Image-Text-to-Text
MLX
Safetensors
mistral3
jangtq
mxtq
jang
apple-silicon
ministral3
pixtral
vision
multimodal
quantized
conversational
Instructions to use OsaurusAI/Mistral-Medium-3.5-128B-JANGTQ with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use OsaurusAI/Mistral-Medium-3.5-128B-JANGTQ with MLX:
# Make sure mlx-vlm is installed # pip install --upgrade mlx-vlm from mlx_vlm import load, generate from mlx_vlm.prompt_utils import apply_chat_template from mlx_vlm.utils import load_config # Load the model model, processor = load("OsaurusAI/Mistral-Medium-3.5-128B-JANGTQ") config = load_config("OsaurusAI/Mistral-Medium-3.5-128B-JANGTQ") # Prepare input image = ["http://images.cocodataset.org/val2017/000000039769.jpg"] prompt = "Describe this image." # Apply chat template formatted_prompt = apply_chat_template( processor, config, prompt, num_images=1 ) # Generate output output = generate(model, processor, formatted_prompt, image) print(output) - Notebooks
- Google Colab
- Kaggle
- Local Apps
- LM Studio
- Pi new
How to use OsaurusAI/Mistral-Medium-3.5-128B-JANGTQ with Pi:
Start the MLX server
# Install MLX LM: uv tool install mlx-lm # Start a local OpenAI-compatible server: mlx_lm.server --model "OsaurusAI/Mistral-Medium-3.5-128B-JANGTQ"
Configure the model in Pi
# Install Pi: npm install -g @mariozechner/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "mlx-lm": { "baseUrl": "http://localhost:8080/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "OsaurusAI/Mistral-Medium-3.5-128B-JANGTQ" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Hermes Agent new
How to use OsaurusAI/Mistral-Medium-3.5-128B-JANGTQ with Hermes Agent:
Start the MLX server
# Install MLX LM: uv tool install mlx-lm # Start a local OpenAI-compatible server: mlx_lm.server --model "OsaurusAI/Mistral-Medium-3.5-128B-JANGTQ"
Configure Hermes
# Install Hermes: curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash hermes setup # Point Hermes at the local server: hermes config set model.provider custom hermes config set model.base_url http://127.0.0.1:8080/v1 hermes config set model.default OsaurusAI/Mistral-Medium-3.5-128B-JANGTQ
Run Hermes
hermes
sidecar: config.json
Browse files- config.json +73 -0
config.json
ADDED
|
@@ -0,0 +1,73 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Mistral3ForConditionalGeneration"
|
| 4 |
+
],
|
| 5 |
+
"dtype": "bfloat16",
|
| 6 |
+
"image_token_index": 10,
|
| 7 |
+
"model_type": "mistral3",
|
| 8 |
+
"multimodal_projector_bias": false,
|
| 9 |
+
"projector_hidden_act": "gelu",
|
| 10 |
+
"spatial_merge_size": 2,
|
| 11 |
+
"text_config": {
|
| 12 |
+
"attention_dropout": 0.0,
|
| 13 |
+
"bos_token_id": 1,
|
| 14 |
+
"eos_token_id": 2,
|
| 15 |
+
"head_dim": 128,
|
| 16 |
+
"hidden_act": "silu",
|
| 17 |
+
"hidden_size": 12288,
|
| 18 |
+
"initializer_range": 0.02,
|
| 19 |
+
"intermediate_size": 28672,
|
| 20 |
+
"max_position_embeddings": 262144,
|
| 21 |
+
"model_type": "ministral3",
|
| 22 |
+
"num_attention_heads": 96,
|
| 23 |
+
"num_hidden_layers": 88,
|
| 24 |
+
"num_key_value_heads": 8,
|
| 25 |
+
"pad_token_id": 11,
|
| 26 |
+
"rms_norm_eps": 1e-05,
|
| 27 |
+
"rope_parameters": {
|
| 28 |
+
"beta_fast": 4.0,
|
| 29 |
+
"beta_slow": 1.0,
|
| 30 |
+
"factor": 64.0,
|
| 31 |
+
"llama_4_scaling_beta": 0,
|
| 32 |
+
"mscale": 1.0,
|
| 33 |
+
"mscale_all_dim": 1.0,
|
| 34 |
+
"original_max_position_embeddings": 4096,
|
| 35 |
+
"rope_theta": 1000000.0,
|
| 36 |
+
"rope_type": "yarn",
|
| 37 |
+
"type": "yarn"
|
| 38 |
+
},
|
| 39 |
+
"sliding_window": null,
|
| 40 |
+
"tie_word_embeddings": false,
|
| 41 |
+
"use_cache": true,
|
| 42 |
+
"vocab_size": 131072
|
| 43 |
+
},
|
| 44 |
+
"tie_word_embeddings": false,
|
| 45 |
+
"transformers_version": "5.6.0.dev0",
|
| 46 |
+
"vision_config": {
|
| 47 |
+
"attention_dropout": 0.0,
|
| 48 |
+
"head_dim": 104,
|
| 49 |
+
"hidden_act": "silu",
|
| 50 |
+
"hidden_size": 1664,
|
| 51 |
+
"image_size": 1540,
|
| 52 |
+
"initializer_range": 0.02,
|
| 53 |
+
"intermediate_size": 8192,
|
| 54 |
+
"model_type": "pixtral",
|
| 55 |
+
"num_attention_heads": 16,
|
| 56 |
+
"num_channels": 3,
|
| 57 |
+
"num_hidden_layers": 48,
|
| 58 |
+
"patch_size": 14,
|
| 59 |
+
"rope_parameters": {
|
| 60 |
+
"rope_theta": 10000.0,
|
| 61 |
+
"rope_type": "default"
|
| 62 |
+
}
|
| 63 |
+
},
|
| 64 |
+
"vision_feature_layer": -1,
|
| 65 |
+
"quantization": {
|
| 66 |
+
"group_size": 64,
|
| 67 |
+
"bits": 8
|
| 68 |
+
},
|
| 69 |
+
"weight_format": "mxtq",
|
| 70 |
+
"mxtq_seed": 42,
|
| 71 |
+
"mxtq_bits": 2,
|
| 72 |
+
"routed_expert_bits": 2
|
| 73 |
+
}
|