Text Generation
Transformers
Safetensors
English
Japanese
qwen3_5
image-text-to-text
qwen3.5
reasoning
efficient-thinking
vision-validation-evaluated
full-weights
text-evaluated
conversational
Instructions to use horiuchinobuyuki/Qwick-3.5-9B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use horiuchinobuyuki/Qwick-3.5-9B with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="horiuchinobuyuki/Qwick-3.5-9B") messages = [ { "role": "user", "content": [ {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"}, {"type": "text", "text": "What animal is on the candy?"} ] }, ] pipe(text=messages)# Load model directly from transformers import AutoProcessor, AutoModelForMultimodalLM processor = AutoProcessor.from_pretrained("horiuchinobuyuki/Qwick-3.5-9B") model = AutoModelForMultimodalLM.from_pretrained("horiuchinobuyuki/Qwick-3.5-9B", device_map="auto") messages = [ { "role": "user", "content": [ {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"}, {"type": "text", "text": "What animal is on the candy?"} ] }, ] inputs = processor.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", ).to(model.device) outputs = model.generate(**inputs, max_new_tokens=40) print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use horiuchinobuyuki/Qwick-3.5-9B with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "horiuchinobuyuki/Qwick-3.5-9B" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "horiuchinobuyuki/Qwick-3.5-9B", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/horiuchinobuyuki/Qwick-3.5-9B
- SGLang
How to use horiuchinobuyuki/Qwick-3.5-9B 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 "horiuchinobuyuki/Qwick-3.5-9B" \ --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": "horiuchinobuyuki/Qwick-3.5-9B", "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 "horiuchinobuyuki/Qwick-3.5-9B" \ --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": "horiuchinobuyuki/Qwick-3.5-9B", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use horiuchinobuyuki/Qwick-3.5-9B with Docker Model Runner:
docker model run hf.co/horiuchinobuyuki/Qwick-3.5-9B
| { | |
| "artifact": { | |
| "architecture": "Qwen3_5ForConditionalGeneration", | |
| "format": "Transformers BF16 safetensors", | |
| "standalone": true, | |
| "weight_index_sha256": "sha256:3d2f0ab780828a41449b9f32fee63e6ee0d27bf98fab0e52aa982f81e6c47cec", | |
| "weight_shards": [ | |
| { | |
| "bytes": 2034237568, | |
| "name": "model-00001-of-00006.safetensors", | |
| "sha256": "sha256:dd63614f1dc80dce2d83be3f3e69af1f8a0ebc8add9b8e0abf3910f563e44344" | |
| }, | |
| { | |
| "bytes": 3999615808, | |
| "name": "model-00002-of-00006.safetensors", | |
| "sha256": "sha256:cbde68fec55b0602ce00941e78643a7ae3333f8d0a07f85403066cd7a4ca0788" | |
| }, | |
| { | |
| "bytes": 3997274128, | |
| "name": "model-00003-of-00006.safetensors", | |
| "sha256": "sha256:0c58eb5fd1ea60c29150a9f1cbdabbc1764c3b62c3f8550637eeb33fe764aaf7" | |
| }, | |
| { | |
| "bytes": 3997290904, | |
| "name": "model-00004-of-00006.safetensors", | |
| "sha256": "sha256:5d2f3c3932606a42eb21fbe79734ff7507c5bf02b066a3ecbfc42e962f0d82ed" | |
| }, | |
| { | |
| "bytes": 3991239264, | |
| "name": "model-00005-of-00006.safetensors", | |
| "sha256": "sha256:800b2058ea95da0dc1026a14e55f6a0c172563764bb0cff7bd121f4f9ad8d79b" | |
| }, | |
| { | |
| "bytes": 800062816, | |
| "name": "model-00006-of-00006.safetensors", | |
| "sha256": "sha256:db688e61fd7575e8537120bc7aaeff043d31d6ca2bafa14df1135c969a081013" | |
| } | |
| ] | |
| }, | |
| "author": { | |
| "name": "Nobuyuki Horiuchi", | |
| "profile": "https://huggingface.co/horiuchinobuyuki" | |
| }, | |
| "base_model": { | |
| "id": "Qwen/Qwen3.5-9B", | |
| "revision": "c202236235762e1c871ad0ccb60c8ee5ba337b9a" | |
| }, | |
| "candidate_id": "g7-nes-d04-p", | |
| "heldout": { | |
| "analysis_sha256": "sha256:a7584f10c35ed6eb60056da52c65d5719de730e7e6568eee184db9f1864567fe", | |
| "eligibility_audit_sha256": "sha256:336e39619eb1778a3af516b8b08f8ceaa122ee26c58b7b35ff5232318759ebfc", | |
| "measurement_interpretation": "The predeclared temperature-zero decode is valid for this frozen internal acceptance rule. It is not the temperature-1.0 measurement profile used for public benchmark claims and must not be compared as though it were.", | |
| "policy_sha256": "sha256:8a929e01ec7f5d60a5b14ff6664f74ce4d86b216d26155ca613e371183ef5701", | |
| "post_result_retuning": false, | |
| "public_performance_benchmark": false, | |
| "release_gate_pass": true, | |
| "rerun": false, | |
| "role": "frozen internal candidate-acceptance release gate", | |
| "used_for_training_or_selection": false, | |
| "used_only_for_postfreeze_final_evaluation": true | |
| }, | |
| "kind": "qwick_3_5_9b_merged_bf16_provenance", | |
| "known_risks": { | |
| "hmmt81920_observed_gate_pass": false, | |
| "hmmt81920_policy": "report_only_by_explicit_release_objective_amendment", | |
| "mmlu_pro_worst_subject": { | |
| "accuracy_delta": -0.02063983488132095, | |
| "base_correct": 753, | |
| "candidate_correct": 733, | |
| "n": 969, | |
| "policy": "reported subgroup risk; aggregate MMLU-Pro remains the frozen selection gate", | |
| "subject": "engineering" | |
| } | |
| }, | |
| "license": { | |
| "derived_model": "Apache-2.0", | |
| "upstream_model_card_declared_license": "apache-2.0", | |
| "upstream_url": "https://huggingface.co/Qwen/Qwen3.5-9B" | |
| }, | |
| "merge": { | |
| "dtype": "bfloat16", | |
| "maximum_shard_size": "4GB", | |
| "method": "PeftModel.merge_and_unload(safe_merge=True)", | |
| "revision": "sft-2a25037cd9c7bbda", | |
| "safe_serialization": true, | |
| "transformers_version": "5.14.1" | |
| }, | |
| "public_model": { | |
| "name": "Qwick-3.5-9B", | |
| "repo_id": "horiuchinobuyuki/Qwick-3.5-9B", | |
| "url": "https://huggingface.co/horiuchinobuyuki/Qwick-3.5-9B" | |
| }, | |
| "release_verification": { | |
| "holdout_release_gate_pass": true, | |
| "holdout_selection_decode_temperature": 0.0, | |
| "holdout_used_as_public_performance_benchmark": false, | |
| "ifeval_base_128k_budget_sensitivity_sha256": "sha256:44654e148106298934b6b2fce08ac3915af7fe2ee80d152700e96bfd62ed975d", | |
| "ifeval_base_128k_decode_audit_sha256": "sha256:02fed0a95d9141d50aee571eeb1f5012a205ac5b455fa33f79a36445486def5a", | |
| "ifeval_base_32k_length_subset_sha256": "sha256:113dfe01f366b13215b1d866979ccd3c857f381c6687262928b30464c5357822", | |
| "ifeval_corrected_decode_audit_sha256": "sha256:102d0d0f27d924be6ae7c248ab17d26c22d9ee83a57d3e94dd688b91fac47740", | |
| "ifeval_report_correction_sha256": "sha256:cc62405d3b198530c4070a8e5256543e064a3462576fe418b15c1d91ce1a1386", | |
| "ifeval_reselection_performed": false, | |
| "ifeval_temperature_zero_rows": 0, | |
| "notice_section_4d_propagation_required": false, | |
| "quantized_vision_validation_analysis_sha256": "sha256:9a47cdf0ce420beb0a715ed10e3ca36ae051160504876cd2d1d5d1bcd9ed6298", | |
| "quantized_vision_validation_candidate_reselection": false, | |
| "quantized_vision_validation_full_trace_missing": 0, | |
| "quantized_vision_validation_records_per_model": 900, | |
| "synthetic_holdout_rerun": false, | |
| "upstream_license_notice_evidence_sha256": "sha256:259822dde74e6ddaa522bb3dd93d4b5d11fe9413810efd419973c973cff941a3", | |
| "upstream_license_verified": "Apache-2.0", | |
| "upstream_notice_present": false, | |
| "upstream_revision_verified": "c202236235762e1c871ad0ccb60c8ee5ba337b9a", | |
| "vision_validation_analysis_sha256": "sha256:1430028ffeb3d2b317150f8b1f48d46f3c63251633c146d28c68fe9be0586250", | |
| "vision_validation_candidate_reselection": false, | |
| "vision_validation_decode_temperature": 1.0, | |
| "vision_validation_full_trace_missing": 0, | |
| "vision_validation_manifest_sha256": "sha256:5fd7d207b1b1c248987e0f79f98060a2c5b5eba55a07fa430f2d3ee92b601013", | |
| "vision_validation_records_per_model": 900 | |
| }, | |
| "schema_version": 1, | |
| "source_compact_artifacts": { | |
| "candidate_freeze": { | |
| "sha256": "sha256:69daadbcc364247579d344fe6ac21beda8d42f8ee2df9f2e5e77489f0c8896d5" | |
| }, | |
| "final_hmmt": { | |
| "sha256": "sha256:fb2b52620c634c45d490b085324589075b9322462a4911c4c20c0a84f7a7ca0c" | |
| }, | |
| "ifeval": { | |
| "sha256": "sha256:c73de9ae0dcf27cbe7557bb6159559432789e2646a1656edd81b6f349c3dd59a" | |
| }, | |
| "jmmlu_base": { | |
| "sha256": "sha256:69aa1a337310755fa493104039abf2762aba3340e13fa946704767a936b5d8ba" | |
| }, | |
| "jmmlu_candidate": { | |
| "sha256": "sha256:b0b92e708f84e5508bedd7471063272108aa00c25dca861b828d387e7a845d73" | |
| }, | |
| "lcb": { | |
| "sha256": "sha256:9f0b2b179e225f57f7c95fdbd816e6a82ca7293a5592430033a647886a953fc4" | |
| }, | |
| "lcb_admission": { | |
| "sha256": "sha256:c412866b96066141def7e94ebd94634cafda827a32275f9bf07c4ebcbf09370a" | |
| }, | |
| "lineage": { | |
| "sha256": "sha256:2469ae746de21ed31b41fee40ca5b44797fb2513fc5dd72c1a7c1d5c1e128418" | |
| }, | |
| "public_core": { | |
| "sha256": "sha256:1b815d52a61b9bf1fb6532c6bda39fc16ef740f3ce63607e4fefe11210572eca" | |
| } | |
| }, | |
| "source_preheldout_release_manifest_sha256": "sha256:2c375b5d019807f35ac718286ddf591d4cd7630d72835da2c9dd5513c5d0ca7d", | |
| "source_status": "materialized_immediately_before_heldout", | |
| "status": "technical_bundle_complete_owner_signoff_pending", | |
| "training_stages": [ | |
| { | |
| "method": "immutable pretrained checkpoint", | |
| "stage": "upstream_base" | |
| }, | |
| { | |
| "method": "difficulty-aware full-trace SFT into one composition-native rank-8 LoRA", | |
| "stage": "difficulty_aware_fulltrace_sft" | |
| }, | |
| { | |
| "method": "composition-native local step-DPO update of the existing SFT factors", | |
| "stage": "composition_native_step_dpo" | |
| }, | |
| { | |
| "method": "serve the unchanged Step-DPO rank-8 tensors at PEFT alpha 1 through 8 on public development prompts", | |
| "stage": "base_interpolated_anchor_screen", | |
| "tensor_sha256": "sha256:fb664321a3b9618e72ddf3b8c9e165f374098468ac19f6727f76abb53e8a648a" | |
| }, | |
| { | |
| "method": "the sealed Generation-1 df173 antithetic NES proposal selected by the earlier public-development search", | |
| "stage": "generation1_df173_good_checkpoint", | |
| "tensor_sha256": "sha256:699a961974399f391c2e40ddcd0a0bd3f8efe8de6e35fb63dbf05952ded355ca" | |
| }, | |
| { | |
| "method": "G2 HMMT81920 repair screens and exact controls established safe/unsafe scale regions; G3 screened eight candidates on GPQA-48 and confirmed four on complete GPQA", | |
| "stage": "full_budget_capability_repair_and_target_search" | |
| }, | |
| { | |
| "method": "G4 froze eight fractional deployment scalars around the df173 and Step alpha-7 centers, screened on paired GPQA-48 with family-specific screen-to-complete calibration, then applied unprojected hard gates on complete GPQA and MMLU-Pro", | |
| "stage": "good_checkpoint_centered_local_scale_search" | |
| }, | |
| { | |
| "method": "fresh six-direction antithetic block-separable NES at alpha 8.0, evaluated with the restored c48-per-GPU sampling contract", | |
| "stage": "g7_reproducible_good_checkpoint_local_nes" | |
| } | |
| ], | |
| "training_update": { | |
| "adapter_config_sha256": "sha256:2695ba5d5c892f5afe7cbb3bd6f043e621d5725a5cf901f2f64d3ee0621333c4", | |
| "adapter_model_sha256": "sha256:2a25037cd9c7bbdacdbefba3b09424e74f7759cb024acfe56f3a412db0f92e0e", | |
| "alpha": 8.0, | |
| "dropout": 0.05, | |
| "merge_topology_config_sha256": "sha256:f8f84eb242e007f1962284ca60781a049b49691c9d73d373e0b1716bf6e3ffc5", | |
| "method": "PEFT LoRA, merged into the publication weights", | |
| "rank": 8, | |
| "target_matrix_count": 104 | |
| }, | |
| "vision_evaluation": { | |
| "analysis_sha256": "sha256:1430028ffeb3d2b317150f8b1f48d46f3c63251633c146d28c68fe9be0586250", | |
| "base_official_budget_accuracy": 0.7377777777777778, | |
| "base_strict_budget_accuracy": 0.7355555555555555, | |
| "candidate_reselection": false, | |
| "claim_scope": "complete 900-item MMMU validation split; not the 10,500-item test split", | |
| "dataset_revision": "98e6ac0cb9b7b2cd2c991b85a50762edc4aedc68", | |
| "full_trace_integrity_pass": true, | |
| "manifest_sha256": "sha256:5fd7d207b1b1c248987e0f79f98060a2c5b5eba55a07fa430f2d3ee92b601013", | |
| "quantized_checkpoint_analysis_sha256": "sha256:9a47cdf0ce420beb0a715ed10e3ca36ae051160504876cd2d1d5d1bcd9ed6298", | |
| "quantized_checkpoint_candidate_reselection": false, | |
| "qwick_fp8_official_budget_accuracy": 0.7277777777777777, | |
| "qwick_minus_base_pp": 0.7777777777777772, | |
| "qwick_nvfp4_official_budget_accuracy": 0.7055555555555556, | |
| "qwick_official_budget_accuracy": 0.7455555555555555, | |
| "qwick_strict_budget_accuracy": 0.7411111111111112, | |
| "records": 900, | |
| "role": "report_only; disclosed regardless of direction; not a release gate" | |
| } | |
| } | |