Text Generation
Transformers
Safetensors
GGUF
hy_v4_flash
hy4-flash
weight-editing
weight-space-programming
reasoning
nvfp4
Mixture of Experts
conversational
8-bit precision
Instructions to use autotrust/hy4-flash-concise with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use autotrust/hy4-flash-concise with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="autotrust/hy4-flash-concise") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# pip install -U transformers accelerate # Load model directly from transformers import AutoModelForSeq2SeqLM model = AutoModelForSeq2SeqLM.from_pretrained("autotrust/hy4-flash-concise", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use autotrust/hy4-flash-concise with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "autotrust/hy4-flash-concise" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "autotrust/hy4-flash-concise", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/autotrust/hy4-flash-concise
- SGLang
How to use autotrust/hy4-flash-concise with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "autotrust/hy4-flash-concise" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "autotrust/hy4-flash-concise", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "autotrust/hy4-flash-concise" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "autotrust/hy4-flash-concise", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use autotrust/hy4-flash-concise with Docker Model Runner:
docker model run hf.co/autotrust/hy4-flash-concise
Download hf_quant_config.json from autotrust/hy4-flash-concise: direct link, hf CLI and curl.
- Browser
- Download file 13.8 kB
-
https://huggingface.co/autotrust/hy4-flash-concise/resolve/main/hf_quant_config.json
- Command line
-
hf download hf://autotrust/hy4-flash-concise/hf_quant_config.json
-
curl -L -o hf_quant_config.json https://huggingface.co/autotrust/hy4-flash-concise/resolve/main/hf_quant_config.json
13.8 kB
| { | |
| "producer": { | |
| "name": "modelopt", | |
| "version": "0.48.0" | |
| }, | |
| "quantization": { | |
| "quant_algo": "NVFP4", | |
| "kv_cache_quant_algo": null, | |
| "group_size": 16, | |
| "exclude_modules": [ | |
| "lm_head", | |
| "model.dfly_layers.0.input_layernorm", | |
| "model.dfly_layers.0.mlp.down_proj", | |
| "model.dfly_layers.0.mlp.gate_proj", | |
| "model.dfly_layers.0.mlp.up_proj", | |
| "model.dfly_layers.0.post_attention_layernorm", | |
| "model.dfly_layers.0.self_attn", | |
| "model.dfly_layers.0.self_attn.kv_a_layernorm", | |
| "model.dfly_layers.0.self_attn.kv_a_proj_with_mqa", | |
| "model.dfly_layers.0.self_attn.kv_b_proj", | |
| "model.dfly_layers.0.self_attn.linear_gate", | |
| "model.dfly_layers.0.self_attn.o_proj", | |
| "model.dfly_layers.0.self_attn.post_proj_norm", | |
| "model.dfly_layers.0.self_attn.q_a_layernorm", | |
| "model.dfly_layers.0.self_attn.q_a_proj", | |
| "model.dfly_layers.0.self_attn.q_b_proj", | |
| "model.dfly_layers.1.input_layernorm", | |
| "model.dfly_layers.1.mlp.down_proj", | |
| "model.dfly_layers.1.mlp.gate_proj", | |
| "model.dfly_layers.1.mlp.up_proj", | |
| "model.dfly_layers.1.post_attention_layernorm", | |
| "model.dfly_layers.1.self_attn", | |
| "model.dfly_layers.1.self_attn.kv_a_layernorm", | |
| "model.dfly_layers.1.self_attn.kv_a_proj_with_mqa", | |
| "model.dfly_layers.1.self_attn.kv_b_proj", | |
| "model.dfly_layers.1.self_attn.linear_gate", | |
| "model.dfly_layers.1.self_attn.o_proj", | |
| "model.dfly_layers.1.self_attn.post_proj_norm", | |
| "model.dfly_layers.1.self_attn.q_a_layernorm", | |
| "model.dfly_layers.1.self_attn.q_a_proj", | |
| "model.dfly_layers.1.self_attn.q_b_proj", | |
| "model.dfly_layers.2.input_layernorm", | |
| "model.dfly_layers.2.mlp.down_proj", | |
| "model.dfly_layers.2.mlp.gate_proj", | |
| "model.dfly_layers.2.mlp.up_proj", | |
| "model.dfly_layers.2.post_attention_layernorm", | |
| "model.dfly_layers.2.self_attn", | |
| "model.dfly_layers.2.self_attn.kv_a_layernorm", | |
| "model.dfly_layers.2.self_attn.kv_a_proj_with_mqa", | |
| "model.dfly_layers.2.self_attn.kv_b_proj", | |
| "model.dfly_layers.2.self_attn.linear_gate", | |
| "model.dfly_layers.2.self_attn.o_proj", | |
| "model.dfly_layers.2.self_attn.post_proj_norm", | |
| "model.dfly_layers.2.self_attn.q_a_layernorm", | |
| "model.dfly_layers.2.self_attn.q_a_proj", | |
| "model.dfly_layers.2.self_attn.q_b_proj", | |
| "model.dfly_layers.3.input_layernorm", | |
| "model.dfly_layers.3.mlp.down_proj", | |
| "model.dfly_layers.3.mlp.gate_proj", | |
| "model.dfly_layers.3.mlp.up_proj", | |
| "model.dfly_layers.3.post_attention_layernorm", | |
| "model.dfly_layers.3.self_attn", | |
| "model.dfly_layers.3.self_attn.kv_a_layernorm", | |
| "model.dfly_layers.3.self_attn.kv_a_proj_with_mqa", | |
| "model.dfly_layers.3.self_attn.kv_b_proj", | |
| "model.dfly_layers.3.self_attn.linear_gate", | |
| "model.dfly_layers.3.self_attn.o_proj", | |
| "model.dfly_layers.3.self_attn.post_proj_norm", | |
| "model.dfly_layers.3.self_attn.q_a_layernorm", | |
| "model.dfly_layers.3.self_attn.q_a_proj", | |
| "model.dfly_layers.3.self_attn.q_b_proj", | |
| "model.dfly_layers.4.input_layernorm", | |
| "model.dfly_layers.4.mlp.down_proj", | |
| "model.dfly_layers.4.mlp.gate_proj", | |
| "model.dfly_layers.4.mlp.up_proj", | |
| "model.dfly_layers.4.post_attention_layernorm", | |
| "model.dfly_layers.4.self_attn", | |
| "model.dfly_layers.4.self_attn.kv_a_layernorm", | |
| "model.dfly_layers.4.self_attn.kv_a_proj_with_mqa", | |
| "model.dfly_layers.4.self_attn.kv_b_proj", | |
| "model.dfly_layers.4.self_attn.linear_gate", | |
| "model.dfly_layers.4.self_attn.o_proj", | |
| "model.dfly_layers.4.self_attn.post_proj_norm", | |
| "model.dfly_layers.4.self_attn.q_a_layernorm", | |
| "model.dfly_layers.4.self_attn.q_a_proj", | |
| "model.dfly_layers.4.self_attn.q_b_proj", | |
| "model.dfly_layers.confidence_head.proj", | |
| "model.dfly_layers.final_norm", | |
| "model.dfly_layers.hidden_correction.down_proj", | |
| "model.dfly_layers.hidden_correction.embed_norm", | |
| "model.dfly_layers.hidden_correction.gate_proj", | |
| "model.dfly_layers.hidden_correction.hidden_norm", | |
| "model.dfly_layers.hidden_correction.up_proj", | |
| "model.dfly_layers.mtp_fusion", | |
| "model.dfly_layers.mtp_fusion.eh_proj", | |
| "model.dfly_layers.mtp_fusion.enorm", | |
| "model.dfly_layers.mtp_fusion.hnorm", | |
| "model.embed_tokens", | |
| "model.layers.0.self_attn*", | |
| "model.layers.1.engram_layer*", | |
| "model.layers.1.mlp.gate", | |
| "model.layers.1.mlp.routed_expert_down_proj", | |
| "model.layers.1.mlp.routed_expert_up_proj", | |
| "model.layers.1.self_attn*", | |
| "model.layers.10.mlp.gate", | |
| "model.layers.10.mlp.routed_expert_down_proj", | |
| "model.layers.10.mlp.routed_expert_up_proj", | |
| "model.layers.10.self_attn*", | |
| "model.layers.11.engram_layer*", | |
| "model.layers.11.mlp.gate", | |
| "model.layers.11.mlp.routed_expert_down_proj", | |
| "model.layers.11.mlp.routed_expert_up_proj", | |
| "model.layers.11.self_attn*", | |
| "model.layers.12.mlp.gate", | |
| "model.layers.12.mlp.routed_expert_down_proj", | |
| "model.layers.12.mlp.routed_expert_up_proj", | |
| "model.layers.12.self_attn*", | |
| "model.layers.13.mlp.gate", | |
| "model.layers.13.mlp.routed_expert_down_proj", | |
| "model.layers.13.mlp.routed_expert_up_proj", | |
| "model.layers.13.self_attn*", | |
| "model.layers.14.mlp.gate", | |
| "model.layers.14.mlp.routed_expert_down_proj", | |
| "model.layers.14.mlp.routed_expert_up_proj", | |
| "model.layers.14.self_attn*", | |
| "model.layers.15.mlp.gate", | |
| "model.layers.15.mlp.routed_expert_down_proj", | |
| "model.layers.15.mlp.routed_expert_up_proj", | |
| "model.layers.15.self_attn*", | |
| "model.layers.16.mlp.gate", | |
| "model.layers.16.mlp.routed_expert_down_proj", | |
| "model.layers.16.mlp.routed_expert_up_proj", | |
| "model.layers.16.self_attn*", | |
| "model.layers.17.mlp.gate", | |
| "model.layers.17.mlp.routed_expert_down_proj", | |
| "model.layers.17.mlp.routed_expert_up_proj", | |
| "model.layers.17.self_attn*", | |
| "model.layers.18.mlp.gate", | |
| "model.layers.18.mlp.routed_expert_down_proj", | |
| "model.layers.18.mlp.routed_expert_up_proj", | |
| "model.layers.18.self_attn*", | |
| "model.layers.19.mlp.gate", | |
| "model.layers.19.mlp.routed_expert_down_proj", | |
| "model.layers.19.mlp.routed_expert_up_proj", | |
| "model.layers.19.self_attn*", | |
| "model.layers.2.mlp.gate", | |
| "model.layers.2.mlp.routed_expert_down_proj", | |
| "model.layers.2.mlp.routed_expert_up_proj", | |
| "model.layers.2.self_attn*", | |
| "model.layers.20.mlp.gate", | |
| "model.layers.20.mlp.routed_expert_down_proj", | |
| "model.layers.20.mlp.routed_expert_up_proj", | |
| "model.layers.20.self_attn*", | |
| "model.layers.21.engram_layer*", | |
| "model.layers.21.mlp.gate", | |
| "model.layers.21.mlp.routed_expert_down_proj", | |
| "model.layers.21.mlp.routed_expert_up_proj", | |
| "model.layers.21.self_attn*", | |
| "model.layers.22.mlp.gate", | |
| "model.layers.22.mlp.routed_expert_down_proj", | |
| "model.layers.22.mlp.routed_expert_up_proj", | |
| "model.layers.22.self_attn*", | |
| "model.layers.23.mlp.gate", | |
| "model.layers.23.mlp.routed_expert_down_proj", | |
| "model.layers.23.mlp.routed_expert_up_proj", | |
| "model.layers.23.self_attn*", | |
| "model.layers.24.mlp.gate", | |
| "model.layers.24.mlp.routed_expert_down_proj", | |
| "model.layers.24.mlp.routed_expert_up_proj", | |
| "model.layers.24.self_attn*", | |
| "model.layers.25.mlp.gate", | |
| "model.layers.25.mlp.routed_expert_down_proj", | |
| "model.layers.25.mlp.routed_expert_up_proj", | |
| "model.layers.25.self_attn*", | |
| "model.layers.26.mlp.gate", | |
| "model.layers.26.mlp.routed_expert_down_proj", | |
| "model.layers.26.mlp.routed_expert_up_proj", | |
| "model.layers.26.self_attn*", | |
| "model.layers.27.mlp.gate", | |
| "model.layers.27.mlp.routed_expert_down_proj", | |
| "model.layers.27.mlp.routed_expert_up_proj", | |
| "model.layers.27.self_attn*", | |
| "model.layers.28.mlp.gate", | |
| "model.layers.28.mlp.routed_expert_down_proj", | |
| "model.layers.28.mlp.routed_expert_up_proj", | |
| "model.layers.28.self_attn*", | |
| "model.layers.29.mlp.gate", | |
| "model.layers.29.mlp.routed_expert_down_proj", | |
| "model.layers.29.mlp.routed_expert_up_proj", | |
| "model.layers.29.self_attn*", | |
| "model.layers.3.mlp.gate", | |
| "model.layers.3.mlp.routed_expert_down_proj", | |
| "model.layers.3.mlp.routed_expert_up_proj", | |
| "model.layers.3.self_attn*", | |
| "model.layers.30.mlp.gate", | |
| "model.layers.30.mlp.routed_expert_down_proj", | |
| "model.layers.30.mlp.routed_expert_up_proj", | |
| "model.layers.30.self_attn*", | |
| "model.layers.31.engram_layer*", | |
| "model.layers.31.mlp.gate", | |
| "model.layers.31.mlp.routed_expert_down_proj", | |
| "model.layers.31.mlp.routed_expert_up_proj", | |
| "model.layers.31.self_attn*", | |
| "model.layers.32.mlp.gate", | |
| "model.layers.32.mlp.routed_expert_down_proj", | |
| "model.layers.32.mlp.routed_expert_up_proj", | |
| "model.layers.32.self_attn*", | |
| "model.layers.33.mlp.gate", | |
| "model.layers.33.mlp.routed_expert_down_proj", | |
| "model.layers.33.mlp.routed_expert_up_proj", | |
| "model.layers.33.self_attn*", | |
| "model.layers.34.mlp.gate", | |
| "model.layers.34.mlp.routed_expert_down_proj", | |
| "model.layers.34.mlp.routed_expert_up_proj", | |
| "model.layers.34.self_attn*", | |
| "model.layers.35.mlp.gate", | |
| "model.layers.35.mlp.routed_expert_down_proj", | |
| "model.layers.35.mlp.routed_expert_up_proj", | |
| "model.layers.35.self_attn*", | |
| "model.layers.36.mlp.gate", | |
| "model.layers.36.mlp.routed_expert_down_proj", | |
| "model.layers.36.mlp.routed_expert_up_proj", | |
| "model.layers.36.self_attn*", | |
| "model.layers.37.mlp.gate", | |
| "model.layers.37.mlp.routed_expert_down_proj", | |
| "model.layers.37.mlp.routed_expert_up_proj", | |
| "model.layers.37.self_attn*", | |
| "model.layers.38.mlp.gate", | |
| "model.layers.38.mlp.routed_expert_down_proj", | |
| "model.layers.38.mlp.routed_expert_up_proj", | |
| "model.layers.38.self_attn*", | |
| "model.layers.39.mlp.gate", | |
| "model.layers.39.mlp.routed_expert_down_proj", | |
| "model.layers.39.mlp.routed_expert_up_proj", | |
| "model.layers.39.self_attn*", | |
| "model.layers.4.mlp.gate", | |
| "model.layers.4.mlp.routed_expert_down_proj", | |
| "model.layers.4.mlp.routed_expert_up_proj", | |
| "model.layers.4.self_attn*", | |
| "model.layers.40.mlp.gate", | |
| "model.layers.40.mlp.routed_expert_down_proj", | |
| "model.layers.40.mlp.routed_expert_up_proj", | |
| "model.layers.40.self_attn*", | |
| "model.layers.41.mlp.gate", | |
| "model.layers.41.mlp.routed_expert_down_proj", | |
| "model.layers.41.mlp.routed_expert_up_proj", | |
| "model.layers.41.self_attn*", | |
| "model.layers.5.mlp.gate", | |
| "model.layers.5.mlp.routed_expert_down_proj", | |
| "model.layers.5.mlp.routed_expert_up_proj", | |
| "model.layers.5.self_attn*", | |
| "model.layers.6.mlp.gate", | |
| "model.layers.6.mlp.routed_expert_down_proj", | |
| "model.layers.6.mlp.routed_expert_up_proj", | |
| "model.layers.6.self_attn*", | |
| "model.layers.7.mlp.gate", | |
| "model.layers.7.mlp.routed_expert_down_proj", | |
| "model.layers.7.mlp.routed_expert_up_proj", | |
| "model.layers.7.self_attn*", | |
| "model.layers.8.mlp.gate", | |
| "model.layers.8.mlp.routed_expert_down_proj", | |
| "model.layers.8.mlp.routed_expert_up_proj", | |
| "model.layers.8.self_attn*", | |
| "model.layers.9.mlp.gate", | |
| "model.layers.9.mlp.routed_expert_down_proj", | |
| "model.layers.9.mlp.routed_expert_up_proj", | |
| "model.layers.9.self_attn*", | |
| "vit*", | |
| "model.layers.22.mlp.shared_experts.down_proj", | |
| "model.layers.24.mlp.shared_experts.down_proj", | |
| "model.layers.26.mlp.shared_experts.down_proj", | |
| "model.layers.28.mlp.shared_experts.down_proj", | |
| "model.layers.30.mlp.shared_experts.down_proj", | |
| "model.layers.32.mlp.shared_experts.down_proj", | |
| "model.layers.34.mlp.shared_experts.down_proj", | |
| "model.layers.36.mlp.shared_experts.down_proj", | |
| "model.layers.38.mlp.shared_experts.down_proj" | |
| ] | |
| } | |
| } |