ictnlp/llama-omni 🔢📝🖼️ → ❓
About
Seamless Speech Interaction with Large Language Models

Example Output
Prompt:
"Please directly answer the questions in the user's speech"
Output
{"text":"The origin of US state names is varied, but most were named by European explorers and settlers. Many were named after Native American tribes, Spanish and Mexican cities, or royal figures. Some states were also named after natural features, like rivers or mountains.","audio":"https://replicate.delivery/pbxt/A1reTPbzlkQiCKrfbuUyI78O1wWDkNPnQPkL9HGjUN7wklfmA/out.wav"}
Performance Metrics
18.50s
Prediction Time
184.81s
Total Time
All Input Parameters
{ "top_p": 0, "prompt": "Please directly answer the questions in the user's speech", "input_audio": "https://replicate.delivery/pbxt/LfbWz5nAdlqDatmo2feweGHjcVyJHdQhqZYRNHqfJ7EyKxXa/helpful_base_1.wav", "temperature": 0, "max_new_tokens": 256 }
Input Parameters
- top_p
- Controls diversity of the output. Valid when temperature > 0. Lower values make the output more focused, higher values make it more diverse.
- prompt
- input_audio (required)
- Input audio
- temperature
- Controls randomness. Lower values make the model more deterministic, higher values make it more random.
- max_new_tokens
- Maximum number of tokens to generate
Output Schema
Example Execution Logs
/root/.pyenv/versions/3.10.15/lib/python3.10/site-packages/transformers/utils/hub.py:127: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead. warnings.warn( /root/.pyenv/versions/3.10.15/lib/python3.10/site-packages/transformers/utils/hub.py:127: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead. warnings.warn( /root/.pyenv/versions/3.10.15/lib/python3.10/site-packages/transformers/utils/hub.py:127: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead. warnings.warn( /root/.pyenv/versions/3.10.15/lib/python3.10/site-packages/transformers/utils/hub.py:127: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead. warnings.warn( /root/.pyenv/versions/3.10.15/lib/python3.10/site-packages/transformers/generation/configuration_utils.py:567: UserWarning: `do_sample` is set to `False`. However, `temperature` is set to `0.0` -- this flag is only used in sample-based generation modes. You should set `do_sample=True` or unset `temperature`. warnings.warn( The attention layers in this model are transitioning from computing the RoPE embeddings internally through `position_ids` (2D tensor with the indexes of the tokens), to using externally computed `position_embeddings` (Tuple of tensors, containing cos and sin). In v4.45 `position_ids` will be removed and `position_embeddings` will be mandatory. The origin of US state names is varied, but most were named by European explorers and settlers. Many were named after Native American tribes, Spanish and Mexican cities, or royal figures. Some states were also named after natural features, like rivers or mountains. output_units: 202 393 946 215 406 538 187 594 908 246 466 503 523 705 11 283 488 620 352 931 932 148 258 436 139 340 483 384 879 70 32 835 683 67 589 702 576 822 89 194 664 506 29 116 281 428 822 89 194 627 545 711 510 169 237 865 641 124 243 526 384 249 466 405 53 664 555 208 417 755 237 193 128 665 547 833 368 945 29 73 324 789 6 908 380 828 835 67 940 118 243 935 101 741 663 575 116 281 428 822 89 194 627 208 944 833 368 837 81 664 258 436 573 391 24 870 188 485 841 488 620 352 487 219 522 589 126 712 593 592 103 466 663 969 198 711 510 362 684 136 912 519 589 26 204 280 576 6 879 185 794 788 402 663 969 198 711 510 297 265 675 237 415 772 497 63 991 162 73 172 871 877 384 879 179 961 207 428 950 321 948 86 787 935 101 741 663 575 116 281 428 822 89 194 627 915 208 944 878 423 27 907 430 70 390 595 600 702 788 663 575 116 281 428 822 89 194 664 539 794 680 910 161 998 885 148 878 565 734 498 172 871 877 384 466 503 487 319 501 137 161 488 352 143 38 777 728 227 647 655 764 837 81 194 664 506 545 711 510 297 265 675 237 307 128 665 780 519 589 126 323 576 384 430 179 961 428 333 432 431 531 362 488 352 136 889 871 384 969 219 522 866 586 314 333 319 990 501 137 333 352 915 912 519 26 204 280 314 333 523 793 50 534 321 948 198 711 510 297 265 675 237 187 594 461 969 156 824 592 103 483 81 327 635 205 521 382 390 479 330 776 333 350 836 74 377 969 377 198 711 510 297 265 675 237 415 772 497 63 780 519 289 26 204 280 668 167 104 896 627 912 519 589 702 874 576 822 89 194 664 506 545 85 510 243 101 663 538 187 594 310 346 540 295 76 614 380 116 281 428 822 89 194 627 915 208 944 907 430 70 958 595 315 702 788 663 575 116 281 62 384 761 430 70 185 477 728 716 205 521 382 390 479 330 776 485 948 86 539 557 477 74 663 969 198 711 510 265 675 237 307 665 991 781 645 837 81 664 219 107 29 156 824 442 333 998 885 692 154 559 663 969 198 711 510 362 461 969 498 889 338 359 6 761 907 597 816 274 794 75 788 15 377 832 758 545 85 510 297 265 675 237 415 772 497 <class 'str'>
Version Details
- Version ID
36c9bcf70a56f40d9a27445c30c769308b18180296749f86ec9b682baf7ad351
- Version Created
- September 22, 2024