Qwen3.8-Flash-Next-mlx-bf16

Qwen/Qwen3.8-Flash-Next ใ‚’ MLX ๅฝขๅผใธๅค‰ๆ›ใ—ใŸใ‚‚ใฎ๏ผˆbf16ใƒป้ž้‡ๅญๅŒ–๏ผ‰ใ€‚

  • ๅค‰ๆ›ๅ…ƒ: Qwen/Qwen3.8-Flash-Next๏ผˆbf16 ๅ…ฌๅผ้‡ใฟ, revision de4b8e4๏ผ‰
  • ๅค‰ๆ›ใƒ„ใƒผใƒซ: mlx-vlm 0.6.17 / mlx 0.32.0
  • ใ‚ตใ‚คใ‚บ: 331 GiB

ๅค‰ๆ›ใ‚ณใƒžใƒณใƒ‰

python -m mlx_vlm convert --hf-path Qwen/Qwen3.8-Flash-Next \
    --mlx-path Qwen3.8-Flash-Next-mlx-bf16 --dtype bfloat16

ไฝฟใ„ๆ–น

pip install -U mlx-vlm
python -m mlx_vlm generate --model ToPo-ToPo/Qwen3.8-Flash-Next-mlx-bf16 \
    --prompt "ใ“ใฎ็”ปๅƒใ‚’่ชฌๆ˜Žใ—ใฆใใ ใ•ใ„" --image path/to/image.jpg --max-tokens 512

ๆŽจ่ซ–ใซใฏ 355 GB ่ถ…ใฎใƒกใƒขใƒชใŒ่ฆใ‚‹ใ€‚้‡ๅญๅŒ–็‰ˆใฏ 4bit๏ผˆ111 GB๏ผ‰/ 8bit๏ผˆ200 GB๏ผ‰ใ€‚

MTP ใƒ˜ใƒƒใƒ‰๏ผˆmtp.*๏ผ‰ใฏๅค‰ๆ›ๆ™‚ใซ้™คๅค–ใ•ใ‚Œใ‚‹ใ€‚ๆŠ•ๆฉŸใƒ‡ใ‚ณใƒผใƒ‰ใซไฝฟใ†ๅ ดๅˆใฏๅ…ฌๅผ bf16 ใ‹ใ‚‰ ๅˆฅ้€”ๅˆ‡ใ‚Šๅ‡บใ™ใ“ใจใ€‚

Downloads last month
162
Safetensors
Model size
177B params
Tensor type
BF16
ยท
I64
ยท
MLX
Hardware compatibility
Log In to add your hardware

Quantized

Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for ToPo-ToPo/Qwen3.8-Flash-Next-mlx-bf16

Finetuned
(141)
this model