",
251
+ "full_attention",
252
+ "full_attention",
253
+ "full_attention",
254
+ "full_attention",
255
+ "full_attention",
256
+ "full_attention",
257
+ "full_attention",
258
+ "full_attention",
259
+ "full_attention"
260
+ ],
261
+ "length_penalty": 1.0,
262
+ "max_length": 20,
263
+ "max_position_embeddings": 32768,
264
+ "max_window_layers": 28,
265
+ "min_length": 0,
266
+ "model_type": "qwen2_5_omni_text",
267
+ "no_repeat_ngram_size": 0,
268
+ "num_attention_heads": 28,
269
+ "num_beam_groups": 1,
270
+ "num_beams": 1,
271
+ "num_hidden_layers": 28,
272
+ "num_key_value_heads": 4,
273
+ "num_return_sequences": 1,
274
+ "output_attentions": false,
275
+ "output_hidden_states": false,
276
+ "output_scores": false,
277
+ "pad_token_id": null,
278
+ "prefix": null,
279
+ "problem_type": null,
280
+ "pruned_heads": {},
281
+ "remove_invalid_values": false,
282
+ "repetition_penalty": 1.0,
283
+ "return_dict": true,
284
+ "return_dict_in_generate": false,
285
+ "rms_norm_eps": 1e-06,
286
+ "rope_scaling": {
287
+ "mrope_section": [
288
+ 16,
289
+ 24,
290
+ 24
291
+ ],
292
+ "rope_type": "default",
293
+ "type": "default"
294
+ },
295
+ "rope_theta": 1000000.0,
296
+ "sep_token_id": null,
297
+ "sliding_window": null,
298
+ "suppress_tokens": null,
299
+ "task_specific_params": null,
300
+ "temperature": 1.0,
301
+ "tf_legacy_loss": false,
302
+ "tie_encoder_decoder": false,
303
+ "tie_word_embeddings": false,
304
+ "tokenizer_class": null,
305
+ "top_k": 50,
306
+ "top_p": 1.0,
307
+ "torch_dtype": null,
308
+ "torchscript": false,
309
+ "typical_p": 1.0,
310
+ "use_bfloat16": false,
311
+ "use_cache": true,
312
+ "use_sliding_window": false,
313
+ "vocab_size": 152064
314
+ },
315
+ "torch_dtype": "bfloat16",
316
+ "user_token_id": 872,
317
+ "video_token_index": 151656,
318
+ "vision_config": {
319
+ "_name_or_path": "",
320
+ "add_cross_attention": false,
321
+ "architectures": null,
322
+ "bad_words_ids": null,
323
+ "begin_suppress_tokens": null,
324
+ "bos_token_id": null,
325
+ "chunk_size_feed_forward": 0,
326
+ "cross_attention_hidden_size": null,
327
+ "decoder_start_token_id": null,
328
+ "depth": 32,
329
+ "diversity_penalty": 0.0,
330
+ "do_sample": false,
331
+ "early_stopping": false,
332
+ "embed_dim": 1280,
333
+ "encoder_no_repeat_ngram_size": 0,
334
+ "eos_token_id": null,
335
+ "exponential_decay_length_penalty": null,
336
+ "finetuning_task": null,
337
+ "forced_bos_token_id": null,
338
+ "forced_eos_token_id": null,
339
+ "fullatt_block_indexes": [
340
+ 7,
341
+ 15,
342
+ 23,
343
+ 31
344
+ ],
345
+ "hidden_act": "silu",
346
+ "hidden_size": 1280,
347
+ "id2label": {
348
+ "0": "LABEL_0",
349
+ "1": "LABEL_1"
350
+ },
351
+ "in_channels": 3,
352
+ "in_chans": 3,
353
+ "init_std": 0.02,
354
+ "initializer_range": 0.02,
355
+ "intermediate_size": 3420,
356
+ "is_decoder": false,
357
+ "is_encoder_decoder": false,
358
+ "label2id": {
359
+ "LABEL_0": 0,
360
+ "LABEL_1": 1
361
+ },
362
+ "length_penalty": 1.0,
363
+ "max_length": 20,
364
+ "min_length": 0,
365
+ "model_type": "qwen2_5_omni_vision_encoder",
366
+ "no_repeat_ngram_size": 0,
367
+ "num_beam_groups": 1,
368
+ "num_beams": 1,
369
+ "num_heads": 16,
370
+ "num_return_sequences": 1,
371
+ "out_hidden_size": 3584,
372
+ "output_attentions": false,
373
+ "output_hidden_states": false,
374
+ "output_scores": false,
375
+ "pad_token_id": null,
376
+ "patch_size": 14,
377
+ "prefix": null,
378
+ "problem_type": null,
379
+ "pruned_heads": {},
380
+ "remove_invalid_values": false,
381
+ "repetition_penalty": 1.0,
382
+ "return_dict": true,
383
+ "return_dict_in_generate": false,
384
+ "sep_token_id": null,
385
+ "spatial_merge_size": 2,
386
+ "spatial_patch_size": 14,
387
+ "suppress_tokens": null,
388
+ "task_specific_params": null,
389
+ "temperature": 1.0,
390
+ "temporal_patch_size": 2,
391
+ "tf_legacy_loss": false,
392
+ "tie_encoder_decoder": false,
393
+ "tie_word_embeddings": true,
394
+ "tokenizer_class": null,
395
+ "tokens_per_second": 25,
396
+ "top_k": 50,
397
+ "top_p": 1.0,
398
+ "torch_dtype": null,
399
+ "torchscript": false,
400
+ "typical_p": 1.0,
401
+ "use_bfloat16": false,
402
+ "window_size": 112
403
+ },
404
+ "vision_end_token_id": 151653,
405
+ "vision_start_token_id": 151652,
406
+ "vision_token_id": 151654
407
+ },
408
+ "token2wav_config": {
409
+ "bigvgan_config": {
410
+ "_name_or_path": "",
411
+ "add_cross_attention": false,
412
+ "architectures": null,
413
+ "bad_words_ids": null,
414
+ "begin_suppress_tokens": null,
415
+ "bos_token_id": null,
416
+ "chunk_size_feed_forward": 0,
417
+ "cross_attention_hidden_size": null,
418
+ "decoder_start_token_id": null,
419
+ "diversity_penalty": 0.0,
420
+ "do_sample": false,
421
+ "early_stopping": false,
422
+ "encoder_no_repeat_ngram_size": 0,
423
+ "eos_token_id": null,
424
+ "exponential_decay_length_penalty": null,
425
+ "finetuning_task": null,
426
+ "forced_bos_token_id": null,
427
+ "forced_eos_token_id": null,
428
+ "id2label": {
429
+ "0": "LABEL_0",
430
+ "1": "LABEL_1"
431
+ },
432
+ "is_decoder": false,
433
+ "is_encoder_decoder": false,
434
+ "label2id": {
435
+ "LABEL_0": 0,
436
+ "LABEL_1": 1
437
+ },
438
+ "length_penalty": 1.0,
439
+ "max_length": 20,
440
+ "mel_dim": 80,
441
+ "min_length": 0,
442
+ "model_type": "qwen2_5_omni_bigvgan",
443
+ "no_repeat_ngram_size": 0,
444
+ "num_beam_groups": 1,
445
+ "num_beams": 1,
446
+ "num_return_sequences": 1,
447
+ "output_attentions": false,
448
+ "output_hidden_states": false,
449
+ "output_scores": false,
450
+ "pad_token_id": null,
451
+ "prefix": null,
452
+ "problem_type": null,
453
+ "pruned_heads": {},
454
+ "remove_invalid_values": false,
455
+ "repetition_penalty": 1.0,
456
+ "resblock_dilation_sizes": [
457
+ [
458
+ 1,
459
+ 3,
460
+ 5
461
+ ],
462
+ [
463
+ 1,
464
+ 3,
465
+ 5
466
+ ],
467
+ [
468
+ 1,
469
+ 3,
470
+ 5
471
+ ]
472
+ ],
473
+ "resblock_kernel_sizes": [
474
+ 3,
475
+ 7,
476
+ 11
477
+ ],
478
+ "return_dict": true,
479
+ "return_dict_in_generate": false,
480
+ "sep_token_id": null,
481
+ "suppress_tokens": null,
482
+ "task_specific_params": null,
483
+ "temperature": 1.0,
484
+ "tf_legacy_loss": false,
485
+ "tie_encoder_decoder": false,
486
+ "tie_word_embeddings": true,
487
+ "tokenizer_class": null,
488
+ "top_k": 50,
489
+ "top_p": 1.0,
490
+ "torch_dtype": null,
491
+ "torchscript": false,
492
+ "typical_p": 1.0,
493
+ "upsample_initial_channel": 1536,
494
+ "upsample_kernel_sizes": [
495
+ 11,
496
+ 7,
497
+ 4,
498
+ 4,
499
+ 4,
500
+ 4
501
+ ],
502
+ "upsample_rates": [
503
+ 5,
504
+ 3,
505
+ 2,
506
+ 2,
507
+ 2,
508
+ 2
509
+ ],
510
+ "use_bfloat16": false,
511
+ "use_bias_at_final": false
512
+ },
513
+ "dit_config": {
514
+ "_name_or_path": "",
515
+ "add_cross_attention": false,
516
+ "architectures": null,
517
+ "bad_words_ids": null,
518
+ "begin_suppress_tokens": null,
519
+ "block_size": 24,
520
+ "bos_token_id": null,
521
+ "chunk_size_feed_forward": 0,
522
+ "cross_attention_hidden_size": null,
523
+ "decoder_start_token_id": null,
524
+ "depth": 22,
525
+ "dim": 1024,
526
+ "diversity_penalty": 0.0,
527
+ "do_sample": false,
528
+ "dropout": 0.1,
529
+ "early_stopping": false,
530
+ "emb_dim": 512,
531
+ "enc_attention_channels": 64,
532
+ "enc_channels": [
533
+ 256,
534
+ 256,
535
+ 256,
536
+ 256,
537
+ 768
538
+ ],
539
+ "enc_dilations": [
540
+ 1,
541
+ 2,
542
+ 3,
543
+ 4,
544
+ 1
545
+ ],
546
+ "enc_dim": 128,
547
+ "enc_emb_dim": 192,
548
+ "enc_global_context": true,
549
+ "enc_kernel_sizes": [
550
+ 5,
551
+ 3,
552
+ 3,
553
+ 3,
554
+ 1
555
+ ],
556
+ "enc_lin_neurons": 192,
557
+ "enc_res2net_scale": 2,
558
+ "enc_se_channels": 64,
559
+ "encoder_no_repeat_ngram_size": 0,
560
+ "eos_token_id": null,
561
+ "exponential_decay_length_penalty": null,
562
+ "ff_mult": 2,
563
+ "finetuning_task": null,
564
+ "forced_bos_token_id": null,
565
+ "forced_eos_token_id": null,
566
+ "head_dim": 64,
567
+ "heads": 16,
568
+ "hidden_size": 1024,
569
+ "id2label": {
570
+ "0": "LABEL_0",
571
+ "1": "LABEL_1"
572
+ },
573
+ "is_decoder": false,
574
+ "is_encoder_decoder": false,
575
+ "label2id": {
576
+ "LABEL_0": 0,
577
+ "LABEL_1": 1
578
+ },
579
+ "length_penalty": 1.0,
580
+ "look_ahead_layers": [
581
+ 10
582
+ ],
583
+ "look_backward_layers": [
584
+ 0,
585
+ 20
586
+ ],
587
+ "max_length": 20,
588
+ "max_position_embeddings": 32768,
589
+ "mel_dim": 80,
590
+ "min_length": 0,
591
+ "model_type": "qwen2_5_omni_dit",
592
+ "no_repeat_ngram_size": 0,
593
+ "num_attention_heads": 16,
594
+ "num_beam_groups": 1,
595
+ "num_beams": 1,
596
+ "num_embeds": 8193,
597
+ "num_hidden_layers": 22,
598
+ "num_return_sequences": 1,
599
+ "output_attentions": false,
600
+ "output_hidden_states": false,
601
+ "output_scores": false,
602
+ "pad_token_id": null,
603
+ "prefix": null,
604
+ "problem_type": null,
605
+ "pruned_heads": {},
606
+ "remove_invalid_values": false,
607
+ "repeats": 2,
608
+ "repetition_penalty": 1.0,
609
+ "return_dict": true,
610
+ "return_dict_in_generate": false,
611
+ "rope_theta": 10000.0,
612
+ "sep_token_id": null,
613
+ "suppress_tokens": null,
614
+ "task_specific_params": null,
615
+ "temperature": 1.0,
616
+ "tf_legacy_loss": false,
617
+ "tie_encoder_decoder": false,
618
+ "tie_word_embeddings": true,
619
+ "tokenizer_class": null,
620
+ "top_k": 50,
621
+ "top_p": 1.0,
622
+ "torch_dtype": "float32",
623
+ "torchscript": false,
624
+ "typical_p": 1.0,
625
+ "use_bfloat16": false
626
+ },
627
+ "model_type": "qwen2_5_omni_token2wav",
628
+ "pad_token_id": 151643,
629
+ "torch_dtype": "bfloat16"
630
+ },
631
+ "torch_dtype": "bfloat16",
632
+ "transformers_version": "4.53.3"
633
+ }
file_list.txt ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ model-00005-of-00005.safetensors
2
+ config.json
3
+ spk_dict.pt
4
+ video_preprocessor_config.json
5
+ model-00004-of-00005.safetensors
6
+ vocab.json
7
+ tokenizer_config.json
8
+ chat_template.jinja
9
+ model-00002-of-00005.safetensors
10
+ model.safetensors.index.json
11
+ model-00003-of-00005.safetensors
12
+ model-00001-of-00005.safetensors
13
+ special_tokens_map.json
14
+ added_tokens.json
15
+ tokenizer.json
16
+ generation_config.json
17
+ preprocessor_config.json
18
+ merges.txt
19
+ args.json
20
+ deploy_result/20250825-143105.jsonl
21
+ deploy_result/20250825-141923.jsonl
22
+ deploy_result/20250827-025556.jsonl
generation_config.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "eos_token_id": [
4
+ 151645,
5
+ 151643
6
+ ],
7
+ "transformers_version": "4.53.3"
8
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
model-00001-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a8f2e6519b786ab2923ff15aa6affef5a35a7e35e7227e92829bdb832fc38e0e
3
+ size 4985055536
model-00002-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f81d393053fcba5d3628ecd88dbee049ec3100e915169d1d4d0d079faba44536
3
+ size 4991496832
model-00003-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:512296d0b7d73bddf32e1a3924ae18ca365c146d68ac8fd6640f9da2542130de
3
+ size 4991496936
model-00004-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2acab89dcc8df7f1683984b9282ffe9efe17decd427e2cb451fe4c65075c39b7
3
+ size 4969489856
model-00005-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:19ac6c0769f8f6a9ffe25540f79d16b8fcb26ac40902c6e5a0ecade50554d635
3
+ size 2425322192
model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
preprocessor_config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "chunk_length": 300,
3
+ "dither": 0.0,
4
+ "feature_extractor_type": "WhisperFeatureExtractor",
5
+ "feature_size": 128,
6
+ "hop_length": 160,
7
+ "image_mean": [
8
+ 0.48145466,
9
+ 0.4578275,
10
+ 0.40821073
11
+ ],
12
+ "image_processor_type": "Qwen2VLImageProcessor",
13
+ "image_std": [
14
+ 0.26862954,
15
+ 0.26130258,
16
+ 0.27577711
17
+ ],
18
+ "max_pixels": 12845056,
19
+ "merge_size": 2,
20
+ "min_pixels": 3136,
21
+ "n_fft": 400,
22
+ "n_samples": 4800000,
23
+ "nb_max_frames": 30000,
24
+ "padding_side": "right",
25
+ "padding_value": 0.0,
26
+ "patch_size": 14,
27
+ "processor_class": "Qwen2_5OmniProcessor",
28
+ "return_attention_mask": true,
29
+ "sampling_rate": 16000,
30
+ "temporal_patch_size": 2
31
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|im_start|>",
4
+ "<|im_end|>",
5
+ "<|AUDIO|>",
6
+ "<|audio_bos|>",
7
+ "<|audio_eos|>",
8
+ "<|box_end|>",
9
+ "<|quad_start|>",
10
+ "<|quad_end|>",
11
+ "<|vision_bos|>",
12
+ "<|vision_eos|>",
13
+ "<|vision_pad|>",
14
+ "<|IMAGE|>",
15
+ "<|VIDEO|>"
16
+ ],
17
+ "audio_bos_token": "<|audio_bos|>",
18
+ "audio_eos_token": "<|audio_eos|>",
19
+ "audio_token": "<|AUDIO|>",
20
+ "eos_token": {
21
+ "content": "<|im_end|>",
22
+ "lstrip": false,
23
+ "normalized": false,
24
+ "rstrip": false,
25
+ "single_word": false
26
+ },
27
+ "image_token": "<|IMAGE|>",
28
+ "pad_token": {
29
+ "content": "<|endoftext|>",
30
+ "lstrip": false,
31
+ "normalized": false,
32
+ "rstrip": false,
33
+ "single_word": false
34
+ },
35
+ "video_token": "<|VIDEO|>",
36
+ "vision_bos_token": "<|vision_bos|>",
37
+ "vision_eos_token": "<|vision_eos|>"
38
+ }
spk_dict.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6a05609b28f5d42b7b748f0f07592545c8f1f6885b9ae8fff64baf56e86b2a18
3
+ size 259544
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8441917e39ae0244e06d704b95b3124795cec478e297f9afac39ba670d7e9d99
3
+ size 11421870
tokenizer_config.json ADDED
@@ -0,0 +1,222 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "151643": {
5
+ "content": "<|endoftext|>",
6
+ "lstrip": false,
7
+ "normalized": false,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ },
12
+ "151644": {
13
+ "content": "<|im_start|>",
14
+ "lstrip": false,
15
+ "normalized": false,
16
+ "rstrip": false,
17
+ "single_word": false,
18
+ "special": true
19
+ },
20
+ "151645": {
21
+ "content": "<|im_end|>",
22
+ "lstrip": false,
23
+ "normalized": false,
24
+ "rstrip": false,
25
+ "single_word": false,
26
+ "special": true
27
+ },
28
+ "151646": {
29
+ "content": "<|AUDIO|>",
30
+ "lstrip": false,
31
+ "normalized": false,
32
+ "rstrip": false,
33
+ "single_word": false,
34
+ "special": true
35
+ },
36
+ "151647": {
37
+ "content": "<|audio_bos|>",
38
+ "lstrip": false,
39
+ "normalized": false,
40
+ "rstrip": false,
41
+ "single_word": false,
42
+ "special": true
43
+ },
44
+ "151648": {
45
+ "content": "<|audio_eos|>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false,
50
+ "special": true
51
+ },
52
+ "151649": {
53
+ "content": "<|box_end|>",
54
+ "lstrip": false,
55
+ "normalized": false,
56
+ "rstrip": false,
57
+ "single_word": false,
58
+ "special": true
59
+ },
60
+ "151650": {
61
+ "content": "<|quad_start|>",
62
+ "lstrip": false,
63
+ "normalized": false,
64
+ "rstrip": false,
65
+ "single_word": false,
66
+ "special": true
67
+ },
68
+ "151651": {
69
+ "content": "<|quad_end|>",
70
+ "lstrip": false,
71
+ "normalized": false,
72
+ "rstrip": false,
73
+ "single_word": false,
74
+ "special": true
75
+ },
76
+ "151652": {
77
+ "content": "<|vision_bos|>",
78
+ "lstrip": false,
79
+ "normalized": false,
80
+ "rstrip": false,
81
+ "single_word": false,
82
+ "special": true
83
+ },
84
+ "151653": {
85
+ "content": "<|vision_eos|>",
86
+ "lstrip": false,
87
+ "normalized": false,
88
+ "rstrip": false,
89
+ "single_word": false,
90
+ "special": true
91
+ },
92
+ "151654": {
93
+ "content": "<|vision_pad|>",
94
+ "lstrip": false,
95
+ "normalized": false,
96
+ "rstrip": false,
97
+ "single_word": false,
98
+ "special": true
99
+ },
100
+ "151655": {
101
+ "content": "<|IMAGE|>",
102
+ "lstrip": false,
103
+ "normalized": false,
104
+ "rstrip": false,
105
+ "single_word": false,
106
+ "special": true
107
+ },
108
+ "151656": {
109
+ "content": "<|VIDEO|>",
110
+ "lstrip": false,
111
+ "normalized": false,
112
+ "rstrip": false,
113
+ "single_word": false,
114
+ "special": true
115
+ },
116
+ "151657": {
117
+ "content": "<tool_call>",
118
+ "lstrip": false,
119
+ "normalized": false,
120
+ "rstrip": false,
121
+ "single_word": false,
122
+ "special": false
123
+ },
124
+ "151658": {
125
+ "content": "</tool_call>",
126
+ "lstrip": false,
127
+ "normalized": false,
128
+ "rstrip": false,
129
+ "single_word": false,
130
+ "special": false
131
+ },
132
+ "151659": {
133
+ "content": "<|fim_prefix|>",
134
+ "lstrip": false,
135
+ "normalized": false,
136
+ "rstrip": false,
137
+ "single_word": false,
138
+ "special": false
139
+ },
140
+ "151660": {
141
+ "content": "<|fim_middle|>",
142
+ "lstrip": false,
143
+ "normalized": false,
144
+ "rstrip": false,
145
+ "single_word": false,
146
+ "special": false
147
+ },
148
+ "151661": {
149
+ "content": "<|fim_suffix|>",
150
+ "lstrip": false,
151
+ "normalized": false,
152
+ "rstrip": false,
153
+ "single_word": false,
154
+ "special": false
155
+ },
156
+ "151662": {
157
+ "content": "<|fim_pad|>",
158
+ "lstrip": false,
159
+ "normalized": false,
160
+ "rstrip": false,
161
+ "single_word": false,
162
+ "special": false
163
+ },
164
+ "151663": {
165
+ "content": "<|repo_name|>",
166
+ "lstrip": false,
167
+ "normalized": false,
168
+ "rstrip": false,
169
+ "single_word": false,
170
+ "special": false
171
+ },
172
+ "151664": {
173
+ "content": "<|file_sep|>",
174
+ "lstrip": false,
175
+ "normalized": false,
176
+ "rstrip": false,
177
+ "single_word": false,
178
+ "special": false
179
+ }
180
+ },
181
+ "additional_special_tokens": [
182
+ "<|im_start|>",
183
+ "<|im_end|>",
184
+ "<|AUDIO|>",
185
+ "<|audio_bos|>",
186
+ "<|audio_eos|>",
187
+ "<|box_end|>",
188
+ "<|quad_start|>",
189
+ "<|quad_end|>",
190
+ "<|vision_bos|>",
191
+ "<|vision_eos|>",
192
+ "<|vision_pad|>",
193
+ "<|IMAGE|>",
194
+ "<|VIDEO|>"
195
+ ],
196
+ "audio_bos_token": "<|audio_bos|>",
197
+ "audio_eos_token": "<|audio_eos|>",
198
+ "audio_token": "<|AUDIO|>",
199
+ "bos_token": null,
200
+ "clean_up_tokenization_spaces": false,
201
+ "eos_token": "<|im_end|>",
202
+ "errors": "replace",
203
+ "extra_special_tokens": {
204
+ "audio_bos_token": "<|audio_bos|>",
205
+ "audio_eos_token": "<|audio_eos|>",
206
+ "audio_token": "<|AUDIO|>",
207
+ "image_token": "<|IMAGE|>",
208
+ "video_token": "<|VIDEO|>",
209
+ "vision_bos_token": "<|vision_bos|>",
210
+ "vision_eos_token": "<|vision_eos|>"
211
+ },
212
+ "image_token": "<|IMAGE|>",
213
+ "model_max_length": 32768,
214
+ "pad_token": "<|endoftext|>",
215
+ "processor_class": "Qwen2_5OmniProcessor",
216
+ "split_special_tokens": false,
217
+ "tokenizer_class": "Qwen2Tokenizer",
218
+ "unk_token": null,
219
+ "video_token": "<|VIDEO|>",
220
+ "vision_bos_token": "<|vision_bos|>",
221
+ "vision_eos_token": "<|vision_eos|>"
222
+ }
video_preprocessor_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "chunk_length": 300,
3
+ "crop_size": null,
4
+ "data_format": "channels_first",
5
+ "default_to_square": true,
6
+ "device": null,
7
+ "dither": 0.0,
8
+ "do_center_crop": null,
9
+ "do_convert_rgb": true,
10
+ "do_normalize": true,
11
+ "do_pad": null,
12
+ "do_rescale": true,
13
+ "do_resize": true,
14
+ "do_sample_frames": false,
15
+ "feature_extractor_type": "WhisperFeatureExtractor",
16
+ "feature_size": 128,
17
+ "fps": null,
18
+ "hop_length": 160,
19
+ "image_mean": [
20
+ 0.48145466,
21
+ 0.4578275,
22
+ 0.40821073
23
+ ],
24
+ "image_processor_type": "Qwen2VLImageProcessor",
25
+ "image_std": [
26
+ 0.26862954,
27
+ 0.26130258,
28
+ 0.27577711
29
+ ],
30
+ "input_data_format": null,
31
+ "max_frames": 768,
32
+ "max_pixels": 12845056,
33
+ "merge_size": 2,
34
+ "min_frames": 4,
35
+ "min_pixels": 3136,
36
+ "n_fft": 400,
37
+ "n_samples": 4800000,
38
+ "nb_max_frames": 30000,
39
+ "num_frames": null,
40
+ "padding_side": "right",
41
+ "padding_value": 0.0,
42
+ "patch_size": 14,
43
+ "processor_class": "Qwen2_5OmniProcessor",
44
+ "resample": 3,
45
+ "rescale_factor": 0.00392156862745098,
46
+ "return_attention_mask": true,
47
+ "sampling_rate": 16000,
48
+ "size": {
49
+ "longest_edge": 12845056,
50
+ "shortest_edge": 3136
51
+ },
52
+ "size_divisor": null,
53
+ "temporal_patch_size": 2,
54
+ "video_metadata": null,
55
+ "video_processor_type": "Qwen2VLVideoProcessor"
56
+ }
vocab.json ADDED
The diff for this file is too large to render. See raw diff