Hugging Face's logo

FastFlowLM
/
D3-Mini-NPU2

Zero-Shot Classification
qwen3_5_text
decision-model
system-one
fastflowlm
amd-npu
q4nx
Model card Files
xet
 Community
FastFlowLM commited on
Commit
9ac5e6e
·
verified ·
1 Parent(s): ae90ba4

Card: FastFlowLM 1.0.11 or later

Browse files
Files changed (2) hide show
  1. MANIFEST.sha256 +1 -1
  2. README.md +1 -1
MANIFEST.sha256 CHANGED
@@ -4,7 +4,7 @@ adad3b9a7d9ca75d882ecb42139b76f45535b41e0d7498ee49e37c83e39f6f1a config.json
4
  c71d239df91726fc519c6eb72d318ec65820627232b2f796219e87dcf35d0ab4 LICENSE
5
  2b217d93ce444e22d4d6ce0460fa0f40524691b1738f962433068790dd3440a1 model.q4nx
6
  612acb7c7f714552de6e3255975e7e163f1ef275520d1171a3c8ddc74d239123 NOTICE
7
- 8e8f2f8319c63d3d65d8245f36189e7d6b52aa342b533088e2f0501f62f94bee README.md
8
  d570665c1da72cb9fca7ad7b23a25720eb691362694dd69d4c80c17d7656ed55 readout.safetensors
9
  316230d6a809701f4db5ea8f8fc862bc3a6f3229c937c174e674ff3ca0a64ac8 tokenizer_config.json
10
  5f9e4d4901a92b997e463c1f46055088b6cca5ca61a6522d1b9f64c4bb81cb42 tokenizer.json
 
4
  c71d239df91726fc519c6eb72d318ec65820627232b2f796219e87dcf35d0ab4 LICENSE
5
  2b217d93ce444e22d4d6ce0460fa0f40524691b1738f962433068790dd3440a1 model.q4nx
6
  612acb7c7f714552de6e3255975e7e163f1ef275520d1171a3c8ddc74d239123 NOTICE
7
+ 97d889e9fadfeba70df673b0ed2c174860ff65a7448bf81106d53ee721b72f66 README.md
8
  d570665c1da72cb9fca7ad7b23a25720eb691362694dd69d4c80c17d7656ed55 readout.safetensors
9
  316230d6a809701f4db5ea8f8fc862bc3a6f3229c937c174e674ff3ca0a64ac8 tokenizer_config.json
10
  5f9e4d4901a92b997e463c1f46055088b6cca5ca61a6522d1b9f64c4bb81cb42 tokenizer.json
README.md CHANGED
@@ -13,7 +13,7 @@ tags:
13
  # D3-Mini-NPU2
14
 
15
  [vllm-sr/d3-mini](https://huggingface.co/vllm-sr/d3-mini) `v3.1.3` (Decision 3.0, the decision model of vLLM Semantic
16
- Router; Qwen3.5-4B text tower) for AMD Ryzen AI NPUs, served by [FastFlowLM](https://github.com/ROCm/FastFlowLM) 1.0.11+.
17
  It answers *System One* requests -- choice, yes / no and score questions about a state, optionally with images, with a
18
  probability for every option -- over the same `POST /v1/systemone` API as the upstream package. Up to 32,768 input
19
  tokens per question. Video is not supported (a request with `videos` is refused with HTTP 422).
 
13
  # D3-Mini-NPU2
14
 
15
  [vllm-sr/d3-mini](https://huggingface.co/vllm-sr/d3-mini) `v3.1.3` (Decision 3.0, the decision model of vLLM Semantic
16
+ Router; Qwen3.5-4B text tower) for AMD Ryzen AI NPUs, served by [FastFlowLM](https://github.com/ROCm/FastFlowLM) 1.0.11 or later.
17
  It answers *System One* requests -- choice, yes / no and score questions about a state, optionally with images, with a
18
  probability for every option -- over the same `POST /v1/systemone` API as the upstream package. Up to 32,768 input
19
  tokens per question. Video is not supported (a request with `videos` is refused with HTTP 422).