docs: update v0.9.34

docs/_config.yml

@@ -122,6 +122,8 @@ docs_nav:
         url: "/docs/models/gemma/"
       - label: MedGemma
         url: "/docs/models/medgemma/"
+      - label: TranslateGemma
+        url: "/docs/models/translategemma/"
       - label: gpt-oss
         url: "/docs/models/gpt-oss/"
       - label: LiquidAI/LFM

docs/docs/models/gemma.md

@@ -47,29 +47,4 @@ flm run gemma3:4b
 /input "file/to/image.jpg" describe this image.
 ```
 
----
-
-## 🧩 Model Card: [gemma-3-270m-it](https://huggingface.co/google/gemma-3-270m-it)
-
-- **Type:** Image-Text-to-Text
-- **Think:** No
-- **Tool Calling Support:** No  
-- **Base Model:** [google/gemma-3-270m-it](https://huggingface.co/google/gemma-3-270m-it)
-- **Quantization:** Q4_1
-- **Max Context Length:** 32k tokens  
-- **Default Context Length:** 32k tokens ([change default](https://fastflowlm.com/docs/instructions/cli/#-change-default-context-length-max))  
-- **[Set Context Length at Launch](https://fastflowlm.com/docs/instructions/cli/#-set-context-length-at-launch)**
-
-▶️ Run with FastFlowLM in PowerShell:  
-
-```shell
-flm run gemma3:270m
-```
-
-⚠️ **Warning:** 
-> – `gemma3:270m` is **Experimental** in FLM  
-> – Limited accuracy; may produce errors  
-> – Can loop on long outputs (quirk from Unsloth weights, also seen in LM Studio)  
-> – Experimenting with different quantization + hyperparameters  
-
 ---

docs/docs/models/index.md

@@ -20,6 +20,7 @@ has_children: true
 - [Qwen](qwen/)
 - [Gemma](gemma/)
 - [MedGemma](medgemma/)
+- [TranslateGemma](translategemma/)
 - [gpt-oss](gpt-oss/)
 - [LiquidAI/LFM](lfm/)
 - [Phi](phi/)

docs/docs/models/medgemma.md

@@ -22,12 +22,6 @@ parent: Models
 flm run medgemma:4b
 ```
 
-🔐 Why It Matters
-
-- **Privacy First** — There is nothing more personal than your health!  
-- **Powered by NPU** — Leverages AMD Ryzen™ AI NPU for fast, low-power inference.  
-- **Healthcare Applications** — A concrete example of how local LLMs + NPUs enable privacy-preserving, research-driven healthcare workflows.
-
 📺 Demo Video
 
 - [MedGemma:4B (Multimodal) on AMD Ryzen™ AI NPU — Demo Video](https://www.youtube.com/watch?v=KWzXZEOcgK4&list=PLf87s9UUZrJoDdz639Yc6w1UTyJ4cFHZ1&index=5&ab_channel=FastFlowLM)  
@@ -39,4 +33,40 @@ flm run medgemma:4b
 This tool (**MedGemma + FastFlowLM**) is **not** a diagnostic or clinical tool. 
 Always consult a licensed medical professional for healthcare decisions.
 
+---
+
+## 🧩 Model Card: [medgemma1.5:4b](https://huggingface.co/google/medgemma-1.5-4b-it)
+
+- **Type:** Image-Text-to-Text
+- **Think:** No
+- **Tool Calling Support:** No  
+- **Base Model:** [google/medgemma1.5-4b-it](https://huggingface.co/google/medgemma-1.5-4b-it)
+- **Quantization:** Q4_1
+- **Max Context Length:** 128k tokens 
+- **Default Context Length:** 64k tokens ([change default](https://fastflowlm.com/docs/instructions/cli/#-change-default-context-length-max))  
+- **[Set Context Length at Launch](https://fastflowlm.com/docs/instructions/cli/#-set-context-length-at-launch)**
+
+▶️ Run with FastFlowLM in PowerShell:  
+
+```shell
+flm run medgemma1.5:4b
+```
+
+⚠️ **Disclaimer** 
+
+This tool (**MedGemma + FastFlowLM**) is **not** a diagnostic or clinical tool. 
+Always consult a licensed medical professional for healthcare decisions.
+
+📺 Demo Video
+
+- [MedGemma1.5:4B (Multimodal) on AMD Ryzen™ AI NPU — Demo Video]()
+- [Reaseach Blog](https://research.google/blog/next-generation-medical-image-interpretation-with-medgemma-15-and-medical-speech-to-text-with-medasr/)
+
+---
+
+🔐 Why They Matter
+
+- **Privacy First** — There is nothing more personal than your health!  
+- **Powered by NPU** — Leverages AMD Ryzen™ AI NPU for fast, low-power inference.  
+- **Healthcare Applications** — A concrete example of how local LLMs + NPUs enable privacy-preserving, research-driven healthcare workflows.
 ---

docs/docs/models/qwen.md

@@ -156,6 +156,27 @@ flm run qwen3-it:4b
 flm run qwen3vl-it:4b
 ```
 
+▶️ Image Resize Options
+
+You can control image resizing when running or serving the model using the `--resize` flag:
+
+```shell
+flm run qwen3vl-it:3b --resize 1
+```
+
+```shell
+flm serve qwen3vl-it:3b --resize 1
+```
+
+The `--resize` option determines the target length of the image’s **longest** side:
+
+- -1 — Disable resizing (use the original image)
+- 0 — Resize so the longest side is **1080 px** (Default)
+- 1 — Resize so the longest side is **1920 px**
+- 2 — Resize so the longest side is **2560 px**
+
+> Don't worry—if your image is already smaller than 1080px, it keeps its original resolution! ✨
+
 📝 **Note**
 
 - Image understanding adapts to image size. Image TTFT can range from under 1 second to ~200 seconds depending on resolution. Use lower-resolution images (720p or below) unless high resolution is required (e.g. OCR on small text).
@@ -199,6 +220,27 @@ flm run qwen2.5-it:3b
 flm run qwen2.5vl-it:3b
 ```
 
+▶️ Image Resize Options
+
+You can control image resizing when running or serving the model using the `--resize` flag:
+
+```shell
+flm run qwen2.5vl-it:3b --resize 1
+```
+
+```shell
+flm serve qwen2.5vl-it:3b --resize 1
+```
+
+The `--resize` option determines the target length of the image’s **longest** side:
+
+- -1 — Disable resizing (use the original image)
+- 0 — Resize so the longest side is **1080 px** (Default)
+- 1 — Resize so the longest side is **1920 px**
+- 2 — Resize so the longest side is **2560 px**
+
+> Don't worry—if your image is already smaller than 1080px, it keeps its original resolution! ✨
+
 📝 **Note**
 
 - Image understanding adapts to image size. Image TTFT can range from under 1 second to ~200 seconds depending on resolution. Use lower-resolution images (720p or below) unless high resolution is required (e.g. OCR on small text).

docs/docs/models/translategemma.md

@@ -0,0 +1,45 @@
+---
+layout: docs
+title: TranslateGemma
+nav_order: 11
+parent: Models
+---
+
+## 🧩 Model Card: [translategemma:4b](https://huggingface.co/google/translategemma-4b-it)
+
+- **Type:** Image-Text-to-Text
+- **Think:** No
+- **Tool Calling Support:** No  
+- **Base Model:** [google/translategemma-4b-it](https://huggingface.co/google/translategemma-4b-it)
+- **Quantization:** Q4_1
+- **Max Context Length:** 128k tokens 
+- **Default Context Length:** 64k tokens ([change default](https://fastflowlm.com/docs/instructions/cli/#-change-default-context-length-max))  
+- **[Set Context Length at Launch](https://fastflowlm.com/docs/instructions/cli/#-set-context-length-at-launch)**
+
+▶️ Run with FastFlowLM in PowerShell:  
+
+```shell
+flm run translategemma:4b
+```
+
+📖 Prompt Guide
+
+**Prompt Format**
+
+TranslateGemma expects a single **user message** with this structure:
+
+    You are a professional {SOURCE_LANG} ({SOURCE_CODE}) to {TARGET_LANG} ({TARGET_CODE}) translator. Your goal is to accurately convey the meaning and nuances of the original {SOURCE_LANG} text while adhering to {TARGET_LANG} grammar, vocabulary, and cultural sensitivities.
+    Produce only the {TARGET_LANG} translation, without any additional explanations or commentary. Please translate the following {SOURCE_LANG} text into {TARGET_LANG}:
+
+
+    {TEXT}
+
+**Important**: There are two blank lines before the text to translate.
+
+🦙Refer to the Ollama model library for detailed usage [examples](https://ollama.com/library/translategemma#:~:text=text%20to%20translate.-,Examples,-English%20to%20Spanish).
+
+📺 Demo Video
+
+- [Translategemma1.5:4B (Multimodal) on AMD Ryzen™ AI NPU — Demo Video]()
+- [Reaseach Blog](https://blog.google/innovation-and-ai/technology/developers-tools/translategemma/)
+- [Technical Report](https://arxiv.org/pdf/2601.09012)