Instructions to use KillerBoss/functiongemma-plantgame-de with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- LiteRT
How to use KillerBoss/functiongemma-plantgame-de with LiteRT:
# No code snippets available yet for this library. # To use this model, check the repository files and the library's documentation. # Want to help? PRs adding snippets are welcome at: # https://github.com/huggingface/huggingface.js
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use KillerBoss/functiongemma-plantgame-de with llama.cpp:
Install (macOS, Linux)
curl -LsSf https://llama.app/install.sh | sh # Start a local OpenAI-compatible server with a web UI: llama serve -hf KillerBoss/functiongemma-plantgame-de:F16 # Run inference directly in the terminal: llama cli -hf KillerBoss/functiongemma-plantgame-de:F16
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf KillerBoss/functiongemma-plantgame-de:F16 # Run inference directly in the terminal: llama cli -hf KillerBoss/functiongemma-plantgame-de:F16
Build from source code
git clone https://github.com/ggml-org/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-app # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama serve -hf KillerBoss/functiongemma-plantgame-de:F16 # Run inference directly in the terminal: ./build/bin/llama cli -hf KillerBoss/functiongemma-plantgame-de:F16
Use Docker
docker model run hf.co/KillerBoss/functiongemma-plantgame-de:F16
- Llama new
- LM Studio
- Jan
- Ollama
How to use KillerBoss/functiongemma-plantgame-de with Ollama:
ollama run hf.co/KillerBoss/functiongemma-plantgame-de:F16
- Unsloth Desktop
- Docker Model Runner
How to use KillerBoss/functiongemma-plantgame-de with Docker Model Runner:
docker model run hf.co/KillerBoss/functiongemma-plantgame-de:F16
- Lemonade
How to use KillerBoss/functiongemma-plantgame-de with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull KillerBoss/functiongemma-plantgame-de:F16
Run and chat with the model
lemonade run user.functiongemma-plantgame-de-F16
List all available models
lemonade list
- Atomic Chat
log: export v6 npu-fail-SM8750
Browse files- logs/export_v6.log +359 -0
logs/export_v6.log
CHANGED
|
@@ -486,3 +486,362 @@ ValueError: apply_plugin failed to apply plugin. See /tmp/tmp97ocbbyw.error for
|
|
| 486 |
|
| 487 |
litertlm-Dateien: ['/kaggle/working/export_npu/model.litertlm']
|
| 488 |
### NPU-Export mit SM8850 fehlgeschlagen (rc=1)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 486 |
|
| 487 |
litertlm-Dateien: ['/kaggle/working/export_npu/model.litertlm']
|
| 488 |
### NPU-Export mit SM8850 fehlgeschlagen (rc=1)
|
| 489 |
+
>>> Log nach HF (npu-fail-SM8850)
|
| 490 |
+
|
| 491 |
+
### NPU-Versuch mit SoC: SM8750
|
| 492 |
+
+ litert-torch export_hf /kaggle/tmp/functiongemma-plantgame-de /kaggle/working/export_npu --task=text_generation --bundle_litert_lm=True --aot_backend=qualcomm --aot_soc_model=SM8750 --use_litert_lm_compiler=True --externalize_embedder=True --prefill_lengths=128 --cache_length=1280 --keep_temporary_files=True
|
| 493 |
+
h: 1280
|
| 494 |
+
}
|
| 495 |
+
state_buffers {
|
| 496 |
+
prefill_input_name: "kv_cache_v_3"
|
| 497 |
+
prefill_output_name: "kv_cache_v_3"
|
| 498 |
+
decode_input_name: "kv_cache_v_3"
|
| 499 |
+
decode_output_name: "kv_cache_v_3"
|
| 500 |
+
type: TYPE_GLOBAL_VALUE_CACHE
|
| 501 |
+
sequence_axis: 2
|
| 502 |
+
maximum_sequence_length: 1280
|
| 503 |
+
}
|
| 504 |
+
state_buffers {
|
| 505 |
+
prefill_input_name: "kv_cache_k_4"
|
| 506 |
+
prefill_output_name: "kv_cache_k_4"
|
| 507 |
+
decode_input_name: "kv_cache_k_4"
|
| 508 |
+
decode_output_name: "kv_cache_k_4"
|
| 509 |
+
type: TYPE_GLOBAL_KEY_CACHE
|
| 510 |
+
sequence_axis: 3
|
| 511 |
+
maximum_sequence_length: 1280
|
| 512 |
+
}
|
| 513 |
+
state_buffers {
|
| 514 |
+
prefill_input_name: "kv_cache_v_4"
|
| 515 |
+
prefill_output_name: "kv_cache_v_4"
|
| 516 |
+
decode_input_name: "kv_cache_v_4"
|
| 517 |
+
decode_output_name: "kv_cache_v_4"
|
| 518 |
+
type: TYPE_GLOBAL_VALUE_CACHE
|
| 519 |
+
sequence_axis: 2
|
| 520 |
+
maximum_sequence_length: 1280
|
| 521 |
+
}
|
| 522 |
+
state_buffers {
|
| 523 |
+
prefill_input_name: "kv_cache_k_5"
|
| 524 |
+
prefill_output_name: "kv_cache_k_5"
|
| 525 |
+
decode_input_name: "kv_cache_k_5"
|
| 526 |
+
decode_output_name: "kv_cache_k_5"
|
| 527 |
+
type: TYPE_GLOBAL_KEY_CACHE
|
| 528 |
+
sequence_axis: 3
|
| 529 |
+
maximum_sequence_length: 1280
|
| 530 |
+
}
|
| 531 |
+
state_buffers {
|
| 532 |
+
prefill_input_name: "kv_cache_v_5"
|
| 533 |
+
prefill_output_name: "kv_cache_v_5"
|
| 534 |
+
decode_input_name: "kv_cache_v_5"
|
| 535 |
+
decode_output_name: "kv_cache_v_5"
|
| 536 |
+
type: TYPE_GLOBAL_VALUE_CACHE
|
| 537 |
+
sequence_axis: 2
|
| 538 |
+
maximum_sequence_length: 1280
|
| 539 |
+
}
|
| 540 |
+
state_buffers {
|
| 541 |
+
prefill_input_name: "kv_cache_k_6"
|
| 542 |
+
prefill_output_name: "kv_cache_k_6"
|
| 543 |
+
decode_input_name: "kv_cache_k_6"
|
| 544 |
+
decode_output_name: "kv_cache_k_6"
|
| 545 |
+
type: TYPE_GLOBAL_KEY_CACHE
|
| 546 |
+
sequence_axis: 3
|
| 547 |
+
maximum_sequence_length: 1280
|
| 548 |
+
}
|
| 549 |
+
state_buffers {
|
| 550 |
+
prefill_input_name: "kv_cache_v_6"
|
| 551 |
+
prefill_output_name: "kv_cache_v_6"
|
| 552 |
+
decode_input_name: "kv_cache_v_6"
|
| 553 |
+
decode_output_name: "kv_cache_v_6"
|
| 554 |
+
type: TYPE_GLOBAL_VALUE_CACHE
|
| 555 |
+
sequence_axis: 2
|
| 556 |
+
maximum_sequence_length: 1280
|
| 557 |
+
}
|
| 558 |
+
state_buffers {
|
| 559 |
+
prefill_input_name: "kv_cache_k_7"
|
| 560 |
+
prefill_output_name: "kv_cache_k_7"
|
| 561 |
+
decode_input_name: "kv_cache_k_7"
|
| 562 |
+
decode_output_name: "kv_cache_k_7"
|
| 563 |
+
type: TYPE_GLOBAL_KEY_CACHE
|
| 564 |
+
sequence_axis: 3
|
| 565 |
+
maximum_sequence_length: 1280
|
| 566 |
+
}
|
| 567 |
+
state_buffers {
|
| 568 |
+
prefill_input_name: "kv_cache_v_7"
|
| 569 |
+
prefill_output_name: "kv_cache_v_7"
|
| 570 |
+
decode_input_name: "kv_cache_v_7"
|
| 571 |
+
decode_output_name: "kv_cache_v_7"
|
| 572 |
+
type: TYPE_GLOBAL_VALUE_CACHE
|
| 573 |
+
sequence_axis: 2
|
| 574 |
+
maximum_sequence_length: 1280
|
| 575 |
+
}
|
| 576 |
+
state_buffers {
|
| 577 |
+
prefill_input_name: "kv_cache_k_8"
|
| 578 |
+
prefill_output_name: "kv_cache_k_8"
|
| 579 |
+
decode_input_name: "kv_cache_k_8"
|
| 580 |
+
decode_output_name: "kv_cache_k_8"
|
| 581 |
+
type: TYPE_GLOBAL_KEY_CACHE
|
| 582 |
+
sequence_axis: 3
|
| 583 |
+
maximum_sequence_length: 1280
|
| 584 |
+
}
|
| 585 |
+
state_buffers {
|
| 586 |
+
prefill_input_name: "kv_cache_v_8"
|
| 587 |
+
prefill_output_name: "kv_cache_v_8"
|
| 588 |
+
decode_input_name: "kv_cache_v_8"
|
| 589 |
+
decode_output_name: "kv_cache_v_8"
|
| 590 |
+
type: TYPE_GLOBAL_VALUE_CACHE
|
| 591 |
+
sequence_axis: 2
|
| 592 |
+
maximum_sequence_length: 1280
|
| 593 |
+
}
|
| 594 |
+
state_buffers {
|
| 595 |
+
prefill_input_name: "kv_cache_k_9"
|
| 596 |
+
prefill_output_name: "kv_cache_k_9"
|
| 597 |
+
decode_input_name: "kv_cache_k_9"
|
| 598 |
+
decode_output_name: "kv_cache_k_9"
|
| 599 |
+
type: TYPE_GLOBAL_KEY_CACHE
|
| 600 |
+
sequence_axis: 3
|
| 601 |
+
maximum_sequence_length: 1280
|
| 602 |
+
}
|
| 603 |
+
state_buffers {
|
| 604 |
+
prefill_input_name: "kv_cache_v_9"
|
| 605 |
+
prefill_output_name: "kv_cache_v_9"
|
| 606 |
+
decode_input_name: "kv_cache_v_9"
|
| 607 |
+
decode_output_name: "kv_cache_v_9"
|
| 608 |
+
type: TYPE_GLOBAL_VALUE_CACHE
|
| 609 |
+
sequence_axis: 2
|
| 610 |
+
maximum_sequence_length: 1280
|
| 611 |
+
}
|
| 612 |
+
state_buffers {
|
| 613 |
+
prefill_input_name: "kv_cache_k_10"
|
| 614 |
+
prefill_output_name: "kv_cache_k_10"
|
| 615 |
+
decode_input_name: "kv_cache_k_10"
|
| 616 |
+
decode_output_name: "kv_cache_k_10"
|
| 617 |
+
type: TYPE_GLOBAL_KEY_CACHE
|
| 618 |
+
sequence_axis: 3
|
| 619 |
+
maximum_sequence_length: 1280
|
| 620 |
+
}
|
| 621 |
+
state_buffers {
|
| 622 |
+
prefill_input_name: "kv_cache_v_10"
|
| 623 |
+
prefill_output_name: "kv_cache_v_10"
|
| 624 |
+
decode_input_name: "kv_cache_v_10"
|
| 625 |
+
decode_output_name: "kv_cache_v_10"
|
| 626 |
+
type: TYPE_GLOBAL_VALUE_CACHE
|
| 627 |
+
sequence_axis: 2
|
| 628 |
+
maximum_sequence_length: 1280
|
| 629 |
+
}
|
| 630 |
+
state_buffers {
|
| 631 |
+
prefill_input_name: "kv_cache_k_11"
|
| 632 |
+
prefill_output_name: "kv_cache_k_11"
|
| 633 |
+
decode_input_name: "kv_cache_k_11"
|
| 634 |
+
decode_output_name: "kv_cache_k_11"
|
| 635 |
+
type: TYPE_GLOBAL_KEY_CACHE
|
| 636 |
+
sequence_axis: 3
|
| 637 |
+
maximum_sequence_length: 1280
|
| 638 |
+
}
|
| 639 |
+
state_buffers {
|
| 640 |
+
prefill_input_name: "kv_cache_v_11"
|
| 641 |
+
prefill_output_name: "kv_cache_v_11"
|
| 642 |
+
decode_input_name: "kv_cache_v_11"
|
| 643 |
+
decode_output_name: "kv_cache_v_11"
|
| 644 |
+
type: TYPE_GLOBAL_VALUE_CACHE
|
| 645 |
+
sequence_axis: 2
|
| 646 |
+
maximum_sequence_length: 1280
|
| 647 |
+
}
|
| 648 |
+
state_buffers {
|
| 649 |
+
prefill_input_name: "kv_cache_k_12"
|
| 650 |
+
prefill_output_name: "kv_cache_k_12"
|
| 651 |
+
decode_input_name: "kv_cache_k_12"
|
| 652 |
+
decode_output_name: "kv_cache_k_12"
|
| 653 |
+
type: TYPE_GLOBAL_KEY_CACHE
|
| 654 |
+
sequence_axis: 3
|
| 655 |
+
maximum_sequence_length: 1280
|
| 656 |
+
}
|
| 657 |
+
state_buffers {
|
| 658 |
+
prefill_input_name: "kv_cache_v_12"
|
| 659 |
+
prefill_output_name: "kv_cache_v_12"
|
| 660 |
+
decode_input_name: "kv_cache_v_12"
|
| 661 |
+
decode_output_name: "kv_cache_v_12"
|
| 662 |
+
type: TYPE_GLOBAL_VALUE_CACHE
|
| 663 |
+
sequence_axis: 2
|
| 664 |
+
maximum_sequence_length: 1280
|
| 665 |
+
}
|
| 666 |
+
state_buffers {
|
| 667 |
+
prefill_input_name: "kv_cache_k_13"
|
| 668 |
+
prefill_output_name: "kv_cache_k_13"
|
| 669 |
+
decode_input_name: "kv_cache_k_13"
|
| 670 |
+
decode_output_name: "kv_cache_k_13"
|
| 671 |
+
type: TYPE_GLOBAL_KEY_CACHE
|
| 672 |
+
sequence_axis: 3
|
| 673 |
+
maximum_sequence_length: 1280
|
| 674 |
+
}
|
| 675 |
+
state_buffers {
|
| 676 |
+
prefill_input_name: "kv_cache_v_13"
|
| 677 |
+
prefill_output_name: "kv_cache_v_13"
|
| 678 |
+
decode_input_name: "kv_cache_v_13"
|
| 679 |
+
decode_output_name: "kv_cache_v_13"
|
| 680 |
+
type: TYPE_GLOBAL_VALUE_CACHE
|
| 681 |
+
sequence_axis: 2
|
| 682 |
+
maximum_sequence_length: 1280
|
| 683 |
+
}
|
| 684 |
+
state_buffers {
|
| 685 |
+
prefill_input_name: "kv_cache_k_14"
|
| 686 |
+
prefill_output_name: "kv_cache_k_14"
|
| 687 |
+
decode_input_name: "kv_cache_k_14"
|
| 688 |
+
decode_output_name: "kv_cache_k_14"
|
| 689 |
+
type: TYPE_GLOBAL_KEY_CACHE
|
| 690 |
+
sequence_axis: 3
|
| 691 |
+
maximum_sequence_length: 1280
|
| 692 |
+
}
|
| 693 |
+
state_buffers {
|
| 694 |
+
prefill_input_name: "kv_cache_v_14"
|
| 695 |
+
prefill_output_name: "kv_cache_v_14"
|
| 696 |
+
decode_input_name: "kv_cache_v_14"
|
| 697 |
+
decode_output_name: "kv_cache_v_14"
|
| 698 |
+
type: TYPE_GLOBAL_VALUE_CACHE
|
| 699 |
+
sequence_axis: 2
|
| 700 |
+
maximum_sequence_length: 1280
|
| 701 |
+
}
|
| 702 |
+
state_buffers {
|
| 703 |
+
prefill_input_name: "kv_cache_k_15"
|
| 704 |
+
prefill_output_name: "kv_cache_k_15"
|
| 705 |
+
decode_input_name: "kv_cache_k_15"
|
| 706 |
+
decode_output_name: "kv_cache_k_15"
|
| 707 |
+
type: TYPE_GLOBAL_KEY_CACHE
|
| 708 |
+
sequence_axis: 3
|
| 709 |
+
maximum_sequence_length: 1280
|
| 710 |
+
}
|
| 711 |
+
state_buffers {
|
| 712 |
+
prefill_input_name: "kv_cache_v_15"
|
| 713 |
+
prefill_output_name: "kv_cache_v_15"
|
| 714 |
+
decode_input_name: "kv_cache_v_15"
|
| 715 |
+
decode_output_name: "kv_cache_v_15"
|
| 716 |
+
type: TYPE_GLOBAL_VALUE_CACHE
|
| 717 |
+
sequence_axis: 2
|
| 718 |
+
maximum_sequence_length: 1280
|
| 719 |
+
}
|
| 720 |
+
state_buffers {
|
| 721 |
+
prefill_input_name: "kv_cache_k_16"
|
| 722 |
+
prefill_output_name: "kv_cache_k_16"
|
| 723 |
+
decode_input_name: "kv_cache_k_16"
|
| 724 |
+
decode_output_name: "kv_cache_k_16"
|
| 725 |
+
type: TYPE_GLOBAL_KEY_CACHE
|
| 726 |
+
sequence_axis: 3
|
| 727 |
+
maximum_sequence_length: 1280
|
| 728 |
+
}
|
| 729 |
+
state_buffers {
|
| 730 |
+
prefill_input_name: "kv_cache_v_16"
|
| 731 |
+
prefill_output_name: "kv_cache_v_16"
|
| 732 |
+
decode_input_name: "kv_cache_v_16"
|
| 733 |
+
decode_output_name: "kv_cache_v_16"
|
| 734 |
+
type: TYPE_GLOBAL_VALUE_CACHE
|
| 735 |
+
sequence_axis: 2
|
| 736 |
+
maximum_sequence_length: 1280
|
| 737 |
+
}
|
| 738 |
+
state_buffers {
|
| 739 |
+
prefill_input_name: "kv_cache_k_17"
|
| 740 |
+
prefill_output_name: "kv_cache_k_17"
|
| 741 |
+
decode_input_name: "kv_cache_k_17"
|
| 742 |
+
decode_output_name: "kv_cache_k_17"
|
| 743 |
+
type: TYPE_GLOBAL_KEY_CACHE
|
| 744 |
+
sequence_axis: 3
|
| 745 |
+
maximum_sequence_length: 1280
|
| 746 |
+
}
|
| 747 |
+
state_buffers {
|
| 748 |
+
prefill_input_name: "kv_cache_v_17"
|
| 749 |
+
prefill_output_name: "kv_cache_v_17"
|
| 750 |
+
decode_input_name: "kv_cache_v_17"
|
| 751 |
+
decode_output_name: "kv_cache_v_17"
|
| 752 |
+
type: TYPE_GLOBAL_VALUE_CACHE
|
| 753 |
+
sequence_axis: 2
|
| 754 |
+
maximum_sequence_length: 1280
|
| 755 |
+
}
|
| 756 |
+
}
|
| 757 |
+
>>>>>>>> end of ExecutorMetadata
|
| 758 |
+
ExecutorMetadataProto.pbtext dumped to: /tmp/tmpsrxymt_f/ExecutorMetadataProto.pbtext
|
| 759 |
+
|
| 760 |
+
|
| 761 |
+
Section 2:
|
| 762 |
+
Items:
|
| 763 |
+
<None>
|
| 764 |
+
Begin Offset: 49152
|
| 765 |
+
End Offset: 5635536
|
| 766 |
+
Data Type: HF_Tokenizer_Zlib
|
| 767 |
+
Section2_HF_Tokenizer_Zlib dumped to: /tmp/tmpsrxymt_f/Section2_HF_Tokenizer_Zlib.zlib
|
| 768 |
+
|
| 769 |
+
|
| 770 |
+
Section 3:
|
| 771 |
+
Items:
|
| 772 |
+
Key: model_type, Value (String): tf_lite_prefill_decode
|
| 773 |
+
Begin Offset: 5636096
|
| 774 |
+
End Offset: 282505680
|
| 775 |
+
Data Type: TFLiteModel
|
| 776 |
+
Section3_TFLiteModel_tf_lite_prefill_decode.tflite dumped to: /tmp/tmpsrxymt_f/Section3_TFLiteModel_tf_lite_prefill_decode.tflite
|
| 777 |
+
|
| 778 |
+
|
| 779 |
+
Section 4:
|
| 780 |
+
Items:
|
| 781 |
+
Key: model_type, Value (String): tf_lite_embedder
|
| 782 |
+
Begin Offset: 282509312
|
| 783 |
+
End Offset: 456578352
|
| 784 |
+
Data Type: TFLiteModel
|
| 785 |
+
Section4_TFLiteModel_tf_lite_embedder.tflite dumped to: /tmp/tmpsrxymt_f/Section4_TFLiteModel_tf_lite_embedder.tflite
|
| 786 |
+
|
| 787 |
+
|
| 788 |
+
Section 5:
|
| 789 |
+
Items:
|
| 790 |
+
Key: model_type, Value (String): tf_lite_aux
|
| 791 |
+
Begin Offset: 456589312
|
| 792 |
+
End Offset: 456704036
|
| 793 |
+
Data Type: TFLiteModel
|
| 794 |
+
Section5_TFLiteModel_tf_lite_aux.tflite dumped to: /tmp/tmpsrxymt_f/Section5_TFLiteModel_tf_lite_aux.tflite
|
| 795 |
+
|
| 796 |
+
|
| 797 |
+
RC=1
|
| 798 |
+
STDERR:
|
| 799 |
+
Skipping import of cpp extensions due to incompatible torch version. Please upgrade to torch >= 2.11.0 (found 2.10.0+cpu).
|
| 800 |
+
|
| 801 |
+
Loading weights: 0%| | 0/236 [00:00<?, ?it/s]
|
| 802 |
+
Loading weights: 100%|██████████| 236/236 [00:00<00:00, 5467.58it/s]
|
| 803 |
+
/usr/lib/python3.12/contextlib.py:144: UserWarning: The tensor attributes self.model.model.rotary_emb.data['sliding_attention'][0], self.model.model.rotary_emb.data['sliding_attention'][1], self.model.model.rotary_emb.data['full_attention'][0], self.model.model.rotary_emb.data['full_attention'][1] were assigned during export. Such attributes must be registered as buffers using the `register_buffer` API (https://pytorch.org/docs/stable/generated/torch.nn.Module.html#torch.nn.Module.register_buffer).
|
| 804 |
+
next(self.gen)
|
| 805 |
+
|
| 806 |
+
Generating Quantization Parameters:: 0%| | 0/3585 [00:00<?, ?it/s]
|
| 807 |
+
Generating Quantization Parameters:: 7%|▋ | 255/3585 [00:00<00:01, 2472.70it/s]
|
| 808 |
+
Generating Quantization Parameters:: 14%|█▍ | 507/3585 [00:00<00:01, 2471.51it/s]
|
| 809 |
+
Generating Quantization Parameters:: 21%|██ | 755/3585 [00:00<00:01, 2461.65it/s]
|
| 810 |
+
Generating Quantization Parameters:: 42%|████▏ | 1510/3585 [00:00<00:00, 4446.10it/s]
|
| 811 |
+
Generating Quantization Parameters:: 55%|█████▍ | 1957/3585 [00:01<00:01, 1280.67it/s]
|
| 812 |
+
Generating Quantization Parameters:: 100%|██████████| 3585/3585 [00:01<00:00, 2799.89it/s]
|
| 813 |
+
|
| 814 |
+
Applying Transformations to tensors:: 0%| | 0/4706 [00:00<?, ?it/s]
|
| 815 |
+
Applying Transformations to tensors:: 100%|██████████| 4706/4706 [00:00<00:00, 162219.60it/s]
|
| 816 |
+
WARNING:root:The model /kaggle/working/export_npu/model_quantized.tflite already exists in the folder. Overwriting the model since overwrite=True.
|
| 817 |
+
WARNING:root:The model /kaggle/working/export_npu/embedder_quantized.tflite already exists in the folder. Overwriting the model since overwrite=True.
|
| 818 |
+
ERROR:absl:Model path /tmp/tmpsrxymt_f/Section3_TFLiteModel_tf_lite_prefill_decode.tflite size: 276869584
|
| 819 |
+
ERROR:absl:Failed to compile prefill_decode: apply_plugin failed to apply plugin. See /tmp/tmpgseixfm9.error for details.
|
| 820 |
+
Traceback (most recent call last):
|
| 821 |
+
File "/usr/local/bin/litert-torch", line 8, in <module>
|
| 822 |
+
sys.exit(main())
|
| 823 |
+
^^^^^^
|
| 824 |
+
File "/usr/local/lib/python3.12/dist-packages/litert_torch/cli.py", line 30, in main
|
| 825 |
+
fire.Fire(CLI())
|
| 826 |
+
File "/usr/local/lib/python3.12/dist-packages/fire/core.py", line 135, in Fire
|
| 827 |
+
component_trace = _Fire(component, args, parsed_flag_args, context, name)
|
| 828 |
+
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
| 829 |
+
File "/usr/local/lib/python3.12/dist-packages/fire/core.py", line 468, in _Fire
|
| 830 |
+
component, remaining_args = _CallAndUpdateTrace(
|
| 831 |
+
^^^^^^^^^^^^^^^^^^^^
|
| 832 |
+
File "/usr/local/lib/python3.12/dist-packages/fire/core.py", line 684, in _CallAndUpdateTrace
|
| 833 |
+
component = fn(*varargs, **kwargs)
|
| 834 |
+
^^^^^^^^^^^^^^^^^^^^^^
|
| 835 |
+
File "/usr/local/lib/python3.12/dist-packages/litert_torch/generative/export_hf/export.py", line 361, in export
|
| 836 |
+
export_utils.run_npu_compilation_pipeline(
|
| 837 |
+
File "/usr/local/lib/python3.12/dist-packages/litert_torch/generative/export_hf/export_utils.py", line 147, in run_npu_compilation_pipeline
|
| 838 |
+
compile_func(
|
| 839 |
+
File "/usr/local/lib/python3.12/dist-packages/litert_torch/generative/export_hf/experimental/litert_lm_npu_compiler/litert_lm_npu_compiler.py", line 406, in compile_litertlm
|
| 840 |
+
compiler(
|
| 841 |
+
File "/usr/local/lib/python3.12/dist-packages/ai_edge_litert/aot/core/apply_plugin.py", line 146, in __call__
|
| 842 |
+
raise ValueError(
|
| 843 |
+
ValueError: apply_plugin failed to apply plugin. See /tmp/tmpgseixfm9.error for details.
|
| 844 |
+
|
| 845 |
+
litertlm-Dateien: ['/kaggle/working/export_npu/model.litertlm']
|
| 846 |
+
### NPU-Export mit SM8750 fehlgeschlagen (rc=1)
|
| 847 |
+
|