Hugging Face's logo

bhanc
/
scratch

ExecuTorch
Model card Files
xet
 Community
scratch / llms

Commit History

30 days ago

  1. Add Gemma 4 Omni E2B MLX int4 model

    Bartosz Hanc
    cb2e07e

  2. Upload llms/gemma4_e2b_mlx_int4.pte with huggingface_hub

    bhanc
    Verified signature
    be94526

  3. Upload llms/gemma_4_e2b_vulkan_8da4w.pte with huggingface_hub

    bhanc
    Verified signature
    ebb364f

  4. Add Gemma 4 E2B XNNPACK 8da4w re-exported for the TypeScript runner (text-only, ctx 4096, seq 1024, fp32 [1, vocab] logits, get_model_schema)

    Bartosz Hanc
    4678943

  5. Re-export llama_3_2_3b_mlx_int4.pte with the SWM ExecuTorch fork (ctx 4096, seq 2048, fp32 logits, get_model_schema)

    Bartosz Hanc
    98e6607

  6. Re-export llama_3_2_1b_mlx_int4.pte with the SWM ExecuTorch fork (ctx 16384, seq 2048, fp32 logits, get_model_schema)

    Bartosz Hanc
    5fe15ce

  7. Re-export qwen_3_4b_mlx_int4.pte with the SWM ExecuTorch fork (ctx 4096, seq 2048, fp32 logits, get_model_schema)

    Bartosz Hanc
    f2c24b3

  8. Re-export phi_4_mini_mlx_int4.pte with the SWM ExecuTorch fork (ctx 4096, seq 2048, fp32 logits, get_model_schema)

    Bartosz Hanc
    447617d

  9. Re-export hammer_2_1_3b_mlx_int4.pte with the SWM ExecuTorch fork (ctx 8192, seq 2048, fp32 logits, get_model_schema)

    Bartosz Hanc
    ee9b2a2

  10. Re-export qwen_2_5_3b_mlx_int4.pte with the SWM ExecuTorch fork (ctx 8192, seq 2048, fp32 logits, get_model_schema)

    Bartosz Hanc
    abbacfb

  11. Re-export qwen_3_1_7b_mlx_int4.pte with the SWM ExecuTorch fork (ctx 4096, seq 2048, fp32 logits, get_model_schema)

    Bartosz Hanc
    900b335

  12. Re-export hammer_2_1_1_5b_mlx_int4.pte with the SWM ExecuTorch fork (ctx 16384, seq 2048, fp32 logits, get_model_schema)

    Bartosz Hanc
    05ac95a

  13. Re-export qwen_2_5_1_5b_mlx_int4.pte with the SWM ExecuTorch fork (ctx 16384, seq 2048, fp32 logits, get_model_schema)

    Bartosz Hanc
    0470455

  14. Re-export llama_3_2_1b_mlx_int4.pte with the SWM ExecuTorch fork (ctx 16384, seq 4096, fp32 logits, get_model_schema)

    Bartosz Hanc
    cbfa64b

  15. Re-export qwen_3_0_6b_mlx_int4.pte with the SWM ExecuTorch fork (ctx 8192, seq 2048, fp32 logits, get_model_schema)

    Bartosz Hanc
    703723a

  16. Re-export hammer_2_1_0_5b_mlx_int4.pte with the SWM ExecuTorch fork (ctx 16384, seq 2048, fp32 logits, get_model_schema)

    Bartosz Hanc
    ef0d01e

  17. Re-export qwen_2_5_0_5b_mlx_int4.pte with the SWM ExecuTorch fork (ctx 16384, seq 2048, fp32 logits, get_model_schema)

    Bartosz Hanc
    df54b6e

  18. Re-export lfm_2_5_1_2b_mlx_int4.pte with the SWM ExecuTorch fork (ctx 16384, seq 65, fp32 logits, get_model_schema)

    Bartosz Hanc
    bef4f8b

  19. Re-export lfm_2_5_350m_mlx_int4.pte with the SWM ExecuTorch fork (ctx 16384, seq 65, fp32 logits, get_model_schema)

    Bartosz Hanc
    6251fc6

  20. Add LFM 2.5 VL 1.6B Vulkan 8da4w re-exported with the SWM ExecuTorch 1.4.1 fork (get_model_schema)

    Bartosz Hanc
    0f2a51c

  21. Add LFM 2.5 VL 450M Vulkan 8da4w re-exported with the SWM ExecuTorch 1.4.1 fork (get_model_schema)

    Bartosz Hanc
    fa1d83e

  22. Add LFM 2.5 VL 1.6B MLX int8 re-exported with the SWM ExecuTorch fork (fp32 method boundaries, get_model_schema)

    Bartosz Hanc
    ee2e023

  23. Add LFM 2.5 VL 1.6B MLX int4 re-exported with the SWM ExecuTorch fork (fp32 method boundaries, get_model_schema)

    Bartosz Hanc
    8074e2c

  24. Add LFM 2.5 VL 450M MLX int4 re-exported with the SWM ExecuTorch fork (fp32 method boundaries, get_model_schema)

    Bartosz Hanc
    9d331da

  25. Add SmolLM2 1.7B MLX int8 re-exported with the SWM ExecuTorch fork (2k context, fp32 logits, get_model_schema)

    Bartosz Hanc
    178f30f

  26. Add SmolLM2 135M/360M MLX int8 re-exported with the SWM ExecuTorch fork (16k context, fp32 logits, get_model_schema)

    Bartosz Hanc
    c05de5d

  27. Rename LLM exports to their canonical react-native-executorch file names

    Bartosz Hanc
    35dabfc

  28. Add lfm_2_5_vl XNNPACK models with get_model_schema (2k context)

    Bartosz Hanc
    bd37518

  1. Add re-exported qwen_3_4b_mlx_4w_4k.pte with get_model_schema

    Bartosz Hanc
    57a5dbf

  2. Add qwen_3_1_7b_mlx_4w_4k.pte

    Bartosz Hanc
    823429d

  3. Add qwen_3_0_6b_mlx_4w_8k.pte

    Bartosz Hanc
    1bd16a1

  4. Add qwen_2_5_3b_mlx_4w_8k.pte

    Bartosz Hanc
    ad6182f

  5. Add qwen_2_5_1_5b_mlx_4w_16k.pte

    Bartosz Hanc
    435f19b

  6. Add qwen_2_5_0_5b_mlx_4w_16k.pte

    Bartosz Hanc
    37991a6

  7. Add phi_4_mini_3_8b_mlx_4w_4k.pte

    Bartosz Hanc
    8d8d3fc

  8. Add llama_3_2_3b_mlx_4w_4k.pte

    Bartosz Hanc
    b96b77f

  9. Add hammer_2_1_3b_mlx_4w_8k.pte

    Bartosz Hanc
    90ce79e

  10. Add hammer_2_1_1_5b_mlx_4w_16k.pte

    Bartosz Hanc
    b9115f8

  11. Add re-exported 16k MLX models with get_model_schema (LFM 350M/1.2B, Llama 3.2 1B, Hammer 0.5B)

    Bartosz Hanc
    2f34584

  1. fix(llms): correct EOS and BOS IDs for Bielik v3 and Phi-4 Mini

    Bartosz Hanc
    bed1b8d

  2. Update target LLMs: enforce seq_len=2048 with target contexts

    Bartosz Hanc
    3f3fd9c

  3. Add target-context XNNPACK LLM models with companion schema

    Bartosz Hanc
    bb1c6df