--- base_model: xtools-at/Qwen3.5-4B-DeepThought-STEM datasets: - xtools-at/Claude-Haiku-5.5-High-Reasoning-TraceInversion-3000x - xtools-at/Xiaomi-MiMo-2.6-Pro-High-Reasoning-250x - xtools-at/ChatGPT-6.1-Terra-High-Reasoning-TraceInversion-80x - Roman1111111/GPT-5.6-luna-reasoning-102881x - Jackrong/GLM-5.1-Reasoning-1M-Cleaned - Kassadin88/GLM-5.1-1000000x - xtools-at/red-pill-drug-discovery-formulation-synthetic-reasoning - saidutta69/red-pill-drug-discovery-formulation - HuggingFaceTB/smoltalk2 language: - en library_name: transformers license: apache-2.0 mradermacher: readme_rev: 1 quantized_by: mradermacher tags: - qwen3.5 - lora - peft - reasoning - mathematics - science - stem - drug-discovery - pharmaceutical-formulation - pharmaceutics - academics - scientific-reasoning - text-generation-inference - transformers - unsloth - qwen3_5 - distillation - deepseek - sft - math - heretic - uncensored - decensored - abliterated --- ## About static quants of https://huggingface.co/xtools-at/Qwen3.5-4B-DeepThought-STEM ***For a convenient overview and download list, visit our [model page for this model](https://hf.tst.eu/model#Qwen3.5-4B-DeepThought-STEM-GGUF).*** weighted/imatrix quants are available at https://huggingface.co/mradermacher/Qwen3.5-4B-DeepThought-STEM-i1-GGUF ## Usage If you are unsure how to use GGUF files, refer to one of [TheBloke's READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for more details, including on how to concatenate multi-part files. ## Provided Quants (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants) | Link | Type | Size/GB | Notes | |:-----|:-----|--------:|:------| | [GGUF](https://huggingface.co/mradermacher/Qwen3.5-4B-DeepThought-STEM-GGUF/resolve/main/Qwen3.5-4B-DeepThought-STEM.mmproj-Q8_0.gguf) | mmproj-Q8_0 | 0.5 | multi-modal supplement | | [GGUF](https://huggingface.co/mradermacher/Qwen3.5-4B-DeepThought-STEM-GGUF/resolve/main/Qwen3.5-4B-DeepThought-STEM.mmproj-f16.gguf) | mmproj-f16 | 0.8 | multi-modal supplement | | [GGUF](https://huggingface.co/mradermacher/Qwen3.5-4B-DeepThought-STEM-GGUF/resolve/main/Qwen3.5-4B-DeepThought-STEM.Q2_K.gguf) | Q2_K | 2.1 | | | [GGUF](https://huggingface.co/mradermacher/Qwen3.5-4B-DeepThought-STEM-GGUF/resolve/main/Qwen3.5-4B-DeepThought-STEM.Q3_K_S.gguf) | Q3_K_S | 2.2 | | | [GGUF](https://huggingface.co/mradermacher/Qwen3.5-4B-DeepThought-STEM-GGUF/resolve/main/Qwen3.5-4B-DeepThought-STEM.Q3_K_M.gguf) | Q3_K_M | 2.4 | lower quality | | [GGUF](https://huggingface.co/mradermacher/Qwen3.5-4B-DeepThought-STEM-GGUF/resolve/main/Qwen3.5-4B-DeepThought-STEM.Q3_K_L.gguf) | Q3_K_L | 2.6 | | | [GGUF](https://huggingface.co/mradermacher/Qwen3.5-4B-DeepThought-STEM-GGUF/resolve/main/Qwen3.5-4B-DeepThought-STEM.IQ4_XS.gguf) | IQ4_XS | 2.7 | | | [GGUF](https://huggingface.co/mradermacher/Qwen3.5-4B-DeepThought-STEM-GGUF/resolve/main/Qwen3.5-4B-DeepThought-STEM.Q4_K_S.gguf) | Q4_K_S | 2.7 | fast, recommended | | [GGUF](https://huggingface.co/mradermacher/Qwen3.5-4B-DeepThought-STEM-GGUF/resolve/main/Qwen3.5-4B-DeepThought-STEM.Q4_K_M.gguf) | Q4_K_M | 2.9 | fast, recommended | | [GGUF](https://huggingface.co/mradermacher/Qwen3.5-4B-DeepThought-STEM-GGUF/resolve/main/Qwen3.5-4B-DeepThought-STEM.Q5_K_S.gguf) | Q5_K_S | 3.2 | | | [GGUF](https://huggingface.co/mradermacher/Qwen3.5-4B-DeepThought-STEM-GGUF/resolve/main/Qwen3.5-4B-DeepThought-STEM.Q5_K_M.gguf) | Q5_K_M | 3.3 | | | [GGUF](https://huggingface.co/mradermacher/Qwen3.5-4B-DeepThought-STEM-GGUF/resolve/main/Qwen3.5-4B-DeepThought-STEM.Q6_K.gguf) | Q6_K | 3.7 | very good quality | | [GGUF](https://huggingface.co/mradermacher/Qwen3.5-4B-DeepThought-STEM-GGUF/resolve/main/Qwen3.5-4B-DeepThought-STEM.Q8_0.gguf) | Q8_0 | 4.7 | fast, best quality | | [GGUF](https://huggingface.co/mradermacher/Qwen3.5-4B-DeepThought-STEM-GGUF/resolve/main/Qwen3.5-4B-DeepThought-STEM.f16.gguf) | f16 | 8.8 | 16 bpw, overkill | Here is a handy graph by ikawrakow comparing some lower-quality quant types (lower is better): ![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png) And here are Artefact2's thoughts on the matter: https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9 ## FAQ / Model Request See https://huggingface.co/mradermacher/model_requests for some answers to questions you might have and/or if you want some other model quantized. ## Thanks I thank my company, [nethype GmbH](https://www.nethype.de/), for letting me use its servers and providing upgrades to my workstation to enable this work in my free time.