nexaai 0.0.8.8__tar.gz → 0.0.9.0__tar.gz

{nexaai-0.0.8.8 → nexaai-0.0.9.0}/.gitignore

@@ -86,4 +86,7 @@ build_*/
 *.sqlite
 
 # Other
-.cache/
+.cache/
+
+# tests
+quantization_test.py

{nexaai-0.0.8.8 → nexaai-0.0.9.0}/.gitmodules

@@ -5,7 +5,7 @@
 [submodule "dependency/llama.cpp"]
 	path = dependency/llama.cpp
 	url = https://github.com/NexaAI/llama.cpp.git
-	branch = experiment-branch
+	branch = master-release
 [submodule "nexa/eval/benchmark_tasks"]
 	path = nexa/eval/benchmark_tasks
 	url = https://github.com/NexaAI/benchmark-tasks.git

{nexaai-0.0.8.8 → nexaai-0.0.9.0}/CLI.md

@@ -13,6 +13,7 @@ positional arguments:
     run                 Run inference for various tasks using GGUF models.
     onnx                Run inference for various tasks using ONNX models.
     embed               Generate embeddings for text.
+    convert             Convert and quantize a Hugging Face model to GGUF format.
     server              Run the Nexa AI Text Generation Service.
     eval                Run the Nexa AI Evaluation Tasks.
     pull                Pull a model from official or hub.
@@ -268,6 +269,63 @@ nexa embed nomic-embed-text-v1.5:fp16 "I love Nexa AI."
 nexa embed sentence-transformers/all-MiniLM-L6-v2:gguf-fp16 "I love Nexa AI." >> generated_embeddings.txt
 ```
 
+### Convert and quantize a Hugging Face Model to GGUF
+
+```
+nexa convert HF_MODEL_PATH [ftype] [output_file]
+usage: nexa convert [-h] [-t NTHREAD] [--convert_type CONVERT_TYPE] [--bigendian] [--use_temp_file] [--no_lazy]
+                    [--metadata METADATA] [--split_max_tensors SPLIT_MAX_TENSORS] [--split_max_size SPLIT_MAX_SIZE]
+                    [--no_tensor_first_split] [--vocab_only] [--dry_run] [--output_tensor_type OUTPUT_TENSOR_TYPE]
+                    [--token_embedding_type TOKEN_EMBEDDING_TYPE] [--allow_requantize] [--quantize_output_tensor]
+                    [--only_copy] [--pure] [--keep_split] input_path [ftype] [output_file]
+
+positional arguments:
+  input_path            Path to the input Hugging Face model directory or GGUF file
+  ftype                 Quantization type (default: q4_0)
+  output_file           Path to the output quantized GGUF file
+
+options:
+  -h, --help            show this help message and exit
+  -t, --nthread NTHREAD Number of threads to use (default: 4)
+  --convert_type CONVERT_TYPE
+                        Conversion type for safetensors to GGUF (default: f16)
+  --bigendian           Use big endian format
+  --use_temp_file       Use a temporary file during conversion
+  --no_lazy             Disable lazy loading
+  --metadata METADATA   Additional metadata as JSON string
+  --split_max_tensors SPLIT_MAX_TENSORS
+                        Maximum number of tensors per split
+  --split_max_size SPLIT_MAX_SIZE
+  --no_tensor_first_split
+                        Disable tensor-first splitting
+  --vocab_only          Only process vocabulary
+  --dry_run             Perform a dry run without actual conversion
+  --output_tensor_type  Output tensor type
+  --token_embedding_type
+                        Token embedding type
+  --allow_requantize    Allow quantizing non-f32/f16 tensors
+  --quantize_output_tensor
+                        Quantize output.weight
+  --only_copy           Only copy tensors (ignores ftype, allow_requantize, and quantize_output_tensor)
+  --pure                Quantize all tensors to the default type
+  --keep_split          Quantize to the same number of shards
+```
+
+#### Example
+
+```
+# Default quantization type (q4_0) and output file in current directory
+nexa convert meta-llama/Llama-3.2-1B-Instruct
+
+# Equivalent to:
+# nexa convert meta-llama/Llama-3.2-1B-Instruct q4_0 ./Llama-3.2-1B-Instruct-q4_0.gguf
+
+# Specifying quantization type and output file
+nexa convert meta-llama/Llama-3.2-1B-Instruct q6_k llama3.2-1b-instruct-q6_k.gguf
+```
+
+Note: When not specified, the default quantization type is set to q4_0, and the output file will be created in the current directory with the name format: `<model_name>-q4_0.gguf`.
+
 ### Start Local Server
 
 Start a local server using models on your local computer.
@@ -329,3 +387,7 @@ For `model_path` in nexa commands, it's better to follow the standard format to
 - `gemma-2b:q4_0`
 - `Meta-Llama-3-8B-Instruct:onnx-cpu-int8`
 - `liuhaotian/llava-v1.6-vicuna-7b:gguf-q4_0`
+
+```
+</rewritten_chunk>
+```

{nexaai-0.0.8.8 → nexaai-0.0.9.0}/CMakeLists.txt

@@ -1,6 +1,6 @@
 cmake_minimum_required(VERSION 3.16)
 
-if (GGML_CUDA OR GGML_METAL)
+if (GGML_CUDA OR GGML_METAL OR GGML_HIPBLAS OR GGML_VULKAN)
     set(SOURCE_EMPTY_FILE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/nexa/gguf/lib/empty_file.txt")
     add_custom_command(
         OUTPUT ${SOURCE_EMPTY_FILE_PATH}

{nexaai-0.0.8.8 → nexaai-0.0.9.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: nexaai
-Version: 0.0.8.8
+Version: 0.0.9.0
 Summary: Nexa AI SDK
 Author-Email: Nexa AI <octopus@nexa4ai.com>
 License: MIT
@@ -11,6 +11,7 @@ Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
 Project-URL: Homepage, https://github.com/NexaAI/nexa-sdk
 Project-URL: Issues, https://github.com/NexaAI/nexa-sdk/issues
 Project-URL: Documentation, https://docs.nexa4ai.com/
@@ -60,16 +61,17 @@ Requires-Dist: flatten_dict; extra == "eval"
 Requires-Dist: colorlog; extra == "eval"
 Requires-Dist: pandas; extra == "eval"
 Requires-Dist: rich; extra == "eval"
+Requires-Dist: codecarbon; extra == "eval"
+Requires-Dist: nexa-gguf; extra == "convert"
 Provides-Extra: onnx
 Provides-Extra: eval
+Provides-Extra: convert
 Description-Content-Type: text/markdown
 
 <div align="center">
 
 <h1>Nexa SDK</h1>
 
-  <img src="https://public-storage.nexa4ai.com/nexa-banner.png" alt="icon" onerror="this.onerror=null; this.src='./assets/banner.png'"/>
-
 [![MacOS][MacOS-image]][release-url] [![Linux][Linux-image]][release-url] [![Windows][Windows-image]][release-url]
 
 [![GitHub Release](https://img.shields.io/github/v/release/NexaAI/nexa-sdk)](https://github.com/NexaAI/nexa-sdk/releases/latest) [![Build workflow](https://img.shields.io/github/actions/workflow/status/NexaAI/nexa-sdk/ci.yaml?label=CI&logo=github)](https://github.com/NexaAI/nexa-sdk/actions/workflows/ci.yaml?query=branch%3Amain) ![GitHub License](https://img.shields.io/github/license/NexaAI/nexa-sdk)
@@ -89,39 +91,46 @@ Description-Content-Type: text/markdown
 
 Nexa SDK is a comprehensive toolkit for supporting **ONNX** and **GGML** models. It supports text generation, image generation, vision-language models (VLM), and speech-to-text (ASR), and text-to-speech (TTS) capabilities. Additionally, it offers an OpenAI-compatible API server with JSON schema mode for function calling and streaming support, and a user-friendly Streamlit UI. Users can run Nexa SDK in any device with Python environment, and GPU acceleration is supported, including CUDA, Metal, and ROCm. An executable version is also available.
 
+<video src="https://user-images.githubusercontent.com/assets/375570dc-0e7a-4a99-840d-c1ef6502e5aa.mp4" autoplay muted loop playsinline style="max-width: 100%;"></video>
+
 ## Latest News 🔥
-* [2024/10] Support embedding model: `nexa embed <model_path> <prompt>`
-* [2024/10] Support pull and run supported Computer Vision models in GGUF format from HuggingFace: `nexa run -hf <model_id> -mt COMPUTER_VISION`
-* [2024/10] Support VLM in local server.
-* [2024/10] Added option to customize maximum context window for NLP and VLM models.
-* [2024/10] Support running model from user's local path
-* [2024/10] Added LoRA support for NLP models. 
-* [2024/10] Added support for whisper-large-v3-turbo: `nexa run faster-whisper-large-turbo`
-* [2024/10] Added support for AMD-Llama-135m: `nexa run AMD-Llama-135m:fp16`
-* [2024/09] Nexa now has executables for easy installation: [Install Nexa SDK](https://nexaai.com/download-sdk) ✨
-* [2024/09] Added support for Llama 3.2 models: `nexa run llama3.2`
-* [2024/09] Added support for Qwen2.5, Qwen2.5-coder and Qwen2.5-Math models: `nexa run qwen2.5`
-* [2024/09] Support pull and run NLP models in GGUF format from HuggingFace: `nexa run -hf <model_id> -mt NLP`
-* [2024/09] Added support for ROCm
-* [2024/09] Added support for Phi-3.5 models: `nexa run phi3.5`
-* [2024/09] Added support for OpenELM models: `nexa run openelm`
-* [2024/09] Introduced logits API support for more advanced model interactions
-* [2024/09] Added support for Flux models: `nexa run flux`
-* [2024/09] Added support for Stable Diffusion 3 model: `nexa run sd3`
-* [2024/09] Added support for Stable Diffusion 2.1 model: `nexa run sd2-1`
+
+- [2024/10] Support embedding model: `nexa embed <model_path> <prompt>`
+- [2024/10] Support pull and run supported Computer Vision models in GGUF format from HuggingFace: `nexa run -hf <model_id> -mt COMPUTER_VISION`
+- [2024/10] Support VLM in local server.
+- [2024/10] Added option to customize maximum context window for NLP and VLM models.
+- [2024/10] Support running model from user's local path
+- [2024/10] Added LoRA support for NLP models.
+- [2024/10] Added support for whisper-large-v3-turbo: `nexa run faster-whisper-large-turbo`
+- [2024/10] Added support for AMD-Llama-135m: `nexa run AMD-Llama-135m:fp16`
+- [2024/09] Nexa now has executables for easy installation: [Install Nexa SDK](https://nexaai.com/download-sdk) ✨
+- [2024/09] Added support for Llama 3.2 models: `nexa run llama3.2`
+- [2024/09] Added support for Qwen2.5, Qwen2.5-coder and Qwen2.5-Math models: `nexa run qwen2.5`
+- [2024/09] Support pull and run NLP models in GGUF format from HuggingFace: `nexa run -hf <model_id> -mt NLP`
+- [2024/09] Added support for ROCm
+- [2024/09] Added support for Phi-3.5 models: `nexa run phi3.5`
+- [2024/09] Added support for OpenELM models: `nexa run openelm`
+- [2024/09] Introduced logits API support for more advanced model interactions
+- [2024/09] Added support for Flux models: `nexa run flux`
+- [2024/09] Added support for Stable Diffusion 3 model: `nexa run sd3`
+- [2024/09] Added support for Stable Diffusion 2.1 model: `nexa run sd2-1`
 
 Welcome to submit your requests through [issues](https://github.com/NexaAI/nexa-sdk/issues/new/choose), we ship weekly.
 
 ## Installation - Executable
 
 ### macOS
+
 [Download](https://public-storage.nexa4ai.com/nexa-sdk-executable-installer/nexa-macos-installer.pkg)
 
 ### Linux
+
 ```bash
-curl -fsSL https://public-storage.nexa4ai.com/install.sh | sh 
+curl -fsSL https://public-storage.nexa4ai.com/install.sh | sh
 ```
+
 ### Windows
+
 Coming soon. Install with Python package below 👇
 
 ## Installation - Python Package
@@ -131,7 +140,8 @@ We have released pre-built wheels for various Python versions, platforms, and ba
 > [!NOTE]
 >
 > 1. If you want to use <strong>ONNX model</strong>, just replace `pip install nexaai` with `pip install "nexaai[onnx]"` in provided commands.
-> 2. For Chinese developers, we recommend you to use <strong>Tsinghua Open Source Mirror</strong> as extra index url, just replace `--extra-index-url https://pypi.org/simple` with `--extra-index-url https://pypi.tuna.tsinghua.edu.cn/simple` in provided commands.
+> 2. If you want to convert and quantize huggingface models to GGUF models, just replace `pip install nexaai` with `pip install "nexaai[nexa-gguf]"`.
+> 3. For Chinese developers, we recommend you to use <strong>Tsinghua Open Source Mirror</strong> as extra index url, just replace `--extra-index-url https://pypi.org/simple` with `--extra-index-url https://pypi.tuna.tsinghua.edu.cn/simple` in provided commands.
 
 #### CPU
 
@@ -164,6 +174,8 @@ CMAKE_ARGS="-DGGML_METAL=ON -DSD_METAL=ON" pip install nexaai --prefer-binary --
 
 #### GPU (CUDA)
 
+To install with CUDA support, make sure you have [CUDA Toolkit 12.0 or later](https://developer.nvidia.com/cuda-12-0-0-download-archive) installed.
+
 For **Linux**:
 
 ```bash
@@ -205,12 +217,36 @@ CMAKE_ARGS="-DCMAKE_CXX_FLAGS=-fopenmp" pip install nexaai
 
 #### GPU (ROCm)
 
+To install with ROCm support, make sure you have [ROCm 6.2.1 or later](https://rocm.docs.amd.com/projects/install-on-linux/en/docs-6.2.1/install/quick-start.html) installed.
+
 For **Linux**:
 
 ```bash
 CMAKE_ARGS="-DGGML_HIPBLAS=on" pip install nexaai --prefer-binary --index-url https://nexaai.github.io/nexa-sdk/whl/rocm621 --extra-index-url https://pypi.org/simple --no-cache-dir
 ```
 
+#### GPU (Vulkan)
+
+To install with Vulkan support, make sure you have [Vulkan SDK 1.3.261.1 or later](https://vulkan.lunarg.com/sdk/home) installed.
+
+For **Windows PowerShell**:
+
+```bash
+$env:CMAKE_ARGS="-DGGML_VULKAN=on"; pip install nexaai --prefer-binary --index-url https://nexaai.github.io/nexa-sdk/whl/vulkan --extra-index-url https://pypi.org/simple --no-cache-dir
+```
+
+For **Windows Command Prompt**:
+
+```bash
+set CMAKE_ARGS="-DGGML_VULKAN=on" & pip install nexaai --prefer-binary --index-url https://nexaai.github.io/nexa-sdk/whl/vulkan --extra-index-url https://pypi.org/simple --no-cache-dir
+```
+
+For **Windows Git Bash**:
+
+```bash
+CMAKE_ARGS="-DGGML_VULKAN=on" pip install nexaai --prefer-binary --index-url https://nexaai.github.io/nexa-sdk/whl/vulkan --extra-index-url https://pypi.org/simple --no-cache-dir
+```
+
 ### Local Build
 
 How to clone this repo
@@ -265,68 +301,69 @@ Below is our differentiation from other similar tools:
 | **User Interface**         |                         ✅                         |                       ❌                       |                          ❌                           |                       ✅                        |
 
 ## Supported Models & Model Hub
+
 Our on-device model hub offers all types of quantized models (text, image, audio, multimodal) with filters for RAM, file size, Tasks, etc. to help you easily explore models with UI. Explore on-device models at [On-device Model Hub](https://model-hub.nexa4ai.com/)
 
 Supported models (full list at [Model Hub](https://nexa.ai/models)):
-| Model                                                                                                   | Type            | Format    | Command                                |
+| Model | Type | Format | Command |
 | ------------------------------------------------------------------------------------------------------- | --------------- | --------- | -------------------------------------- |
-| [octopus-v2](https://www.nexaai.com/NexaAI/Octopus-v2/gguf-q4_0/readme)                                 | NLP             | GGUF      | `nexa run octopus-v2`                  |
-| [octopus-v4](https://www.nexaai.com/NexaAI/Octopus-v4/gguf-q4_0/readme)                                 | NLP             | GGUF      | `nexa run octopus-v4`                  |
-| [gpt2](https://nexaai.com/openai/gpt2/gguf-q4_0/readme)                                                 | NLP             | GGUF      | `nexa run gpt2`                        |
-| [tinyllama](https://www.nexaai.com/TinyLlama/TinyLlama-1.1B-Chat-v1.0/gguf-fp16/readme)                 | NLP             | GGUF      | `nexa run tinyllama`                   |
-| [llama2](https://www.nexaai.com/meta/Llama2-7b-chat/gguf-q4_0/readme)                                   | NLP             | GGUF/ONNX | `nexa run llama2`                      |
-| [llama2-uncensored](https://www.nexaai.com/georgesung/Llama2-7b-chat-uncensored/gguf-q4_0/readme)       | NLP             | GGUF      | `nexa run llama2-uncensored`           |
-| [llama2-function-calling](https://www.nexaai.com/Trelis/Llama2-7b-function-calling/gguf-q4_K_M/readme)  | NLP             | GGUF      | `nexa run llama2-function-calling`     |
-| [llama3](https://www.nexaai.com/meta/Llama3-8B-Instruct/gguf-q4_0/readme)                               | NLP             | GGUF/ONNX | `nexa run llama3`                      |
-| [llama3.1](https://www.nexaai.com/meta/Llama3.1-8B-Instruct/gguf-q4_0/readme)                           | NLP             | GGUF/ONNX | `nexa run llama3.1`                    |
-| [llama3.2](https://nexaai.com/meta/Llama3.2-3B-Instruct/gguf-q4_0/readme)                               | NLP             | GGUF      | `nexa run llama3.2`                    |
-| [llama3-uncensored](https://www.nexaai.com/Orenguteng/Llama3-8B-Lexi-Uncensored/gguf-q4_K_M/readme)     | NLP             | GGUF      | `nexa run llama3-uncensored`           |
-| [gemma](https://www.nexaai.com/google/gemma-1.1-2b-instruct/gguf-q4_0/readme)                           | NLP             | GGUF/ONNX | `nexa run gemma`                       |
-| [gemma2](https://www.nexaai.com/google/gemma-2-2b-instruct/gguf-q4_0/readme)                            | NLP             | GGUF      | `nexa run gemma2`                      |
-| [qwen1.5](https://www.nexaai.com/Qwen/Qwen1.5-7B-Instruct/gguf-q4_0/readme)                             | NLP             | GGUF      | `nexa run qwen1.5`                     |
-| [qwen2](https://www.nexaai.com/Qwen/Qwen2-1.5B-Instruct/gguf-q4_0/readme)                               | NLP             | GGUF/ONNX | `nexa run qwen2`                       |
-| [qwen2.5](https://www.nexaai.com/Qwen/Qwen2.5-1.5B-Instruct/gguf-q4_0/readme)                           | NLP             | GGUF      | `nexa run qwen2.5`                     |
-| [mathqwen](https://nexaai.com/Qwen/Qwen2.5-Math-1.5B-Instruct/gguf-q4_0/readme)                         | NLP             | GGUF      | `nexa run mathqwen`                    |
-| [codeqwen](https://www.nexaai.com/Qwen/CodeQwen1.5-7B-Instruct/gguf-q4_0/readme)                        | NLP             | GGUF      | `nexa run codeqwen`                    |
-| [mistral](https://www.nexaai.com/mistralai/Mistral-7B-Instruct-v0.3/gguf-q4_0/readme)                   | NLP             | GGUF/ONNX | `nexa run mistral`                     |
-| [dolphin-mistral](https://www.nexaai.com/CognitiveComputations/dolphin-2.8-mistral-7b/gguf-q4_0/readme) | NLP             | GGUF      | `nexa run dolphin-mistral`             |
-| [codegemma](https://www.nexaai.com/google/codegemma-2b/gguf-q4_0/readme)                                | NLP             | GGUF      | `nexa run codegemma`                   |
-| [codellama](https://www.nexaai.com/meta/CodeLlama-7b-Instruct/gguf-q2_K/readme)                         | NLP             | GGUF      | `nexa run codellama`                   |
-| [deepseek-coder](https://www.nexaai.com/DeepSeek/deepseek-coder-1.3b-instruct/gguf-q4_0/readme)         | NLP             | GGUF      | `nexa run deepseek-coder`              |
-| [phi2](https://www.nexaai.com/microsoft/Phi-2/gguf-q4_0/readme)                                         | NLP             | GGUF      | `nexa run phi2`                        |
-| [phi3](https://www.nexaai.com/microsoft/Phi-3-mini-128k-instruct/gguf-q4_0/readme)                      | NLP             | GGUF/ONNX | `nexa run phi3`                        |
-| [phi3.5](https://nexaai.com/microsoft/Phi-3.5-mini-instruct/gguf-q4_0/readme)                           | NLP             | GGUF      | `nexa run phi3.5`                      |
-| [openelm](https://nexaai.com/apple/OpenELM-3B/gguf-q4_K_M/readme)                                       | NLP             | GGUF      | `nexa run openelm`                     |
-| [AMD-Llama-135m](https://nexaai.com/amd/AMD-Llama-135m/gguf-fp16/readme)                                | NLP             | GGUF      | `nexa run AMD-Llama-135m:fp16`         |
-| [nanollava](https://www.nexaai.com/qnguyen3/nanoLLaVA/gguf-fp16/readme)                                 | Multimodal      | GGUF      | `nexa run nanollava`                   |
-| [llava-phi3](https://www.nexaai.com/xtuner/llava-phi-3-mini/gguf-q4_0/readme)                           | Multimodal      | GGUF      | `nexa run llava-phi3`                  |
-| [llava-llama3](https://www.nexaai.com/xtuner/llava-llama-3-8b-v1.1/gguf-q4_0/readme)                    | Multimodal      | GGUF      | `nexa run llava-llama3`                |
-| [llava1.6-mistral](https://www.nexaai.com/liuhaotian/llava-v1.6-mistral-7b/gguf-q4_0/readme)            | Multimodal      | GGUF      | `nexa run llava1.6-mistral`            |
-| [llava1.6-vicuna](https://www.nexaai.com/liuhaotian/llava-v1.6-vicuna-7b/gguf-q4_0/readme)              | Multimodal      | GGUF      | `nexa run llava1.6-vicuna`             |
-| [stable-diffusion-v1-4](https://www.nexaai.com/runwayml/stable-diffusion-v1-4/gguf-q4_0/readme)         | Computer Vision | GGUF      | `nexa run sd1-4`                       |
-| [stable-diffusion-v1-5](https://www.nexaai.com/runwayml/stable-diffusion-v1-5/gguf-q4_0/readme)         | Computer Vision | GGUF/ONNX | `nexa run sd1-5`                       |
-| [stable-diffusion-v2-1](https://nexaai.com/StabilityAI/stable-diffusion-v2-1/gguf-q4_0/readme)          | Computer Vision | GGUF      | `nexa run sd2-1`                       |
-| [stable-diffusion-3-medium](https://nexaai.com/StabilityAI/stable-diffusion-3-medium/gguf-q4_0/readme)  | Computer Vision | GGUF      | `nexa run sd3`                         |
-| [FLUX.1-schnell](https://nexaai.com/BlackForestLabs/FLUX.1-schnell/gguf-q4_0/readme)                    | Computer Vision | GGUF      | `nexa run flux`                        |
-| [lcm-dreamshaper](https://www.nexaai.com/SimianLuo/lcm-dreamshaper-v7/gguf-fp16/readme)                 | Computer Vision | GGUF/ONNX | `nexa run lcm-dreamshaper`             |
-| [hassaku-lcm](https://nexaai.com/stablediffusionapi/hassaku-hentai-model-v13-LCM/gguf-fp16/readme)      | Computer Vision | GGUF      | `nexa run hassaku-lcm`                 |
-| [anything-lcm](https://www.nexaai.com/Linaqruf/anything-v30-LCM/gguf-fp16/readme)                       | Computer Vision | GGUF      | `nexa run anything-lcm`                |
-| [faster-whisper-tiny](https://www.nexaai.com/Systran/faster-whisper-tiny/bin-cpu-fp16/readme)           | Audio           | BIN       | `nexa run faster-whisper-tiny`         |
-| [faster-whisper-small](https://www.nexaai.com/Systran/faster-whisper-small/bin-cpu-fp16/readme)         | Audio           | BIN       | `nexa run faster-whisper-small`        |
-| [faster-whisper-medium](https://www.nexaai.com/Systran/faster-whisper-medium/bin-cpu-fp16/readme)       | Audio           | BIN       | `nexa run faster-whisper-medium`       |
-| [faster-whisper-base](https://www.nexaai.com/Systran/faster-whisper-base/bin-cpu-fp16/readme)           | Audio           | BIN       | `nexa run faster-whisper-base`         |
-| [faster-whisper-large](https://www.nexaai.com/Systran/faster-whisper-large-v3/bin-cpu-fp16/readme)      | Audio           | BIN       | `nexa run faster-whisper-large`        |
-| [whisper-large-v3-turbo](https://nexaai.com/Systran/faster-whisper-large-v3-turbo/bin-cpu-fp16/readme)  | Audio           | BIN       | `nexa run faster-whisper-large-turbo`  |
-| [whisper-tiny.en](https://nexaai.com/openai/whisper-tiny.en/onnx-cpu-fp32/readme)                       | Audio           | ONNX      | `nexa run whisper-tiny.en`             |
-| [whisper-tiny](https://nexaai.com/openai/whisper-tiny/onnx-cpu-fp32/readme)                             | Audio           | ONNX      | `nexa run whisper-tiny`                |
-| [whisper-small.en](https://nexaai.com/openai/whisper-small.en/onnx-cpu-fp32/readme)                     | Audio           | ONNX      | `nexa run whisper-small.en`            |
-| [whisper-small](https://nexaai.com/openai/whisper-small/onnx-cpu-fp32/readme)                           | Audio           | ONNX      | `nexa run whisper-small`               |
-| [whisper-base.en](https://nexaai.com/openai/whisper-base.en/onnx-cpu-fp32/readme)                       | Audio           | ONNX      | `nexa run whisper-base.en`             |
-| [whisper-base](https://nexaai.com/openai/whisper-base/onnx-cpu-fp32/readme)                             | Audio           | ONNX      | `nexa run whisper-base`                |
-| [mxbai-embed-large-v1](https://nexa.ai/mixedbread-ai/mxbai-embed-large-v1/gguf-fp16/readme)             | Embedding       | GGUF      | `nexa embed mxbai`                     |
-| [nomic-embed-text-v1.5](https://nexa.ai/nomic-ai/nomic-embed-text-v1.5/gguf-fp16/readme)                | Embedding       | GGUF      | `nexa embed nomic`                     |
-| [all-MiniLM-L6-v2](https://nexa.ai/sentence-transformers/all-MiniLM-L6-v2/gguf-fp16/readme)             | Embedding       | GGUF      | `nexa embed all-MiniLM-L6-v2:fp16`     |
-| [all-MiniLM-L12-v2](https://nexa.ai/sentence-transformers/all-MiniLM-L12-v2/gguf-fp16/readme)           | Embedding       | GGUF      | `nexa embed all-MiniLM-L12-v2:fp16`    |
+| [octopus-v2](https://www.nexaai.com/NexaAI/Octopus-v2/gguf-q4_0/readme) | NLP | GGUF | `nexa run octopus-v2` |
+| [octopus-v4](https://www.nexaai.com/NexaAI/Octopus-v4/gguf-q4_0/readme) | NLP | GGUF | `nexa run octopus-v4` |
+| [gpt2](https://nexaai.com/openai/gpt2/gguf-q4_0/readme) | NLP | GGUF | `nexa run gpt2` |
+| [tinyllama](https://www.nexaai.com/TinyLlama/TinyLlama-1.1B-Chat-v1.0/gguf-fp16/readme) | NLP | GGUF | `nexa run tinyllama` |
+| [llama2](https://www.nexaai.com/meta/Llama2-7b-chat/gguf-q4_0/readme) | NLP | GGUF/ONNX | `nexa run llama2` |
+| [llama2-uncensored](https://www.nexaai.com/georgesung/Llama2-7b-chat-uncensored/gguf-q4_0/readme) | NLP | GGUF | `nexa run llama2-uncensored` |
+| [llama2-function-calling](https://www.nexaai.com/Trelis/Llama2-7b-function-calling/gguf-q4_K_M/readme) | NLP | GGUF | `nexa run llama2-function-calling` |
+| [llama3](https://www.nexaai.com/meta/Llama3-8B-Instruct/gguf-q4_0/readme) | NLP | GGUF/ONNX | `nexa run llama3` |
+| [llama3.1](https://www.nexaai.com/meta/Llama3.1-8B-Instruct/gguf-q4_0/readme) | NLP | GGUF/ONNX | `nexa run llama3.1` |
+| [llama3.2](https://nexaai.com/meta/Llama3.2-3B-Instruct/gguf-q4_0/readme) | NLP | GGUF | `nexa run llama3.2` |
+| [llama3-uncensored](https://www.nexaai.com/Orenguteng/Llama3-8B-Lexi-Uncensored/gguf-q4_K_M/readme) | NLP | GGUF | `nexa run llama3-uncensored` |
+| [gemma](https://www.nexaai.com/google/gemma-1.1-2b-instruct/gguf-q4_0/readme) | NLP | GGUF/ONNX | `nexa run gemma` |
+| [gemma2](https://www.nexaai.com/google/gemma-2-2b-instruct/gguf-q4_0/readme) | NLP | GGUF | `nexa run gemma2` |
+| [qwen1.5](https://www.nexaai.com/Qwen/Qwen1.5-7B-Instruct/gguf-q4_0/readme) | NLP | GGUF | `nexa run qwen1.5` |
+| [qwen2](https://www.nexaai.com/Qwen/Qwen2-1.5B-Instruct/gguf-q4_0/readme) | NLP | GGUF/ONNX | `nexa run qwen2` |
+| [qwen2.5](https://www.nexaai.com/Qwen/Qwen2.5-1.5B-Instruct/gguf-q4_0/readme) | NLP | GGUF | `nexa run qwen2.5` |
+| [mathqwen](https://nexaai.com/Qwen/Qwen2.5-Math-1.5B-Instruct/gguf-q4_0/readme) | NLP | GGUF | `nexa run mathqwen` |
+| [codeqwen](https://www.nexaai.com/Qwen/CodeQwen1.5-7B-Instruct/gguf-q4_0/readme) | NLP | GGUF | `nexa run codeqwen` |
+| [mistral](https://www.nexaai.com/mistralai/Mistral-7B-Instruct-v0.3/gguf-q4_0/readme) | NLP | GGUF/ONNX | `nexa run mistral` |
+| [dolphin-mistral](https://www.nexaai.com/CognitiveComputations/dolphin-2.8-mistral-7b/gguf-q4_0/readme) | NLP | GGUF | `nexa run dolphin-mistral` |
+| [codegemma](https://www.nexaai.com/google/codegemma-2b/gguf-q4_0/readme) | NLP | GGUF | `nexa run codegemma` |
+| [codellama](https://www.nexaai.com/meta/CodeLlama-7b-Instruct/gguf-q2_K/readme) | NLP | GGUF | `nexa run codellama` |
+| [deepseek-coder](https://www.nexaai.com/DeepSeek/deepseek-coder-1.3b-instruct/gguf-q4_0/readme) | NLP | GGUF | `nexa run deepseek-coder` |
+| [phi2](https://www.nexaai.com/microsoft/Phi-2/gguf-q4_0/readme) | NLP | GGUF | `nexa run phi2` |
+| [phi3](https://www.nexaai.com/microsoft/Phi-3-mini-128k-instruct/gguf-q4_0/readme) | NLP | GGUF/ONNX | `nexa run phi3` |
+| [phi3.5](https://nexaai.com/microsoft/Phi-3.5-mini-instruct/gguf-q4_0/readme) | NLP | GGUF | `nexa run phi3.5` |
+| [openelm](https://nexaai.com/apple/OpenELM-3B/gguf-q4_K_M/readme) | NLP | GGUF | `nexa run openelm` |
+| [AMD-Llama-135m](https://nexaai.com/amd/AMD-Llama-135m/gguf-fp16/readme) | NLP | GGUF | `nexa run AMD-Llama-135m:fp16` |
+| [nanollava](https://www.nexaai.com/qnguyen3/nanoLLaVA/gguf-fp16/readme) | Multimodal | GGUF | `nexa run nanollava` |
+| [llava-phi3](https://www.nexaai.com/xtuner/llava-phi-3-mini/gguf-q4_0/readme) | Multimodal | GGUF | `nexa run llava-phi3` |
+| [llava-llama3](https://www.nexaai.com/xtuner/llava-llama-3-8b-v1.1/gguf-q4_0/readme) | Multimodal | GGUF | `nexa run llava-llama3` |
+| [llava1.6-mistral](https://www.nexaai.com/liuhaotian/llava-v1.6-mistral-7b/gguf-q4_0/readme) | Multimodal | GGUF | `nexa run llava1.6-mistral` |
+| [llava1.6-vicuna](https://www.nexaai.com/liuhaotian/llava-v1.6-vicuna-7b/gguf-q4_0/readme) | Multimodal | GGUF | `nexa run llava1.6-vicuna` |
+| [stable-diffusion-v1-4](https://www.nexaai.com/runwayml/stable-diffusion-v1-4/gguf-q4_0/readme) | Computer Vision | GGUF | `nexa run sd1-4` |
+| [stable-diffusion-v1-5](https://www.nexaai.com/runwayml/stable-diffusion-v1-5/gguf-q4_0/readme) | Computer Vision | GGUF/ONNX | `nexa run sd1-5` |
+| [stable-diffusion-v2-1](https://nexaai.com/StabilityAI/stable-diffusion-v2-1/gguf-q4_0/readme) | Computer Vision | GGUF | `nexa run sd2-1` |
+| [stable-diffusion-3-medium](https://nexaai.com/StabilityAI/stable-diffusion-3-medium/gguf-q4_0/readme) | Computer Vision | GGUF | `nexa run sd3` |
+| [FLUX.1-schnell](https://nexaai.com/BlackForestLabs/FLUX.1-schnell/gguf-q4_0/readme) | Computer Vision | GGUF | `nexa run flux` |
+| [lcm-dreamshaper](https://www.nexaai.com/SimianLuo/lcm-dreamshaper-v7/gguf-fp16/readme) | Computer Vision | GGUF/ONNX | `nexa run lcm-dreamshaper` |
+| [hassaku-lcm](https://nexaai.com/stablediffusionapi/hassaku-hentai-model-v13-LCM/gguf-fp16/readme) | Computer Vision | GGUF | `nexa run hassaku-lcm` |
+| [anything-lcm](https://www.nexaai.com/Linaqruf/anything-v30-LCM/gguf-fp16/readme) | Computer Vision | GGUF | `nexa run anything-lcm` |
+| [faster-whisper-tiny](https://www.nexaai.com/Systran/faster-whisper-tiny/bin-cpu-fp16/readme) | Audio | BIN | `nexa run faster-whisper-tiny` |
+| [faster-whisper-small](https://www.nexaai.com/Systran/faster-whisper-small/bin-cpu-fp16/readme) | Audio | BIN | `nexa run faster-whisper-small` |
+| [faster-whisper-medium](https://www.nexaai.com/Systran/faster-whisper-medium/bin-cpu-fp16/readme) | Audio | BIN | `nexa run faster-whisper-medium` |
+| [faster-whisper-base](https://www.nexaai.com/Systran/faster-whisper-base/bin-cpu-fp16/readme) | Audio | BIN | `nexa run faster-whisper-base` |
+| [faster-whisper-large](https://www.nexaai.com/Systran/faster-whisper-large-v3/bin-cpu-fp16/readme) | Audio | BIN | `nexa run faster-whisper-large` |
+| [whisper-large-v3-turbo](https://nexaai.com/Systran/faster-whisper-large-v3-turbo/bin-cpu-fp16/readme) | Audio | BIN | `nexa run faster-whisper-large-turbo` |
+| [whisper-tiny.en](https://nexaai.com/openai/whisper-tiny.en/onnx-cpu-fp32/readme) | Audio | ONNX | `nexa run whisper-tiny.en` |
+| [whisper-tiny](https://nexaai.com/openai/whisper-tiny/onnx-cpu-fp32/readme) | Audio | ONNX | `nexa run whisper-tiny` |
+| [whisper-small.en](https://nexaai.com/openai/whisper-small.en/onnx-cpu-fp32/readme) | Audio | ONNX | `nexa run whisper-small.en` |
+| [whisper-small](https://nexaai.com/openai/whisper-small/onnx-cpu-fp32/readme) | Audio | ONNX | `nexa run whisper-small` |
+| [whisper-base.en](https://nexaai.com/openai/whisper-base.en/onnx-cpu-fp32/readme) | Audio | ONNX | `nexa run whisper-base.en` |
+| [whisper-base](https://nexaai.com/openai/whisper-base/onnx-cpu-fp32/readme) | Audio | ONNX | `nexa run whisper-base` |
+| [mxbai-embed-large-v1](https://nexa.ai/mixedbread-ai/mxbai-embed-large-v1/gguf-fp16/readme) | Embedding | GGUF | `nexa embed mxbai` |
+| [nomic-embed-text-v1.5](https://nexa.ai/nomic-ai/nomic-embed-text-v1.5/gguf-fp16/readme) | Embedding | GGUF | `nexa embed nomic` |
+| [all-MiniLM-L6-v2](https://nexa.ai/sentence-transformers/all-MiniLM-L6-v2/gguf-fp16/readme) | Embedding | GGUF | `nexa embed all-MiniLM-L6-v2:fp16` |
+| [all-MiniLM-L12-v2](https://nexa.ai/sentence-transformers/all-MiniLM-L12-v2/gguf-fp16/readme) | Embedding | GGUF | `nexa embed all-MiniLM-L12-v2:fp16` |
 
 ## CLI Reference
 
@@ -334,6 +371,7 @@ Here's a brief overview of the main CLI commands:
 
 - `nexa run`: Run inference for various tasks using GGUF models.
 - `nexa onnx`: Run inference for various tasks using ONNX models.
+- `nexa convert`: Convert and quantize huggingface models to GGUF models.
 - `nexa server`: Run the Nexa AI Text Generation Service.
 - `nexa eval`: Run the Nexa AI Evaluation Tasks.
 - `nexa pull`: Pull a model from official or hub.